Search

International audience; We continue studies of the uncertainty quantification problem in emission tomographies such aspositron emission tomography (PET) or single photon emission computed tomography (SPECT)when additional multimodal data (anatomical magnetic resonance imaging (MRI) images) are avail-able. To solve the aforementioned problem we adapt the recently proposed nonparametric posteriorlearning technique to the context of Poisson-type data in emission tomography. Using this approachwe derive sampling algorithms which are trivially parallelizable, scalable and very easy to implement.In addition, we prove conditional consistency and tightness for the distribution of produced samplesin the small noise limit (i.e., when the acquisition time tends to infinity) and derive new geometricaland necessary condition on how MRI images must be used. This condition arises naturally in thecontext of identifiability problem for misspecified generalized Poisson models with wrong design. Wealso contrast our approach with Bayesian Markov chain Monte Carlo sampling based on one dataaugmentation scheme which is very popular in the context of expectation-maximization algorithmsfor PET or SPECT. We show theoretically and also numerically that such data augmentation sig-nificantly increases mixing times for the Markov chain. In view of this, our algorithms seem to givea reasonable trade-off between design complexity, scalability, numerical load and assessment for theuncertainty

While a myriad of effective techniques exist to aid in symphyseal age estimation for those 40 years and younger, few offer similar levels of efficacy for those beyond that threshold. Through the application of a novel technique, this study sought to determine whether a closer inspection of degenerative change may help to improve precision in age estimation for post-epiphyseal adults. Results show that the combination of five distinct areas of interest, plus a correction for density, accurately estimated age 87.75% of the time (averaged amongst four observers [spread: 72–100%]) for a subset of 50 living British females. An adjusted R2 value of 0.85, an RSME value of 5.62 years, and a PCC value of 0.92 also confirmed the trialed technique to be a good predictor of age for the entirety of the larger female sample (n = 533). Low inaccuracy (3.86 years) and Bias (0.69 years) further indicate that a continuum-based approach, without pre-set phases or ranges, such as was utilized by this research holds the potential to be at least as effective as the currently available methodologies but with the added advantage of allowing for increased variation at the individual level. Age estimation by linear regression, or by simple addition, yielded estimation envelopes (intervals) of 22–23 and 24 years, respectively, which remain narrow enough to be forensically useful while still wide enough to maximize accuracy in mature adults.

[EN] The increased accuracy and exhaustivity of modern Artificial Intelligence techniques in supporting the analysis of complex data, such as medical images, have exponentially increased real-world data collection for research purposes. This fact has led to the development of international repositories and high-performance computing solutions to deal with the computational demand for training models. However, other stages in the development of medical imaging biomarkers do not require such intensive computing resources, which has led to the convenience of integrating different computing backends tailored for the processing demands of the various stages of processing workflows. We present in this article a distributed and federated repository architecture for the development and application of medical image biomarkers that combines multiple cloud storages with cloud and HPC processing backends. The architecture has been deployed to serve the PRIMAGE (H2020 826494) project, aiming to collect and manage data from paediatric cancer. The repository seamlessly integrates distributed storage backends, an elastic Kubernetes cluster on a cloud on-premises and a supercomputer. Processing jobs are handled through a single control platform, synchronising data on demand. The article shows the specification of the different types of applications and a validation through a use case that make use of most of the features of the platform., The work presented in this paper has been partially funded by the project PRIMAGE (PRedictive In-silico Multiscale Analytics to support cancer persona-lised diaGnosis and prognosis, empowered by imaging biomarkers) Business Place is a Horizon 2020, Research and Innovation Actions (RIA) (Topic SC1-DTH-07-2018) with grant agreement no: 826494. This research was supported in part by PLGrid Infrastructure.

Objective To determine the diagnostic accuracy of Necrotizing Otitis Externa (NOE) based on radiologic studies. Methods The PubMed, Cochrane, Embase, Web of Science, SCOPUS, and Google Scholar databases were searched. True-positive and false-negative results were extracted for each study. Methodological quality was evaluated using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) tool. Results The included studies contained data on 37 studies diagnosed with NOE. The sensitivity of gallium-67, technetium-99m, and Magnetic Resonance Imaging (MRI) was 0.9378 (0.7688–0.9856), 0.9699 (0.8839–0.9927), and 0.9417 (0.6968–0.9913), respectively. For Computed Tomography (CT), the positive criteria consisted of bony erosion alone and bony erosion plus any soft tissue abnormality. The sensitivity of CT based only on bony erosion was 0.7062 (0.5954–0.7971); it was higher 0.9572 (0.9000–0.9823) when based on bony erosion plus any soft tissue abnormality. Conclusion The diagnostic sensitivity of technetium-99m, gallium-67, and MRI was favorable. On CT, the presence of bony erosion may be a useful diagnostic marker of NOE, but the diagnostic sensitivity will be even higher if the criterion of any soft tissue abnormality is also included; however, care should be taken when interpreting the results. Our study demonstrates the potential utility of radiology studies for diagnosing NOE, but their lack of specificity must be considered, and standardized anatomic criteria are still needed. Level of evidence 2A.

Time-to-digital converters (TDCs) are high-performance mixed-signal circuits capable of timestamping events with sub-gate delay resolution. As a result of their high-performance, in recent years TDCs were integrated in complementary metal-oxide-semiconductor (CMOS) technology with highly sensitive photodetectors known as single-photon avalanche diodes (SPADs), to form digital silicon photomultipliers (dSiPMs) and SPAD imagers. Time-resolved SPAD-based sensors are capable of detecting the absorption of a single photon and timestamping it with picosecond resolution. As such, SPAD-based sensors are very useful in the field of biomedical imaging, using time-of-flight (ToF) information to produce data that can be used to reconstruct high-quality biological images. Additionally, the capability of integration in standard CMOS technologies, allows SPAD-based sensors to provide high-performance, while maintaining low cost. In this paper, we present an overview of fundamental TDC principles, and an analysis of state-of-the-art TDCs. Furthermore, the integration of TDCs into dSiPMs and SPAD imagers will be discussed, with an analysis of the current results of TDCs in different biomedical imaging applications. Finally, several important research challenges for TDCs in biomedical imaging applications are presented.

Cancer is one of the world's most critical health issues. Many in the past decade More accurately, diagnostic tests and methodologies have been enhanced. Tests are categorised into imaging tests, endoscopic procedures Includes testing for biopsy and cytology. These diagnostic tests yield huge numbers. volumes of data that must be evaluated and differentiated by specialists Between tumours that are benign and malignant. Artificial intelligence now (AI) Large quantities of diagnostic imaging may be examined using applications Improved accuracy to enhance health system efficiency. New AI algorithms technical advances and computer enhancements Hardware may be used to train diagnostic neural artificial networks Experience with a wide range of scans. Already a thorough understanding and Machines can educate computers to compare and assess models Huge quantity of cancer scanning data. Highly diagnostic skills Software has been evaluated and compared to the conventional Cancer specialists' diagnostic tools. Their precision is much increased and regarded highly effective in early diagnosis and extended forecasting Different cancers. Scientists have become systems of artificial intelligence (AI). that may exceed human specialists in the prognosis of breast cancer and A great deal earlier diagnosis. Similarly, informatics developed an AI algorithm. and profound learning algorithms that may forecast to which persons Develop lung cancer using low dose CT analysis (computerized tomography) Lung scans. Lung scans. Use of the convolutionary neural network recently (CNNs) was employed in the invasion depth diagnosis of Gastric endoscopy-based gastric cancer. Studies have also been shown AI techniques paired with imagery may have a significant influence early diagnosis with algorithm-guided identification of oral cancer results Heterogeneity of the oral lesion. This review gathered some pretty interesting information Research publications on the topic.

A 69-year-old man with dementia presented to our hospital with fever and left back pain. He was diagnosed with acute pyelonephritis based on microscopy and CT findings indicating bacteriuria and elevated fat density around the right kidney, respectively. Three days post admission, the diagnosis of a urinary tract infection was queried and the patient was re-examined. Marked tenderness was observed in the left buttock; contrast CT scan was subsequently performed and a left piriformis abscess was diagnosed. This diagnostic delay occurred due to a communication error—the referral letter to the radiologist stated ‘fever and back pain’—and the physician’s hyposkillia; hence, differential diagnoses were not considered. Physicians can reduce the frequency of related errors by conducting proper physical examinations and not relying on imaging findings alone. Furthermore, to ensure accurate diagnoses, the comments to the radiologist should be clear and detailed when requesting diagnostic imaging.

