Search

Study Design A retrospective cohort study of patients with surgically treated thoracolumbar fractures. Purpose This study aimed to describe the incidence of adverse events (AEs) after surgical stabilization of thoracolumbar spine injuries and to identify predictive factors for the occurrence of AEs. Overview of Literature Thoracolumbar spine fractures are frequently present in patients with blunt trauma and are associated with significant morbidity. AEs can occur due to the initial spinal injury or secondary to surgical treatment. There is a lack of emphasis in the literature on the AEs that can occur after operative management of thoracolumbar fractures. Methods We performed a retrospective review of 199 patients with surgically treated thoracolumbar fractures operated between January 2007 and January 2018. The potential risk factors for the development of AEs as well as the development of common complications were evaluated by univariate analysis, and a multivariate logistic regression analysis was performed to identify independent risk factors predictive of the above. Results The overall rate of AEs was 46.7%; 83 patients (41.7%) had nonsurgical AEs, whereas 24 (12.1%) had surgical adverse events. The most common AEs were urinary tract infections in 43 patients (21.6%), and hospital-acquired pneumonia in 21 patients (10.6%). On multivariate logistic regression, a Thoracolumbar Injury Classification and Severity (TLICS) score of 8–10 (odds ratio [OR], 6.39; 95% confidence interval [CI], 2.33–17.51), the presence of polytrauma (OR, 2.64; 95% CI, 1.17–5.99), and undergoing open surgery (OR, 2.31; 95% CI, 1.09–4.88) were significant risk factors for AEs. The absence of neurological deficit was associated with a lower rate of AEs (OR, 0.47; 95% CI, 0.31–0.70). Conclusions This study suggests the presence of polytrauma, preoperative American Spinal Injury Association score, and TLICS score are predictive of AEs in patients with surgically treated thoracolumbar fractures. The results might also suggest a role for minimally invasive surgical methods in reducing AEs in these patients.

We extend existing techniques by using generative adversarial network (GAN) models to reduce the appearance of cast shadows in radiographs across various age groups. We retrospectively collected 11,500 adult and paediatric wrist radiographs, evenly divided between those with and without casts. The test subset consisted of 750 radiographs with cast and 750 without cast. We extended the results from a previous study that employed CycleGAN by enhancing the model using a perceptual loss function and a self-attention layer. The CycleGAN model which incorporates a self-attention layer and perceptual loss function delivered a similar quantitative performance as the original model. This model was applied to images from 20 cases where the original reports recommended CT scanning or repeat radiographs without the cast, which were then evaluated by radiologists for qualitative assessment. The results demonstrated that the generated images could improve radiologists' diagnostic confidence, in some cases leading to more decisive reports. Where available, the reports from follow-up imaging were compared with those produced by radiologists reading AI-generated images. Every report, except two, provided identical diagnoses as those associated with follow-up imaging. The ability of radiologists to perform robust reporting with downsampled AI-enhanced images is clinically meaningful and warrants further investigation. Additionally, radiologists were unable to distinguish AI-enhanced from unenhanced images. These findings suggest the cast suppression technique could be integrated as a tool to augment clinical workflows, with the potential benefits of reducing patient doses, improving operational efficiencies, reducing delays in diagnoses, and reducing the number of patient visits., Competing Interests: Declarations. Ethics Approval: This study was exempt from Human Research Ethics Committee review as a retrospective quality improvement project. It was consistent with the NHMRC Ethical Considerations in Quality Assurance and Evaluation Activities (2014) guideline. Consent to Participate: Not applicable. Consent for Publication: Not applicable. Competing Interests: The authors declare no competing interests., (© 2024. The Author(s) under exclusive licence to Society for Imaging Informatics in Medicine.)