Different types of medical imaging are increasingly being used to explain specific aspects of injuries to patients during consultations. However, there are no validated questionnaires available yet that specifically measure patients’ satisfaction with the use of such images. The objective of this study was to develop and evaluate a patient-centred measure of satisfaction with the use of medical imaging modalities in clinical treatment. A preliminary 22-item version of the Questionnaire for Patient Satisfaction with Imaging (QPSI) was developed based on extant literature and interviews with trauma patients. Final item selection and psychometric evaluations were conducted amongst a sample of 106 hospital trauma patients who were shown medical images of their injuries. The psychometric analyses resulted in a final 13-item questionnaire comprising two subscales that measure the importance of seeing the images (9 items) and the clearness of the image (4 items). Both subscales showed adequate internal consistency (α = 0.84 and 0.75). The subscales were weakly intercorrelated (ρ = 0.34) and were both significantly and independently associated with patients’ global ratings of satisfaction with the use of imaging. The final two-dimensional QPSI is an innovative, reliable and valid questionnaire for measuring patients’ satisfaction with imaging-based information during clinical consultations.

Saliency methods, which “explain” deep neural networks by producing heat maps that highlight the areas of the medical image that influence model prediction, are often presented to clinicians as an aid in diagnostic decision-making. Although many saliency methods have been proposed for medical imaging interpretation, rigorous investigation of the accuracy and reliability of these strategies is necessary before they are integrated into the clinical setting. In this work, we quantitatively evaluate seven saliency methods—including Grad-CAM, Grad-CAM++, and Integrated Gradients—across multiple neural network architectures using two evaluation metrics. We establish the first human benchmark for chest X-ray segmentation in a multilabel classification set up, and examine under what clinical conditions saliency maps might be more prone to failure in localizing important pathologies compared to a human expert benchmark. We find that (i) while Grad-CAM generally localized pathologies better than the other evaluated saliency methods, all seven performed significantly worse compared with the human benchmark; (ii) the gap in localization performance between Grad-CAM and the human benchmark was largest for pathologies that were smaller in size and had shapes that were more complex; (iii) model confidence was positively correlated with Grad-CAM localization performance. While it is difficult to know whether poor localization performance is attributable to the model or to the saliency method, our work demonstrates that several important limitations of saliency methods must be addressed before we can rely on them for deep learning explainability in medical imaging.

We propose a strategy for magnetic navigation of collective stem cell microrobots in blood environments under ultrasound Doppler imaging. The cell microrobots are fabricated through a coculture process of stem cells and iron microparticles, they have spheroidal structures and are actuated under external magnetic fields. The collective cell microrobots can be reversibly gathered and spread by adjusting the magnetic interaction, and these microrobots are able to exhibit collective motion in whole blood under rotating magnetic fields. Simulation results indicate that the induced blood flow around the collective pattern affects the motion of red blood cells, and experimental results show that Doppler signals are observed when emitting ultrasound waves to the microrobots. Due to the induced three-dimensional blood flow, Doppler signals can also be observed when the imaging plane is above the collective microrobots, which enables indirect localization when performing navigation on uneven surfaces. Moreover, collective patterns of different numbers of microrobots are formed in flowing conditions, and these collective biohybrid agents are actively navigated against flowing blood with a flow rate up to 4.5 mL/min (mean velocity: 5.97 mm/s). Our study investigates a strategy for pattern formation and navigation of collective microrobots under ultrasound Doppler imaging, demonstrating the integration of collective control approach and medical imaging holds great potential for real-time active delivery tasks.

Diabetic Retinopathy (DR) is the condition caused due to uncontrolled diabetes that can lead to vision impairment. It greatly affects the retinal blood vessels and diminishes the fundus light-sensitive inner coating. Early diagnosis and regular screening of this disease are essential for prompt processing through artificial intelligence techniques. This paper targets assessing the latest techniques for screening and diagnosing DR, including 94 articles based on the Detection and grading of DR. For every analyzed approach, tables are summarized detailing imaging procedure used, datasets, performance metrics used. The research gaps are also highlighted in this paper. Despite the consistent progression and methods actualized in this field, a couple of issues actually should be centered on. The noise and contrast of the image in Image enhancement are still in the infancy stage for high resolution. This study covers a review of existing image techniques, the gold standard and private datasets available, performance measures used for detection and grading of DR. Now the future research focuses on the amalgamation of the dataset as well as techniques to make the generalized technique for detecting the lesion in DR through an automated system. Moreover, various research gaps have also been taken into account for further research. This review is beneficial to the researchers working in the field of medical imaging to screen and diagnose diseases.

Rationale and Objectives: The temporomandibular joint (TMJ) is commonly involved in children with juvenile idiopathic arthritis. The diagnosis and evaluation of the disease progression is dependent on medical imaging. The precision of this imaging is under debate. SeveralR scoring systems have been proposed but transparent testing of the precision of the constituents of the scoring systems is lacking. The present study aims to test the precision of 25 imaging features based on magnetic resonance imaging (MRI). Materials and Methods: Clinical data and imaging were obtained from the Norwegian juvenile idiopathic arthritis study, The NorJIA study. Twenty-five imaging features of the TMJ in MRI datasets from 86 study participants were evaluated by two experienced radiologists for inter- and intraobserver agreement. Agreement of ordinal variables was measured with Cohen s linear or weighted Kappa as appropriate. Agreement of continuous measurements was assessed with 95% limit of agreement according to Bland-Altman. Results: In the osteochondral domain, the ordinal imaging variables “loss of condylar volume,” “condylar shape,” “condylar irregularities,” “shape of the eminence/fossa,” “disk abnormalities,” and “condylar inclination” showed inter- and intraobserver agreement above Kappa 0.5. In the inflammatory domain, the ordinal imaging variables “joint fluid,” “overall impression of inflammation,” “synovial enhancement” and “bone marrow oedema” showed inter- and intraobserver agreement above Kappa 0.5. Continuous measurements performed poorly with wide limits of agreement. Conclusion: A precise MRI-based scoring system for assessment of TMJ in JIA is proposed consisting of seven variables in the osteochondral domain and four variables in the inflammatory domain. Further testing of the clinical validity of the variables is needed.

The extensive growth of data in the health domain has increased the utility of Deep Learning in health. Deep learning is a highly advanced successor of artificial neural networks, having powerful computing ability. Due to the availability of fast data storage and hardware parallelism its popularity grows in the last five years. This in article presents a comprehensive literature review of research deploying deep learning medical imaging and medical NLP including tasks, pipelines, and challenges. In this work, we have presented an extensive survey of deep learning architecture deployed in the fields of medical imaging and medical natural language processing. This paper helps in identifying suitable combination of Deep learning, Natural language processing and medical imaging to enhance diagnosis. We have highlighted the major challenges in deploying deep learning in medical imaging and medical natural language processing. All the results are presented in pictorial form. This survey is very helpful for novices working in the area of health informatics.

Objective In this research paper, the brain MRI images are going to classify by considering the excellence of CNN on a public dataset to classify Benign and Malignant tumors. Materials and Methods Deep learning (DL) methods due to good performance in the last few years have become more popular for Image classification. Convolution Neural Network (CNN), with several methods, can extract features without using handcrafted models, and eventually, show better accuracy of classification. The proposed hybrid model combined CNN and support vector machine (SVM) in terms of classification and with threshold-based segmentation in terms of detection. Result The findings of previous studies are based on different models with their accuracy as Rough Extreme Learning Machine (RELM)-94.233%, Deep CNN (DCNN)-95%, Deep Neural Network (DNN) and Discrete Wavelet Autoencoder (DWA)-96%, k-nearest neighbors (kNN)-96.6%, CNN-97.5%. The overall accuracy of the hybrid CNN-SVM is obtained as 98.4959%. Conclusion In today's world, brain cancer is one of the most dangerous diseases with the highest death rate, detection and classification of brain tumors due to abnormal growth of cells, shapes, orientation, and the location is a challengeable task in medical imaging. Magnetic resonance imaging (MRI) is a typical method of medical imaging for brain tumor analysis. Conventional machine learning (ML) techniques categorize brain cancer based on some handicraft property with the radiologist specialist choice. That can lead to failure in the execution and also decrease the effectiveness of an Algorithm. With a brief look came to know that the proposed hybrid model provides more effective and improvement techniques for classification.

BACKGROUND: Low frequency vibrations from motorized vehicles and heavy equipment have been associated with musculoskeletal disorders. Spine degeneration on diagnostic imaging provides direct and objective measures of the possible effects of such exposures on the spine. OBJECTIVE: The objective of this systematic review was to evaluate the association of exposure to whole-body vibration (WBV) with spine degeneration on imaging. METHODS: We conducted electronic searches in MEDLINE, CINAHL, EMBASE, and Web of Science to July 2021. Two reviewers independently screened search results, assessed quality, and extracted data. Studies evaluating the exposure to WBV and lumbar spine degeneration on imaging were included. RESULTS: Fifteen studies (16 manuscripts) were included. Seven studies including a meta-demonstrated moderate quality evidence of no association between WBV and disc degeneration. There was also moderate quality evidence of no association between WBV and disc height narrowing and osteophytes. Overall, there was low level evidence of no association between WBV and other degenerations findings. CONCLUSIONS: There was moderate to low quality evidence suggesting no association between WBV exposures with spine degeneration on imaging. The results of this study currently do not support assertion that motorized vehicle and WBV exposure accelerates degeneration and causes structural damage to the spine.

Hundreds of imaging-based artificial intelligence (AI) models have been developed in response to the COVID-19 pandemic. AI systems that incorporate imaging have shown promise in primary detection, severity grading, and prognostication of outcomes in COVID-19, and have enabled integration of imaging with a broad range of additional clinical and epidemiologic data. However, systematic reviews of AI models applied to COVID-19 medical imaging have highlighted problems in the field, including methodologic issues and problems in real-world deployment. Clinical use of such models should be informed by both the promise and potential pitfalls of implementation. How does a practicing radiologist make sense of this complex topic, and what factors should be considered in the implementation of AI tools for imaging of COVID-19? This critical review aims to help the radiologist understand the nuances that impact the clinical deployment of AI for imaging of COVID-19. We review imaging use cases for AI models in COVID-19 (e.g., diagnosis, severity assessment, and prognostication) and explore considerations for AI model development and testing, deployment infrastructure, clinical user interfaces, quality control, and institutional review board and regulatory approvals, with a practical focus on what a radiologist should consider when implementing an AI tool for COVID-19.

Background Evaluation of acute appendicitis (AA) in pregnancy is supported with diagnostic imaging. Typically, ultrasound (US) is performed first, and then often followed by magnetic resonance imaging (MRI) due to continued diagnostic uncertainty. The purpose of our study was to evaluate the sensitivity, specificity, and accuracy of US as compared to MRI and to evaluate the inter-radiologist agreement amongst body Radiologists with varying levels of expertise. Methods We performed a retrospective study of 364 consecutive pregnant patients with clinical suspicion of AA at a single center over a 6-year period. Sensitivity, Specificity, accuracy, positive and negative predictive values were calculated for US and MRI. Inter Radiologist agreement was determined using Cohen's Kappa analysis between original interpreting Radiologist and retrospective review by expert Radiologist. Results 31 of 364 patients (8.5%) underwent appendectomy based on preoperative diagnosis, with confirmation of acute appendicitis (AA) by pathology in 19. US was able to visualize the appendix in only 6 (1.65%), 5 of whom had appendicitis. 141 patients underwent MRI, and correctly diagnosed appendicitis in 9. No patient with a negative MRI diagnosis had AA. The sensitivity, and negative predictive value for diagnosing AA with MRI was 100%. The MRI inter-reader agreement for appendix visualization and overall accuracy were 87.9 and 98% with Cohen Kappa of 0.7 and 0.56 respectively. Conclusion Our data suggests that MRI should be considered the first line imaging modality in pregnant patients suspected of having AA. Body Radiologists with varied levels of experience in MRI readouts had substantial agreement.

Background: A variety of emerging medical imaging technologies based on artificial intelligence have been widely applied in many diseases, but they are still limitedly used in the cerebrovascular field even though the diseases can lead to catastrophic consequences. Objective: This work aims to discuss the current challenges and future directions of artificial intelligence technology in cerebrovascular diseases through reviewing the existing literature related to applications in terms of computer-aided detection, prediction and treatment of cerebrovascular diseases. Methods: Based on artificial intelligence applications in four representative cerebrovascular diseases including intracranial aneurysm, arteriovenous malformation, arteriosclerosis and moyamoya disease, this paper systematically reviews studies published between 2006 and 2021 in five databases: National Center for Biotechnology Information, Elsevier Science Direct, IEEE Xplore Digital Library, Web of Science and Springer Link. And three refinement steps were further conducted after identifying relevant literature from these databases. Results: For the popular research topic, most of the included publications involved computer-aided detection and prediction of aneurysms, while studies about arteriovenous malformation, arteriosclerosis and moyamoya disease showed an upward trend in recent years. Both conventional machine learning and deep learning algorithms were utilized in these publications, but machine learning techniques accounted for a larger proportion. Conclusion: Algorithms related to artificial intelligence, especially deep learning, are promising tools for medical imaging analysis and will enhance the performance of computer-aided detection, prediction and treatment of cerebrovascular diseases.

Covid-19 lockdowns dramatically accelerated demand for companion family animals. But increased selective breeding of flat-faced dogs has led to a crisis in associated neurological, skeletal, and airway disorders, where canine quality of life is inadvertently sacrificed for cuteness in appearance. It is suggested that some physical traits are more likely to be found in pedigree dogs afflicted with several genetic developmental disorders, and the exaggeration of these traits worsen the severity of such disorders. However, identifying and grading these traits is impractical without large-scale medical imaging and invasive surgical procedures. A database comprising CT scans obtained from Cavalier King Charles Spaniels was provided to this study, from which cranial models can be generated with computer vision software. A novel low-cost AI methodology has been developed to identify key physical characteristics, present in crania, associated with genetic developmental diseases affecting the Cavalier. Early AI-led results found a significant bulge on the top of the skull linked to neurological disease with near-perfect sensitivity. Continuing developments of this methodology will assist breeders to better develop sustainable, ethical breeding practices for at-risk pedigree dogs and contribute to reducing quality of life issues arising from genetic developmental disorders.&nbsp

— Dermatology is the branch of bioscience that deals with treating and diagnosing skin disorders and allergies that can be caused due to various geographical and biological factors. However, human skin is a complex and unpredictable domain to define, synthesize and analyze and to rather to distinguish disorders affecting with skin. To address this issue, Researchers are still working to figure out and model human skin behaviour to automate the process of skin disease detecting through victimization that uses vision techniques to meet the criteria of medical science. Patients in remote areas where access to quality medical facilities is still limited may ignore symptoms and exacerbate their conditions. Automated medical paradigms are necessary to effectively address these issues and provide convenient and accessible solutions to people's needs. The emergence of automated paradigms for detecting skin diseases is a promising solution to ensure accurate diagnosis of skin-related issues. A multiclass deep learning model has been proposed to develop that can differentiate between healthy skin and skin affected by a disease, and also categorize various types of skin diseases. We used Deep Learning, a type of machine learning that uses large datasets to reduce the number of classifiers required to train our model. This approach lets the machine to learn and divide the provided data into levels of predictions and in a very short period of time gives delivers accurate results, thus our work aims to promote and support development of Dermatology. The algorithm we used here is Convolutional Neural Network (CNN) as it is one of the most preferred algorithms for image classification.&nbsp

Deep learning models for radiology are typically deployed either through cloud-based platforms, through on-premises infrastructures, or though heavyweight viewers. This tends to restrict the audience of deep learning models to radiologists working in state-of-the-art hospitals, which raises concerns about the democratization of deep learning for medical imaging, most notably in the context of research and education. We show that complex deep learning models can be applied directly inside Web browsers, without resorting to any external computation infrastructure, and we release our code as free and open-source software. This opens the path to the use of teleradiology solutions as an effective way to distribute, teach, and evaluate deep learning architectures.

Presentation "Artificial intelligence (deep learning) for image synthesis" at the session "AI for (online?) IGRT" held onSunday May 14 15:15 - 16:30, Strauss 1 at ESTRO2023, along with an "Overview of deep learning-based products in radiotherapy: segmentation and synthetic CT". Versioning Version 0.1presentation + 1.0 overview product, released for ESTRO 2023, Vienna, 11-16 May 2023 on 13-05-2023. &nbsp