Without careful dissection of the ways in which biases can be encoded into artificial intelligence (AI) health technologies, there is a risk of perpetuating existing health inequalities at scale. One major source of bias is the data that underpins such technologies. The STANDING Together recommendations aim to encourage transparency regarding limitations of health datasets and proactive evaluation of their effect across population groups. Draft recommendation items were informed by a systematic review and stakeholder survey. The recommendations were developed using a Delphi approach, supplemented by a public consultation and international interview study. Overall, more than 350 representatives from 58 countries provided input into this initiative. 194 Delphi participants from 25 countries voted and provided comments on 32 candidate items across three electronic survey rounds and one in-person consensus meeting. The 29 STANDING Together consensus recommendations are presented here in two parts. Recommendations for Documentation of Health Datasets provide guidance for dataset curators to enable transparency around data composition and limitations. Recommendations for Use of Health Datasets aim to enable identification and mitigation of algorithmic biases that might exacerbate health inequalities. These recommendations are intended to prompt proactive inquiry rather than acting as a checklist. We hope to raise awareness that no dataset is free of limitations, so transparent communication of data limitations should be perceived as valuable, and absence of this information as a limitation. We hope that adoption of the STANDING Together recommendations by stakeholders across the AI health technology lifecycle will enable everyone in society to benefit from technologies which are safe and effective., Competing Interests: Declaration of interests JEA is a named researcher on grants funded by the Engineering and Physical Sciences Research Council (EPSRC), the National Institute for Health and Care Research (NIHR), and the Medical Research Council (MRC); is co-organiser of the Alan Turing Institute Clinical AI Interest Group; and was an NIHR Academic Clinical Fellow from 2017 to 2020. EL is employed as a research fellow at University Hospitals Birmingham NHS Foundation Trust and received support to attend the Symposium for Artificial Intelligence Learning Health Systems conference (May, 2023). JO is an employee of Roche Diagnostics but contributed while an employee of the UK MHRA; declares support from the AIRIS 2024 committee; and has shares in Roche Group. MG is a Canadian Institute for Advanced Research (CIFAR) AI Chair, CIFAR Azrieli Global Scholar, Herman L F von Helmholtz Career Development Professor, and Jameel Clinic Affiliate, and acknowledges support from these programmes. SRP is an employee of Google Research and has stock in Google. JS is an employee of Harrison.ai. NR is an employee of Google Research. HC-L is an employee of Google and has stock or stock options in Google. DS is Acting Deputy Editor of The Lancet Digital Health, Elsevier. BG was scientific advisor for Kheiron Medical Technologies (January, 2018, to September, 2021) and has had part-time employment with stock options as part of the standard compensation package (since October, 2021); has part-time employment at HeartFlow with stock options as part of the standard compensation package (since September, 2018); and is the Royal Academy of Engineering Research Chair in Safe Deployment of Medical Imaging AI. MCa is Director of the Birmingham Health Partners Centre for Regulatory Science and Innovation, Director of the Centre for Patient Reported Outcomes Research, and an NIHR Senior Investigator. MCa, NIHR Birmingham Biomedical Research Centre, Applied Research Collaboration West Midlands, UK SPINE, Research England, the European Regional Development Fund DEMAND Hub at the University of Birmingham, the University Hospitals Birmingham NHS Foundation Trust, and the NIHR Blood and Transplant Research Unit in Precision Transplant and Cellular Therapeutics; funding from Health Data Research UK, Innovate UK (part of UKRI), Macmillan Cancer Support, UCB Pharma, GSK, Anthony Nolan, Gilead Sciences, European Commission, European Federation of Pharmaceutical Industries and Associations, Janssen, Merck, and The Brain Tumor Charity; and personal fees from Aparito, ICON, CIS Oncology, Takeda Pharmaceuticals, Merck, Daiichi Sankyo, Glaukos, GSK, Halfloop, the Patient-Centered Outcomes Research Institute (PCORI), Pfizer, Genentech, and Vertex Pharmaceuticals, outside of the submitted work. MCa also declares royalties via revenue share from the commercial licence of Symptom Burden Questionnaire-Long COVID; payment or honoraria from the University of Maastricht, South-Eastern Norway Regional Health Authority, Cochrane Portugal, Singapore National Medical Research Council; and stock in GSK held by a family member. MCa has a leadership or fiduciary role in PROTEUS consortium (paid via a consultancy fee from Genetech and PCORI). TJP declares NIH awards (OT2OD032701 and R01EB030362), and grants or contracts from Google Cloud and Amazon Web Services. SK was an employee of Hardian Health. L-AF received consulting fees from Therapeutic Goods Administration, Australia, and WHO, and is a member of the AI in Healthcare Committee for Standards Australia and Director of Tethyan Consulting. CS received research funding from UKRI, NIHR, and the Wellcome Trust to support their research programme. BAM reports grants or contracts from MRC, the British Heart Foundation, the United States Agency for International Development, and HDRUK; consulting fees and support for attending meetings from the Bill & Melinda Gates Foundation; and was an employee of the Wellcome Trust at the time of the Delphi study. KH is an employee of Google Research. AK is an employee of Google and receives Google stock reimbursement. MM was an employee of Genomics England, Founder and Director of One HealthTech and Data Science for Health Equity (within One HealthTech), and received support from Nature Journals and Conde Nast. RP is an employee of MHRA. AKM is Deputy Editor of NEJM AI, NEJM Group. PM has a data custodian role as Director of the Clinical Practice Research Datalink and is an employee of MHRA. ZK was employed by Health Research Authority at the time of the consensus process. AAk received grants from the Economic and Social Care Research Council (ADR Wales ES/W012227/1) and HDRUK (HDR-9006). LB received support from the US National Institutes of Health (NIH) All of Us Research Programme to attend STANDING Together meetings. JWG is on the 2022 Robert Wood Johnson Foundation Harold Amos Medical Faculty Development Programme; declares support from the Radiological Society of North America Health Disparities grant (EIHD2204), Lacuna Fund (67), Gordon and Betty Moore Foundation, and NIH (NIBIB) Medical Imaging and Data Resource Center grant under contracts 75N92020C00008 and 75N92020C00021; has received honoraria from the National Bureau of Economic Research for authorship in their 2022 conference collection; has participated on the American Heart Association Deracing Algorithms advisory board; and is a board member of Hl7 and SIIM. LR is Senior Editor of the journal Nature Medicine, Nature Research, Springer Nature. MCh is an NIHR Academic Clinical Fellow. AAr is a panel member for various NIH Research committees, has received honoraria from HDRUK for attending panel meetings, and has an honorary affiliation with Moorfield's Eye Hospital. SP is employed by WHO. ES has had research funded by UK Research and Innovation (UKRI), NIHR, Wellcome Trust, Health Data Research UK (HDRUK), MRC, EPSRC, and Innovate UK, and declares support from the European Respiratory Society, British Thoracic Society, and Gilead. VT is Head of UK Approved Body as at the British Standards Institution. AD reports institutional research grants awarded by the NIHR, MRC, and EPSRC, and is an NIHR Senior Investigator, Director of the UK's incubator for Regulatory Science in AI and Digital Healthcare, deputy director of the Centre for Regulatory Science and Innovation, Birmingham Health Partners, system lead for AI diagnostics to NHS England, and a member of the UK Government's Regulatory Horizons Council. XL reports funding from the NIHR, the UK National Health Service AI Lab, The Health Foundation, Research England, the MHRA, the National Institute for Health and Care Excellence, and Hardian Health, and was previously a Health Studies Scientist at Apple; reports grants from the Medical Research Council, Wellcome Trust, NIHR, and Moorfields Eye Hospital Charity; received payment or honoraria for talks and book chapter reviews from the University of Turku and Elsevier; and received support from Harvard Medical School, SingHealth, and The Ada Lovelace Institute. All other authors declare no competing interests., (© 2024 World Health Organization. Published by Elsevier Ltd. This is an Open Access article published under the CC BY 3.0 IGO license which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. In any use of this article, there should be no suggestion that WHO endorses any specific organisation, products or services. The use of the WHO logo is not permitted. This notice should be preserved along with the article's original URL.)