Introduction Both my STS research project and technical capstone focused on the implementation of participatory design in the software development process. My STS research project took a much larger view of a more complicated area of software development, while my technical report focused on my specific experiences with applying this framework at my internship from this past summer. As an intern at the time, the framework of participatory design was not mentioned by name, but after learning about it during my STS classes this year, I was able to reflect on the project and see how did apply it in many ways to our project. In particular, I realized the emphasis we placed on engaging with the different teams within the company that our application was going to be used by. Our design process lasted half of the summer, with the other half dedicated to development. STS Project My STS project centers on the issues that have arisen and will continue to arise from the application of machine learning image recognition algorithms to medical imaging analysis. Specifically, I highlight issues of bias, trust, and data privacy. For example, minority groups are, as they are in the general population, often underrepresented in datasets used to train algorithms, leading to decreased accuracy for their images. I talk about how algorithms like these are notoriously hard to tweak or change, due to their black-box nature and their unsupervised training process. To address these issues, I discuss how principles of the framework participatory design could be applied. I argue that by following these principles, many practical solutions can be reached such as increasing representation on development teams, gathering balanced datasets, and involving doctors and patients in the design process of the algorithms. While the framework is not a catch-all solution for the ethical concerns explored, I argue that it can be a positive step towards solving many of them. Technical Report Last summer, a large McLean, Virginia-based financial company decided to streamline the complicated, manual process of adding new data models that wasted tons of man-hours. As an intern, I worked on a team was tasked with creating an enterprise application to onboard new data models and view existing models. To accomplish this goal, we built a full-stack application using a combination of Python and Vue.js, that would be deployed on AWS and connected to a DynamoDB database. We spent an extensive time designing the application, involving all affected teams in the process. At the end of the summer, our application was deployed to a development environment and was able to perform the basic features we outlined. To bring the application into use, the database needs to be fully populated with the necessary data, and the application needs to be deployed through the pipeline into a production environment. Conclusion Working on these projects at the same time has been extremely valuable to me, as it allowed me to have insights on their connectedness that I otherwise would not have had. If I had not decided to work on participatory design, my technical report would have been a much more dry and technical paper focusing strictly on the development of our application. Looking back over my experiences through the perspective of this framework allowed me to better understand what things my team did that were successful and why they were successful. At the time, my team’s extended design process without getting to start coding frustrated me, but now I realize that process was invaluable in creating an application that meets the needs of those intended to engage with it. These realizations will inform my future work of developing software.

A key challenge in training neural networks for a given medical imaging task is often the difficulty of obtaining a sufficient number of manually labeled examples. In contrast, textual imaging reports, which are often readily available in medical records, contain rich but unstructured interpretations written by experts as part of standard clinical practice. We propose using these textual reports as a form of weak supervision to improve the image interpretation performance of a neural network without requiring additional manually labeled examples. We use an image-text matching task to train a feature extractor and then fine-tune it in a transfer learning setting for a supervised task using a small labeled dataset. The end result is a neural network that automatically interprets imagery without requiring textual reports during inference. This approach can be applied to any task for which text-image pairs are readily available. We evaluate our method on three classification tasks and find consistent performance improvements, reducing the need for labeled data by 67%-98%., This work is accepted to the IEEE Journal of Biomedical and Health Informatics

Design of the SynthRAD2023 Grand Challengehttps://synthrad2023.grand-challenge.org/ following the structured challenge submission provided at https://www.biomedical-challenges.org/. The challenge will be part of MICCAI 2023. Versions v0.1design submitted on 4-11-2022 to MICCAI; v0.2 revised on 02-2023 as part of the rebuttal; v0.3automatically updated by MICCAI: https://doi.org/10.5281/zenodo.7835406; v1.0textually updated on 03-2023 according to the approved submitted changes by MICCAI (Zenodo:10.5281/zenodo.7746020); v1.1 - latesttextual revision + update timeline figure, added abstract and reference to MICCAI automatically generated Zenodo; Abstract The impact of medical imaging on oncological patients' diagnosis and therapy, has grown significantly over the last decades. Especially in radiotherapy (RT), imaging plays a crucial role in the entire workflow, from treatment simulation to patient positioning and monitoring. Traditionally, 3D computed tomography (CT) is considered the primary imaging modality in RT, providing accurate and high-resolution patient geometry and enabling direct electron density conversion needed for dose calculations and plan optimization [Chernak et al., 1975]. X-ray-based imaging has been widely adopted for patient positioning and monitoring before, during, and after dose delivery. Recently, 3D cone-beam computed tomography (CBCT) - often integrated with the dose delivery machine - has been widely adopted, playing a vital role in traditional and more advanced image-guided adaptive radiation therapy (IGART) workflows in both photon and proton therapy. A key challenge in using CBCT is that several artifacts, such as shading, streaking, and cupping, affect image reconstruction. As a result, CBCT quality is insufficient to perform accurate dose calculations or replanning. Consequently, patients must be referred to a rescan CT when anatomical differences are noted between daily images and the planning CT [Ramella et al., 2017]. As an alternative, image synthesis has been proposed to improve the quality of CBCT to the CT level, producing the so-called “synthetic CT” (sCT) [Kida et al., 2018]. Additionally, CBCT-based sCT allows online adaptive CBCT-based RT workflows to improve the quality of IGART provided to the patients. In parallel, over the last decades, magnetic resonance imaging (MRI) has also proved its added value for tumor and organs-at-risk delineation thanks to its superb soft-tissue contrast [Schmidt et al., 2015]. MRI can be acquired to match patient positioning to the planned one and monitor changes before, during, or after the dose delivery [Lagendijk et al., 2004]. To benefit from the complementary advantages offered by different imaging modalities, MRI is generally registered to CT. Such a workflow requires obtaining both CT and MRI, increasing workload, and introducing additional radiation to the patient. Recently, MRI-only based RT has been proposed to simplify and speed up the workflow, decreasing patients' exposure to ionizing radiation. This is particularly relevant for repeated simulations or fragile populations like children. MRI-only RT may reduce treatment costs and workload and eliminate residual registration errors using both imaging modalities. Additionally, MRI-only techniques can benefit MRI-guided RT [Edmund and Nyholm, 2017]. The main obstacle in introducing MRI-only RT is the lack of tissue attenuation information required for accurate dose calculations. Many methods have been proposed to convert MR to CT-equivalent images, yielding sCTs suitable for treatment planning and dose calculation. Artificial intelligence algorithms such as machine learning or deep learning have become the best-performing methods for deriving sCT from MRI or CBCT. With many algorithms available, all tested on different datasets, it is unclear which algorithms are better than others. Unfortunately, no public datasets or challenges have been designed to benchmark and compare different approaches. A recent review of deep learning-based sCT generation also advocated for public challenges to provide data and evaluation metrics for such open comparison [Spadea & Maspero et al., 2021]. We now designed a challenge to provide the first platform offering public datasets and evaluation metrics to benchmark and compare the latest algorithms in sCT generation. For this purpose, two tasks were defined: 1) MRI-to-sCT generation to facilitate MRI-only RT and 2) CBCT-to-sCT generation to facilitate IGART and online adaptive RT. A multi-center dataset of matched input (CBCT or MRI) and target (CT) image pairs with heterogeneous acquisition protocols will be divided into balanced training, validation, and test sets. We will share the inputs and targets from the training set to design and evaluate algorithms generating sCT. In the first evaluation round, called validation, input images will be shared. The participating team can upload their outputs to validate their performance and see how they rank on the leaderboard on image similarity metrics. Finally, independent test cases will be shared to generate sCT for our final evaluation. Brain and pelvic datasets from three Dutch centers (UMC Utrecht, UMC Groningen, and Radboud Nijmegen) have been collected, providing more than 500 patients for MRI-to-CT (Task 1) and 500 patients for CBCT-to-CT (Task 2) undergoing radiotherapy treatments in the corresponding departments. The challenge runs onhttps://synthrad2023.grand-challenge.org/, and the pre-processing and evaluation code used to rank the submissions based on image and dose evaluations will be shared. Challenge participants may choose to participate only in task 1 (MRI-to-sCT), only in task 2 (CBCT-to-sCT), or in both, for both anatomical regions. In the initial training/validation phases, training and validation data will be made available to submit to an open leaderboard to enable the development of algorithms. The test input MRI/CBCT and ground truth CT will not be shared with the participants to avoid optimistic biases. We envision that this challenge will enable a fair and open assessment of different approaches. References Chernak E.S., Rodriguez-Antunez A., Jelden G.L., Dhaliwal R.S., Lavik P.S. The use of computed tomography for radiation therapy treatment planning. Radiology. 1975 Dec;117(3):613-4. https://doi.org/10.1148/117.3.613 Ramella, S., Fiore, M., Silipigni, S., Zappa, M. C., Jaus, M., Alberti, A. M., ... & D’Angelillo, R. M. (2017). Local control and toxicity of adaptive radiotherapy using weekly CT imaging: results from the LARTIA trial in stage III NSCLC. Journal of Thoracic Oncology, 12(7), 1122-1130. https://doi.org/10.1016/j.jtho.2017.03.025 Kida, S., Nakamoto, T., Nakano, M., Nawa, K., Haga, A., Kotoku, J. I., ... & Nakagawa, K. (2018). Cone beam computed tomography image quality improvement using a deep convolutional neural network. Cureus, 10(4). https://doi.org/10.7759%2Fcureus.2548 Schmidt M. A., Payne G. S. Radiotherapy planning using MRI. Phys Med Biol. 2015;60:R323 Lagendijk, J. J., Raaymakers, B. W., Van den Berg, C. A., Moerland, M. A., Philippens, M. E., & Van Vulpen, M. (2014). MR guidance in radiotherapy. Physics in Medicine & Biology, 59(21), R349. https://doi.org/10.1088/0031-9155/59/21/r349 Edmund, J. M., & Nyholm, T. (2017). A review of substitute CT generation for MRI-only radiation therapy. Radiation Oncology, 12(1), 1-15. https://doi.org/10.1186/s13014-016-0747-y Spadea, M. F. & Maspero, M., Zaffino, P., & Seco, J. (2021). Deep learning-based synthetic-CT generation in radiotherapy and PET: A review. Medical Physics, 48(11), 6537–6566. https://doi.org/10.1002/mp.15150 Schwarz, C. G., Kremers, W. K., Wiste, H. J., Gunter, J. L., Vemuri, P., Spychalla, A. J., ... & Alzheimer's Disease Neuroimaging Initiative. (2021). Changing the face of neuroimaging research: Comparing a new MRI de-facing technique with popular alternatives. NeuroImage, 231, 117845. https://doi.org/10.1016/j.neuroimage.2021.117845 Low, D.A., Harms, W.B., Mutic, S., and Purdy, J.A. (1998), A technique for the quantitative evaluation of dose distributions. Med. Phys., 25: 656-661. https://doi.org/10.1118/1.598248 Hall, W. A., Paulson, E., Davis, B. J., Spratt, D. E., Morgan, T. M., Dearnaley, D., ... & Lawton, C. A. (2021). NRG oncology updated international consensus atlas on pelvic lymph node volumes for intact and postoperative prostate cancer. International Journal of Radiation Oncology* Biology* Physics, 109(1), 174-185.https://doi.org/10.1016/j.ijrobp.2020.08.034 Lambrecht, M., Eekers, D. B., Alapetite, C., Burnet, N. G., Calugaru, V., Coremans, I. E., ... & Troost, E. G. (2018). Radiation dose constraints for organs at risk in neuro-oncology; the European Particle Therapy Network consensus. Radiotherapy and Oncology, 128(1), 26-36 https://doi.org/10.1016/j.radonc.2018.05.001 Ezzell, G. A., Burmeister, J. W., Dogan, N., LoSasso, T. J., Mechalakos, J. G., Mihailidis, D., ... & Xiao, Y. (2009). IMRT commissioning: multiple institution planning and dosimetry comparisons, a report from AAPM Task Group 119. Medical physics, 36(11), 5359-5373. https://doi.org/10.1118/1.3238104 Drzymala, R. E., Mohan, R., Brewster, L., Chu, J., Goitein, M., Harms, W., & Urie, M. (1991). Dose-volume histograms. International Journal of Radiation Oncology* Biology* Physics, 21(1), 71-78. https://doi.org/10.1016/0360-3016(91)90168-4 International Commission on Radiation Units and Measurements, ICRU Report 83, Prescribing, recording, andreporting intensity-modulated photon-beam therapy (IMRT)(ICRU Report 83), Bethesda, MD (2010) https://www.fnkv.cz/soubory/216/icru-83.pdf