Volvox carteri is an excellent system for investigating the origins of cell differentiation because it possesses just two cell types, reproductive gonidia and motile somatic cells, which evolved relatively recently. The somatic phenotype depends on the regA gene, which represses cell growth and reproduction, preventing cells expressing it from growing large enough to become gonidia. regA encodes a putative transcription factor and was generated in an undifferentiated ancestor of V. carteri through duplication of a progenitor gene whose ortholog in V. carteri is named rlsD. Here we analyze the function of rlsD through knockdown, overexpression, and RNA-seq experiments, to gain clues into the function of a member of an understudied putative transcription factor family and to obtain insight into the origins of cell differentiation in the volvocine algae. rlsD knockdown was lethal, while rlsD overexpression dramatically reduced gonidial growth. rlsD overexpression led to differential expression of approximately one-fourth of the genome, with repressed genes biased for those typically overexpressed in gonidia relative to somatic cells, and upregulated genes biased toward expression in soma, where regA expression is high. Notably, rlsD overexpression affects accumulation of transcripts for genes/Pfam domains involved in ribosome biogenesis, photosynthetic light harvesting, and sulfate generation, functions related to organismal growth, and responses to resource availability. We also found that in the wild type, rlsD expression is induced by light deprivation. These findings are consistent with the idea that cell differentiation in V. carteri evolved when a resource-responsive, growth-regulating gene was amplified, and a resulting gene duplicate was co-opted to repress growth in a constitutive, spatial context., (© 2024 Society for Experimental Biology and John Wiley & Sons Ltd.)

The chest X-ray (CXR) has a wide range of clinical indications in the field of cardiology, from the assessment of acute pathology to disease surveillance and screening. Despite many technological advancements, CXR interpretation error rates have remained constant for decades. The application of machine learning has the potential to substantially improve clinical workflow efficiency, pathology detection accuracy, error rates and clinical decision making in cardiology. To date, machine learning has been developed to improve image processing, facilitate pathology detection, optimize the clinical workflow, and facilitate risk stratification. This review explores the current and potential future applications of machine learning for chest radiography to facilitate clinical decision making in cardiology. It maps the current state of the science and considers additional potential use cases from the perspective of clinicians and technologists actively engaged in the development and deployment of deep learning driven clinical decision support systems.

Background: Serratia marcescens is an aerobic Gram-negative Enterobacterales bacillus that has emerged as a cause of hospital-associated infections., Aim: To report the epidemiological, diagnostic, and genetic investigation of an outbreak involving five neonatal patients infected or colonized with S. marcescens including the infection control interventions., Methods: The outbreak occurred in a 28-bedded neonatal unit in an acute care tertiary hospital in Singapore divided into three areas: two negative-pressure airborne infection isolation rooms with a shared anteroom, 10 neonatal intensive care unit (NICU) beds, and 16 high-dependency beds. In-flight patients and their immediate environment were screened for S. marcescens to determine probable environmental sources. Whole-genome sequencing (WGS) analysis of resulting isolates was performed to determine clone relatedness and possible transmission patterns. Implementation of infection control interventions included prompt isolation of cases, enhanced equipment and environmental disinfection, use of alcohol-based hand rub as the preferred hand hygiene mode, enhanced infection prevention orientation for parents, review of practices, audits, and immediate feedback on non-compliance., Findings: Five neonates infected or colonized with S. marcescens were involved in this outbreak. Four were infection cases and one was identified through contact tracing. Three NICU sinks and the milk preparation room sink were tested positive for S. marcescens. WGS confirmed clonality of strains from two NICU sinks, and milk preparation room sink with that of the five neonates., Conclusion: A multi-prong strategy was required to contain this outbreak. WGS analysis showed association of biofilms in sinks with the outbreak., Competing Interests: Conflict of interest statement None declared., (Copyright © 2024 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved.)

Objective: The aim of this study was to describe the time series of broad-spectrum antibiotic utilisation and incidence of antibiotic-resistant organisms during the implementation of antimicrobial stewardship programmes (ASP) in Singapore., Methods: An observational study was conducted using data from 2011 to 2020 in seven acute-care public hospitals. We applied joinpoint regressions to investigate changes in antibiotic utilisation rate and incidence density of antibiotic-resistant organisms., Results: Across the seven hospitals, quarterly broad-spectrum antibiotic utilisation rate remained stable. Half-yearly incidence density of antibiotic-resistant organisms with two joinpoints at first half (H1) of 2012 and second half (H2) of 2014 decreased significantly in the second and third period with a half-yearly percentage change (HPC) of -2.9% and - 0.5%, respectively. Across the five hospitals with complete data, half-yearly broad-spectrum antibiotic utilisation rate with one joinpoint decreased significantly from H1 of 2011 to H2 of 2018 (HPC - 4.0%) and H2 of 2018 to H2 2020 (HPC - 0.5%). Incidence density of antibiotic-resistant organisms decreased significantly in the two joinpoint periods from H1 of 2012 to H2 of 2014 (HPC - 2.7%) and H2 of 2014 to H2 of 2020 (HPC - 1.0%). Ceftriaxone with one joinpoint decreased significantly from H1 of 2011 to H1 of 2014 (HPC - 6.0%) and H1 of 2014 to H2 of 2020 (HPC - 1.8%) and ceftriaxone-resistant E. coli and K. pneumoniae decreased significantly in later periods, from H2 of 2016 to H2 of 2020 (HPC - 2.5%) and H1 of 2012 to H2 of 2015 (HPC - 4.6%) respectively. Anti-pseudomonal antibiotics with one joinpoint decreased significantly from H1 of 2011 to H2 of 2014 (HPC - 4.5%) and H2 of 2014 to H2 of 2020 (HPC - 0.8%) and that of quinolones with one joinpoint at H1 of 2015 decreased significantly in the first period. C. difficile with one joinpoint increased significantly from H1 of 2011 to H1 of 2015 (HPC 3.9%) and decreased significantly from H1 of 2015 to H2 of 2020 (HPC - 4.9%)., Conclusions: In the five hospitals with complete data, decrease in broad-spectrum antibiotic utilisation rate was followed by decrease in incidence density of antibiotic-resistant organisms. ASP should continue to be nationally funded as a key measure to combat antimicrobial resistance in acute care hospitals., (© 2023. BioMed Central Ltd., part of Springer Nature.)