Design of the SynthRAD2023 Grand Challengehttps://synthrad2023.grand-challenge.org/ following the structured challenge submission provided at https://www.biomedical-challenges.org/. The challenge will be part of MICCAI 2023. This design has been first submited on 4-11-2022, revised on 02-2023 and textually updated on 03-2023 according to the approved submitted changes by MICCAI. &nbsp

A medical imaging system is often required to monitor bone growth or bone healing in patients. X-ray imaging method is one of the most used imaging systems but it affects human health due to the ionizing effect of X-rays on matter. In this paper, authors investigated the healing period of a bone structure by using electromagnetic antennas. Two different design approaches were examined which are the monopole antenna approach and the horn antenna approach with the frequency bands of 4-10 and 3-18 GHz, respectively. According to simulation results of the first scenario, radius change of a circular bone from 1 to 8 mm causes a transmission reduction from -13.5 dB to -20 dB at 7 GHz. In the second scenario, the thickness of the bone changed from 0.05 to 20 mm and the corresponding transmission value is reduced from -5 to -25 dB at 10 GHz of constant frequency. All these two approaches showed that the bone variation with time changes the transmission characteristics of detector antennas at the operating frequencies. In the last step of this study, an experimental measurement is carried out by using two opposite horn antennas and a planar bone structure in microwave laboratory. The measured results are in a good agreement with the simulated results in the second scenario. This study gives an opportunity to determine bone healing periods in medical applications by microwave techniques.

The detection of Brain cancer is an essential process, which is based on the clinician’s knowledge and experience. An automatic tumor classification model is important to handle radiologists to detect the brain tumors. However, the precision of present model should be enhanced for appropriate treatments. Numerous computer-aided diagnosis (CAD) models are offered in the literary works of medical imaging to help radiologists concerning their patients. This paper proposes an optimization-driven technique, namely Whale Harris Hawks optimization (WHHO) for brain tumor detection using MR images. Here, segmentation is performed using cellular automata and rough set theory. In addition, the features are extracted from the segments, which include tumor size, Local Optical Oriented Pattern (LOOP), Mean, Variance, and Kurtosis. In addition, the brain tumor detection is carried out using deep convolutional neural network (DeepCNN), wherein the training is performed using proposed WHHO. The proposed WHHO is designed by integrating Whale optimization algorithm (WOA) and Harris hawks optimization (HHO) algorithm. The proposed WHHO-based DeepCNN outperformed other methods with maximal accuracy of 0.816, maximal specificity of 0.791, and maximal sensitivity of 0.974, respectively.

With the advancements in modern medical imaging systems, the diagnostic image data has increased exponentially. The future medical applications seek medical imaging devices to be portable. Image quality and real-time processing are of prime importance in medical image compression. Therefore the volumetric medical images need to be compressed in perceptually lossless manner. Also the time taken to compress the images before transmitting (or storing), should be small. In this paper, an algorithm for lossless and perceptually-lossless medical image compression is proposed. The proposed algorithm uses two small lists and two small state-tables to encode an image. The compression efficiency is comparable to the state-of-the-art lossless compression techniques. Also the computational complexity and memory requirement are realistic for portable medical imaging devices. Combining all the three features, it is obvious that the proposed algorithm is a better candidate for image compression in comparison to all the state-of-art compression algorithms that we know of, for volumetric-medical imaging systems.

The interest on graph matching has not stopped growing since the late seventies. The basic idea of graph matching consists of generating graph representations of different data or structures and compare those representations by searching correspondences between them. There are manifold techniques that have been developed to find those correspondences and the choice of one or another depends on the characteristics of the application of interest. These applications range from pattern recognition (e.g. biometric identification) to signal processing or artificial intelligence. One of the aspects that make graph matching so attractive is its ability to facilitate data analysis, and medical imaging is one of the fields that can benefit from this in a greater extent. The potential of graph matching to find similarities and differences between data acquired at different points in time shows its potential to improve diagnosis, follow-up of human diseases or any other of the clinical scenarios that require comparison between different datasets. In spite of the large amount of papers that were published in this field to the date there is no survey paper of graph matching for clinical applications. This survey aims to fill this gap.

Background: Cancer is one of the life threatening disease which is affecting a large number of population worldwide. The cancer cells multiply inside the body without showing much symptoms on the surface of the skin thereby making it difficult to predict and detect at the onset of disease. Many organizations are working towards automating the process of cancer detection with minimal false detection rates. Introduction: The machine learning algorithms serve to be a promising alternative to support health care practitioners to rule out the disease and predict the growth with various imaging and statistical analysis tools. The medical practitioners are utilizing the output of these algorithms to diagnose and design the course of treatment. These algorithms are capable of finding out the risk level of the patient and can reduce the mortality rate concerning to cancer disease. Method: This article presents the existing state of art techniques for identifying cancer affecting human organs based on machine learning models. The supported set of imaging operations are also elaborated for each type of Cancer. Conclusion: The CAD tools are the aid for the diagnostic radiologists for preliminary investigations and detecting the nature of tumor cells.

Background and Overview Trigeminal neuralgia due to a metastatic lesion is a relatively rare occurrence. These cases pose a diagnostic challenge for the clinician due to a complex clinical presentation. Case Description The authors describe the case of a 65-year-old woman with left-sided facial pain and occasional numbness. The patient also reported autonomic features associated with facial pain, facial muscular weakness, and hearing and visual impairment. The patient's history of breast cancer, for which she was receiving treatment for bone metastasis, warranted diagnostic imaging. Magnetic resonance imaging with and without contrast was performed and revealed a possible metastatic lesion. Conclusions and Practical Implications Unusual clinical features must be identified, and prompt appropriate imaging is instrumental in making at an accurate diagnosis and management plan.

Background: The new coronavirus disease 2019 (COVID-19) is spreading rapidly around the world. Artificial Intelligence (AI) assisted identification and detection of diseases is an effective method of medical diagnosis. Objectives: To present recent advances in AI-assisted diagnosis of COVID-19, we introduce major aspects of AI in the process of diagnosing COVID-19. Methods: In this paper, we firstly cover the latest collection and processing methods of datasets of COVID-19. The processing methods mainly include building public datasets, transfer learning, unsupervised learning and weakly supervised learning, semi-supervised learning methods and so on. Secondly, we introduce the algorithm application and evaluation metrics of AI in medical imaging segmentation and automatic screening. Then, we introduce the quantification and severity assessment of infection in COVID-19 patients based on image segmentation and automatic screening. Finally, we analyze and point out the current AI-assisted diagnosis of COVID-19 problems, which may provide useful clues for future work. Conclusion: AI is critical for COVID-19 diagnosis. Combining chest imaging with AI can not only save time and effort, but also provide more accurate and efficient medical diagnosis results.

The most common burn wound assessment continues to be the clinical inspection and the tactile examination, which are subjective and remain challenging even for experienced burn surgeons. Recently, hyperspectral imaging camera systems have been increasingly used to support the evaluation of burn wounds. The aim of our study was to determine if hyperspectral imaging analysis differentiates and objectifies the assessment of burn wounds in burns of the upper extremities. We included 97 superficial partial, deep partial dermal burns, and full thickness burns. Hyperspectral imaging analysis was performed for all burns using proprietary software. The software recorded parameters for tissue oxygenation (StO2), tissue hemoglobin index, and near-infrared perfusion. These values were compared with the recordings for healthy, non-burned skin. We found that hyperspectral imaging analysis effectively differentiates burn wounds and shows the ability to distinguish even superficial partial burns from deep partial burns in the near-infrared perfusion analysis feature. Although, it was not possible to differentiate burn wounds in all features. Currently, it is important to optimize the respective reference values of the individual burn degrees for an objectified assessment.

Purpose: The present study compared the performance of computed tomography (CT), magnetic resonance imaging-derived proton density fat fraction (MRI-PDFF), controlled attenuation parameter (CAP), grayscale ultrasonography (US), and attenuation imaging (ATI) for the diagnosis of hepatic steatosis (HS).Methods: In total, 120 prospectively recruited patients who underwent hepatic resection between June 2018 and June 2020 were retrospectively analyzed. CT, MRI-PDFF, CAP, grayscale US, and ATI were performed within 3 months before surgery. Diagnostic performance for HS ≥5% and HS >33% was compared using the area under the curve (AUC) of receiver operating characteristic curves. Histopathologic examinations served as the reference standard for the degree of HS.Results: For detecting HS ≥5%, MRI-PDFF (AUC, 0.946) significantly outperformed CT, CAP and grayscale US (AUC, 0.807, 0.829, and 0.761, respectively) (P33%, all the modalities provided good diagnostic performance without significant differences (AUC, 0.887-0.947; P>0.05 for all).Conclusion: For detecting HS ≥5%, MRI-PDFF was the best imaging modality, while ATI outperformed grayscale US. For detecting HS >33%, all five imaging tools demonstrated good diagnostic performance.

With the emergence of immune-modulating therapies, brain tumors present significant diagnostic imaging challenges. These challenges include planning personalized treatment and adjudicating accurate monitoring approaches and therapeutically specific response criteria. This has been due, in part, to the reliance on nonspecific imaging metrics, such as gadolinium-contrast-enhanced MRI or FDG PET, and rapidly evolving biologic understanding of neuroinflammation. The importance of the tumor-immune interaction and ability to therapeutically augment inflammation to improve clinical outcomes necessitates that the radiologist develop a working knowledge of the immune system and its role in clinical neuroimaging. In this article, we review relevant biologic concepts of the tumor microenvironment of primary and metastatic brain tumors, these tumors' interactions with the immune system, and MRI and PET methods for imaging inflammatory elements associated with these malignancies. Recognizing the growing fields of immunotherapeutics and precision oncology, we highlight clinically translatable imaging metrics for the diagnosis and monitoring of brain tumor neuroinflammation. Practical guidance is provided for implementing iron nanoparticle imaging, including imaging indications, protocol, interpretation, and pitfalls. A comprehensive understanding of the inflammatory mechanisms within brain tumors and their imaging features will facilitate the development of innovative non-invasive prognostic and predictive imaging strategies for precision oncology.

From the first report of the respiratory disease caused by SARS-Cov-2 towards the end of 2019, in Wuhan, China, to June 8, 2020, 6,931,000 people had been affected in the world, and 113,619 in Mexico, where there is a lack of literature describing the tomographic behavior of this disease throughout its evolution and its correlation with patient clinical data.To identify, describe and classify the manifestations detected on chest computed tomography images of patients with pneumonia caused by SARS-CoV-2.Hospitalized patients with RT-PCR-confirmed SARS-CoV-2 infection and who underwent a plain chest tomography were included.The most common tomographic patterns were ground glass opacity, consolidation, cobblestone pattern and mixed pattern, with peripheral and subpleural distribution and bilateral involvement in up to 95 %. Subpleural bands that distorted the architecture were identified in up to 62 % of patients, with an average of seven days of evolution. The most common extrapulmonary finding was an increase in pulmonary artery transverse axis.Knowing the behavior of this type of pneumonia allows becoming familiarized with the characteristic tomographic patterns, in order to aid timely detection and treatment.Desde el primer reporte de la enfermedad respiratoria causada por SARS-Cov-2 a finales de 2019, en Wuhan, China, hasta el 8 de junio de 2020, habían sido afectadas 6 931 000 personas en el mundo y 113 619 en México, donde se carece de informes que describan el comportamiento tomográfico de esta enfermedad a lo largo de su evolución y su correlación con los datos clínicos del paciente.Identificar, describir y clasificar las manifestaciones en las imágenes de tomografía de tórax de pacientes con neumonía causada por SARS-CoV-2.Se incluyeron pacientes hospitalizados con infección por SARS-CoV-2 confirmada mediante PCR-TR y en quienes se realizó tomografía simple de tórax.Los patrones tomográficos más comunes fueron opacidad de vidrio despulido, consolidación, empedrado y patrón mixto, con distribución periférica, subpleural y afectación bilateral hasta en 95 %. Se identificaron bandas subpleurales que distorsionaban la arquitectura hasta en 62 % de los pacientes, con un promedio de siete días de evolución. El aumento del eje transverso de la arteria pulmonar constituyó el hallazgo extrapulmonar más frecuente.Conocer el comportamiento de este tipo de neumonía permite la familiarización con los patrones tomográficos característicos, con el fin de coadyuvar a la detección y tratamiento oportunos.

Background Omission errors in medical imaging can lead to missed diagnosis and harm to patients. The subject has been studied in conventional imaging, but no data is available for functional imaging in general and for PET/CT in particular. In this work, we evaluated the frequency and characteristics of perceptual omission errors in the PET component of oncologic PET/CT imaging and we analyze the hazardous scenarios prone to such modality-specific errors. Materials and methods Perceptual omission errors were collected in one tertiary center PET/CT clinic during routine PET/CT reporting over a 26-month period. The omissions were detected either in reporting follow-up PET/CT studies of the same patient or during multidisciplinary meetings. Results Significant omission errors were found in 1.2 % of the 2100 reports included in the study. The most common omissions were bone metastases and focal colon uptake. We identified six PET-specific causative factors contributing to the occurrence of omissions, and we propose solutions to minimize their influence. Conclusions The data presented here can help to promote the awareness of interpreting physicians to body areas that require higher attention and to implement reading strategies for improving the accuracy of PET/CT interpretation.

BackgroundIdiopathic pulmonary fibrosis (IPF) is a progressive, irreversible, and usually fatal lung disease of unknown reasons, generally affecting the elderly population. Early diagnosis of IPF is crucial for triaging patients' treatment planning into anti-fibrotic treatment or treatments for other causes of pulmonary fibrosis. However, current IPF diagnosis workflow is complicated and time-consuming, which involves collaborative efforts from radiologists, pathologists, and clinicians and it is largely subject to inter-observer variability.PurposeThe purpose of this work is to develop a deep learning-based automated system that can diagnose subjects with IPF among subjects with interstitial lung disease (ILD) using an axial chest computed tomography (CT) scan. This work can potentially enable timely diagnosis decisions and reduce inter-observer variability.MethodsOur dataset contains CT scans from 349 IPF patients and 529 non-IPF ILD patients. We used 80% of the dataset for training and validation purposes and 20% as the holdout test set. We proposed a two-stage model: at stage one, we built a multi-scale, domain knowledge-guided attention model (MSGA) that encouraged the model to focus on specific areas of interest to enhance model explainability, including both high- and medium-resolution attentions; at stage two, we collected the output from MSGA and constructed a random forest (RF) classifier for patient-level diagnosis, to further boost model accuracy. RF classifier is utilized as a final decision stage since it is interpretable, computationally fast, and can handle correlated variables. Model utility was examined by (1) accuracy, represented by the area under the receiver operating characteristic curve (AUC) with standard deviation (SD), and (2) explainability, illustrated by the visual examination of the estimated attention maps which showed the important areas for model diagnostics.ResultsDuring the training and validation stage, we observe that when we provide no guidance from domain knowledge, the IPF diagnosis model reaches acceptable performance (AUC±SD=0.93±0.07), but lacks explainability; when including only guided high- or medium-resolution attention, the learned attention maps are not satisfactory; when including both high- and medium-resolution attention, under certain hyperparameter settings, the model reaches the highest AUC among all experiments (AUC±SD=0.99±0.01) and the estimated attention maps concentrate on the regions of interests for this task. Three best-performing hyperparameter selections according to MSGA were applied to the holdout test set and reached comparable model performance to that of the validation set.ConclusionsOur results suggest that, for a task with only scan-level labels available, MSGA+RF can utilize the population-level domain knowledge to guide the training of the network, which increases both model accuracy and explainability.

PurposeHindsight bias-where people falsely believe they can accurately predict something once they know about it-is a pervasive decision-making phenomenon, including in the interpretation of radiological images. Evidence suggests it is not only a decision-making phenomenon but also a visual perception one, where prior information about an image enhances our visual perception of the contents of that image. The current experiment investigates to what extent expert radiologists perceive mammograms with visual abnormalities differently when they know what the abnormality is (a visual hindsight bias), above and beyond being biased at a decision level.ApproachN=40 experienced mammography readers were presented with a series of unilateral abnormal mammograms. After each case, they were asked to rate their confidence on a 6-point scale that ranged from confident mass to confident calcification. We used the random image structure evolution method, where the images repeated in an unpredictable order and with varied noise, to ensure any biases were visual, not cognitive.ResultsRadiologists who first saw an original image with no noise were more accurate in the max noise level condition [area under the curve (AUC)=0.60] than those who first saw the degraded images (AUC=0.55; difference: p=0.005), suggesting that radiologists' visual perception of medical images is enhanced by prior visual experience with the abnormality.ConclusionsOverall, these results provide evidence that expert radiologists experience not only decision level but also visual hindsight bias, and have potential implications for negligence lawsuits.

Early diagnosis of dental caries progression can prevent invasive treatment and enable preventive treatment. In this regard, dental radiography is a widely used tool to capture dental visuals that are used for the detection and diagnosis of caries. Different deep learning (DL) techniques have been used to automatically analyse dental images for caries detection. However, most of these techniques require large-scale annotated data to train DL models. On the other hand, in clinical settings, such medical images are scarcely available and annotations are costly and time-consuming. To this end, we present an efficient self-training-based method for caries detection and segmentation that leverages a small set of labelled images for training the teacher model and a large collection of unlabelled images for training the student model. We also propose to use centroid cropped images of the caries region and different augmentation techniques for the training of self-supervised models that provide computational and performance gains as compared to fully supervised learning and standard self-supervised learning methods. We present a fully labelled dental radiographic dataset of 141 images that are used for the evaluation of baseline and proposed models. Our proposed self-supervised learning strategy has provided performance improvement of approximately 6% and 3% in terms of average pixel accuracy and mean intersection over union, respectively as compared to standard self-supervised learning. Data and code will be made available to facilitate future research. 2023, The Author(s). This work is supported in parts by EPSRC grant no: EP/T021063/1 and Ajman University Internal Research Grants No. [2021-IRGDEN-7 and RTG-2022-DEN-01]. The research findings presented in this article are solely the author(s) responsibility. Scopus

Many eye diseases like Diabetic Macular Edema (DME), Age-related Macular Degeneration (AMD), and Glaucoma manifest in the retina, can cause irreversible blindness or severely impair the central version. The Optical Coherence Tomography (OCT), a 3D scan of the retina with high qualitative information about the retinal morphology, can be used to diagnose and monitor changes in the retinal anatomy. Many Deep Learning (DL) methods have shared the success of developing an automated tool to monitor pathological changes in the retina. However, the success of these methods depend mainly on large datasets. To address the challenge from very small and limited datasets, we proposed a DL architecture termed CoNet (Coherent Network) for joint segmentation of layers and fluids in retinal OCT images on very small datasets (less than a hundred training samples). The proposed model was evaluated on the publicly available Duke DME dataset consisting of 110 B-Scans from 10 patients suffering from DME. Experimental results show that the proposed model outperformed both the human experts' annotation and the current state-of-the-art architectures by a clear margin with a mean Dice Score of 88% when trained on 55 images without any data augmentation., Comment: Submitted to Bioimaging 2023

Objectives: The purpose of this systematic review was to answer the focus question: “Could the gray values (GVs) from CBCT (cone beam computed tomography) be converted to Hounsfield units (HUs) in multidetector computed tomography (MDCT)?” Methods: The included studies try to answer the research question according to the PICO strategy. Studies were gathered by searching several electronic databases and partial grey literature up to January 2021 without language or time restrictions. The methodological assessment of the studies was performed using The Oral Health Assessment Tool (OHAT) for in vitro studies and the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) for in vivo studies. The Grading of Recommendations Assessment, Development and Evaluation (GRADE system) instrument was applied to assess the level of evidence across the studies. Results: 2710 articles were obtained in Phase 1, and 623 citations remained after removing duplicates. Only three studies were included in this review using a two-phase selection process and after applying the eligibility criteria. All studies were methodologically acceptable, although in general terms with low risks of bias. There are some included studies with quite low and limited evidence estimations and recommendation forces; evidencing the need for clinical studies with diagnostic capacity to support its use. Conclusions: This systematic review demonstrated that the GVs from CBCT cannot be converted to HUs due to the lack of clinical studies with diagnostic capacity to support its use. However, it is evidenced that three conversion steps (equipment calibration, prediction equation models, and a standard formula (converting GVs to HUs)) are needed to obtain pseudo Hounsfield values instead of only obtaining them from a regression or directly from the software.

By promising more accurate diagnostics and individual treatment recommendations, deep neural networks and in particular convolutional neural networks have advanced to a powerful tool in medical imaging. Here, we first give an introduction into methodological key concepts and resulting methodological promises including representation and transfer learning, as well as modelling domain-specific priors. After reviewing recent applications within neuroimaging-based psychiatric research, such as the diagnosis of psychiatric diseases, delineation of disease subtypes, normative modeling, and the development of neuroimaging biomarkers, we discuss current challenges. This includes for example the difficulty of training models on small, heterogeneous and biased data sets, the lack of validity of clinical labels, algorithmic bias, and the influence of confounding variables.

Due to their high resolution, dynamic medical 2D+t and 3D+t volumes from computed tomography (CT) and magnetic resonance tomography (MR) reach a size which makes them very unhandy for teleradiologic applications. A lossless scalable representation offers the advantage of a down-scaled version which can be used for orientation or previewing, while the remaining information for reconstructing the full resolution is transmitted on demand. The wavelet transform offers the desired scalability. A very high quality of the lowpass sub-band is crucial in order to use it as a down-scaled representation. We propose an approach based on compensated wavelet lifting for obtaining a scalable representation of dynamic CT and MR volumes with very high quality. The mesh compensation is feasible to model the displacement in dynamic volumes which is mainly given by expansion and contraction of tissue over time. To achieve this, we propose an optimized estimation of the mesh compensation parameters to optimally fit for dynamic volumes. Within the lifting structure, the inversion of the motion compensation is crucial in the update step. We propose to take this inversion directly into account during the estimation step and can improve the quality of the lowpass sub-band by 0.63 and 0.43 dB on average for our tested dynamic CT and MR volumes at the cost of an increase of the rate by 2.4% and 1.2% on average.

Tomographic coherent imaging requires the reconstruction of a series of two-dimensional projections of the object. We show that using the solution for the image of one projection as the starting point for the reconstruction of the next projection offers a reliable and rapid approach to the image reconstruction. The method is demonstrated on simulated and experimental data. This technique also simplifies reconstructions using data with curved incident wavefronts.

BACKGROUND Medical errors are defined by Grober and Bohnen (2005, p. 42) as, “an act of omission or commission in planning or execution that contributes or could contribute to an unintended result”. Subsequently, jeopardization of patient safety and economic problems can occur, regardless of if the incident was adverse (Grober & Bohnen, 2005). Adverse events are caused by unforeseen medical incidents and may result in death or permanent disability (Wilson et al., 1995). The only in-depth retrospective quantitative study assessing the quality of the Australian healthcare system in 1992, revealed the rate of adverse events was 16.6% where 51% of these events were preventable out of 14,179 hospital admissions in New South Wales and South Australia (Wilson et al., 1995). Notably, 13.7% of patients resulted in permanent disability and 4.9% suffered death (Wilson et al., 1995). Radiologists face litigation due to radiological, external and psychological errors and each new modality introduces new reasons for litigation (Littlefair, 2015). Yasgur (2022) identified out of 29 medical specialties, radiology ranked ninth in malpractice lawsuit involvement. Many radiology claims were due to delayed or misdiagnosis, with the remaining claims involving complications from treatment, surgery, poor outcome or disease progression (Yasgur, 2022). Numerous studies revealed lawsuits against radiologists are primarily due to misdiagnosis or miscommunication, resulting in poor patient care (Biswas et al., 2023; Cannavale et al., 2013; Kwee & Kwee, 2020). Tarkiainen et al. (2021) conducted a retrospective study to survey patient injury claims relating to medical imaging (MI) in Finland from 1991-2017. Out of 1054 claims, most claims were related to mammography and screening mammography (29.8%), radiography (16.1%) and magnetic resonance imaging (MRI) (15.4%), while claims involving ultrasound and computed tomography (CT) were 9.8% and 5.9% respectively. Despite diagnosis-related errors (55.1%) being the most common reason for filing lawsuits, medical malpractice errors (36.3%) and improper examinations (6%) were also emphasized. In the United Kingdom, the total value of 10,284 clinical negligence claims in 2021/22 was £6.03 billion, where 4% of claims were related to radiology, making up only 1% (£60.3 million) of the total value (National Health Service Resolution, 2022). Harvey et al. (2016) conducted a quantitative study on the frequency and cost of radiology malpractice claims in the United States from 2008 to 2012 identifying out of 300,000 medical claims, 879 claims were related to radiology. 37.7% of radiology claims involved indemnity payments to the claimant with the median indemnity payment equating to $175,000. Staggeringly, the total paid loss (indemnity payment, defence cost and administrative costs) for all claims was $180.7 million with a median loss of $30,091 per case. Medical imaging examinations are performed by sonographers and medical imaging technologists (MITs), which include diagnostic radiographers and nuclear medicine technologists (NMTs). Diagnostic radiographers utilise ionising radiation to perform CT and projection radiography in a variety of clinical settings to generate diagnostic images (Medical Radiation Practice Board of Australia (MRPBA), 2019). In contrast, NMTs are required to prepare the radiopharmaceutical and conduct nuclear medicine examinations and therapies within their respective imaging modalities (MRPBA, 2019). MITs must be competent in performing MRI and/or ultrasound if required as part of their practice (MRPBA, 2019), while sonographers need to perform diagnostic ultrasound examinations with sound waves to generate images of various body systems (Australian Sonographer Accreditation Registry, 2023). In the radiology department, patient incident reports may occur due to miscommunication between radiology staff and patients, lack of radiographic and/or pathological knowledge and improper imaging examinations (Wallin et al., 2018). Errors may occur during the preprocedural, procedural and postprocedural phases of the general radiography imaging cycle for radiographers (Zhou et al., 2015). During the preprocedural phase, wrong examination, wrong anatomical side, patient misidentification and request form errors may occur, while the procedural phase is associated with radiation incidents and repeats of plain radiographs (Zhou et al., 2015). Kjelle et al. (2021) revealed radiographers rejected or repeated radiographs due to suboptimal positioning, collimation, centering, artefacts, and exposure errors. These are examples of radiographic errors that could be performed by radiographers, creating possible avenues for patient complaints and misdiagnosis. Multidetector CT is complex as it requires appropriate CT acquisition parameters (collimation, reconstruction interval, etc.) and determining the relevant anatomical volume to be covered concerning the clinical question (Mazzei & Volterrani, 2015). Therefore, modifying CT examinations to tailor both the patient’s characteristics and the specific pathology being investigated while minimising dose to obtain diagnostic images can be challenging (Mazzei & Volterrani, 2015). One Saudi Arabian study concluded MRI technologists lack the education and knowledge in identifying MRI quality errors and suggesting appropriate correction techniques (Alsharif et al., 2016), making the identification of potential pathologies difficult due to poor image quality. Mammographers encounter challenges in optimising technical and radiographic techniques to produce diagnostic images, while minimising pain caused by breast compression during mammograms (Breast Screen South Australia, 2023; Metsälä et al., 2017). This may exacerbate errors in mammogram interpretations, which is already difficult as breast tissue composition varies among different women, making it difficult to distinguish abnormal growths from normal tissue (National Institute of Biomedical Imaging and Bioengineering, 2023). Zylstra et al. (2005) identified between 1999 and 2004, the aggregate indemnity payment was over $47 million from 132 breast cancer cases involving misdiagnosis and inadequate follow-up, which are still common reasons for filing litigation claims in recent years (Tarkiainen et al., 2021). Regarding NMTs, errors involving medication/IV fluids mainly involved omitted medicine or dose, while incorrect clinical administrations and clinical procedures mainly involved wrong patients and incomplete examinations respectively (Kasalak et al., 2020). Benacerraf (2004) highlights that ultrasound examinations are operator dependent and require adequate experience to perform coherent examinations. Common sources of error relating to the sonographer’s expertise include inadequate knowledge of the ultrasound equipment, failure to optimise images, usage of inappropriate probes, lack of communication with the patient and over-estimation of one’s capabilities (Farina & Sparano, 2012; Pinto et al., 2013). Between 1985 and the early 2000s, there was an upward trend in the number of litigation cases relating to ultrasound, with a survey study revealing that many sonographers believed obstetrics was the most likely area to result in litigation (Harrison & Mitcalf, 2003). This is further supported by Sanders (2003), who identified common reasons for filing litigation claims in obstetric and gynaecologic ultrasound were due to misdiagnosis, misinterpretation of images, the invention of lesions, delay in communicating findings to physicians, cases involving sonographers (particularly alleged sexual assault with the use of the vaginal probe) and invasive procedures done incorrectly due to improper ultrasound guidance. With advances in technology and rigorous protocols for practice, it is not clear if there have been any recent changes in trends for sonographer litigation. Preliminary searches identified an abundance of literature concerning the types of errors and challenges involved for MITs. However, the direct link between errors and malpractice litigation for MITs is poorly understood. Identification of potential causes of litigation also has implications for policy and practice, where more rigorous protocols can be developed and implemented to improve safety for patient centred care. Hence, this research seeks to minimise the knowledge gap in the current literature on the causes of malpractice litigation. AIM The aim of this scoping review is to map the existing literature about the common trends in malpractice litigation of allied health professionals working in MI, including identifying the causes of malpractice litigation. To reduce the knowledge gap surrounding this topic, the research question, “What are the trends in malpractice litigation in medical imaging technologists and sonographers and why do they occur?” will allow a detailed scope of this study.

Fan-Slicer (https://github.com/UCL/fan-slicer) is a Python package that enables the fast sampling (slicing) of 2D ultrasound-shaped images from a 3D volume. To increase sampling speed, CUDA kernel functions are used in conjunction with the Pycuda package. The main features include functions to generate images from both 3D surface models and 3D volumes. Additionally, the package also allows for the sampling of images from curvilinear (fan shaped planes) and linear (rectangle shaped planes) ultrasound transducers. Potential uses of Fan-slicer include the generation of large datasets of 2D images from 3D volumes and the simulation of intra-operative data among others.

A small ultra-wideband (UWB) patch antenna for microwave breast imaging (MWI) applications is shown off in this paper. However, to improve the antenna performance relating to the bandwidth (BW), the radiating element of the suggested antenna is modified by adding slits in the patch as well as the ground plane. The proposed prototype has a relatively small size of 20 × 19 × 1.6 mm3 and it accomplishes a return loss below −10 dB (S11< −10 dB) at an overall BW of 7 GHz (4–11 GHz) with more than 3 dBi realized gain. The antenna is designed and simulated by using a finite integration technique-based simulator. In this way, the characteristics of the fabricated antenna are measured to examine the antenna performance. Indeed, the fidelity factor of face-to-face (FtF) and side-by-side (SbS) scenarios are also noticed for the same frequency range. In the final analysis, a simulation model of the antennas, that operate as a transceiver, and a breast phantom model with tumor sample is proposed for detecting cancerous tumor cells within the breast. Hence, the proposed approach is suitable for UWB based MWI applications in tumor cell detection.