Rapid detection of intracranial haemorrhage (ICH) is crucial for assessing patients with neurological symptoms. Prioritising these urgent scans for reporting presents a challenge for radiologists. Artificial intelligence (AI) offers a solution to enable radiologists to triage urgent scans and reduce reporting errors. This study aims to evaluate the accuracy of an ICH-detection AI software and whether it benefits a high-volume trauma centre in terms of triage and reducing diagnostic errors. A peer review of head CT scans performed prior to the implementation of the AI was conducted to identify the department's current miss-rate. Once implemented, the AI software was validated using CT scans performed over one month, and was reviewed by a neuroradiologist. The turn-around-time was calculated as the time taken from scan completion to report finalisation. 2916 head CT scans and reports were reviewed as part of the audit. The AI software flagged 20 cases that were negative-by-report. Two of these were true-misses that had no follow-up imaging. Both patients were followed up and exhibited no long-term neurological sequelae. For ICH-positive scans, there was an increase in TAT in the total sample (35.6%), and a statistically insignificant decrease in TAT in the emergency (- 5.1%) and outpatient (- 14.2%) cohorts. The AI software was tested on a sample of real-world data from a high-volume Australian centre. The diagnostic accuracy was comparable to that reported in literature. The study demonstrated the institution's low miss-rate and short reporting time, therefore any improvements from the use of AI would be marginal and challenging to measure., (© 2022. Crown.)

Introduction: Institutional surgical antibiotic prophylaxis (SAP) guidelines are in place at all public hospitals in Singapore, but variations exist and adherence to guidelines is not tracked consistently. A national point prevalence survey carried out in 2020 showed that about 60% of surgical prophylactic antibiotics were administered for more than 24 hours. This guideline aims to align best practices nationally and provides a framework for audit and surveillance., Method: This guideline was developed by the National Antimicrobial Stewardship Expert Panel's National Surgical Antibiotic Prophylaxis Guideline Development Workgroup Panel, which comprises infectious diseases physicians, pharmacists, surgeons and anaesthesiologists. The Workgroup adopted the ADAPTE methodology framework with modifications for the development of the guideline. The recommended duration of antibiotic prophylaxis was graded according to the strength of consolidated evidence based on the scoring system of the Singapore Ministry of Health Clinical Practice Guidelines., Results: This National SAP Guideline provides evidence-based recommendations for the rational use of antibiotic prophylaxis. These include recommended agents, dose, timing and duration for patients undergoing common surgeries based on surgical disciplines. The Workgroup also provides antibiotic recommendations for special patient population groups (such as patients with β-lactam allergy and patients colonised with methicillin-resistant Staphylococcus aureus ), as well as for monitoring and surveillance of SAP., Conclusion: This evidence-based National SAP Guideline for hospitals in Singapore aims to align practices and optimise the use of antibiotics for surgical prophylaxis for the prevention of surgical site infections while reducing adverse events from prolonged durations of SAP.

Objective: To evaluate the mechanisms associated with reflux events after sleeve gastrectomy (SG)., Summary Background Data: Gastro-esophageal reflux (GERD) post-SG is a critical issue due to symptom severity, impact on quality of life, requirement for reoperation, and potential for Barrett esophagus. The pathophysiology is incompletely delineated., Methods: Post-SG patients, stratified into asymptomatic and symptomatic, underwent protocolized nuclear scintigraphy (n = 83), 24-hour esophageal pH monitoring, and stationary manometry (n = 143) to characterize reflux patterns. Ten patients underwent fasting and postprandial concurrent manometry and pH for detailed analysis of reflux events., Results: Baseline demographics between cohorts were similar: Age 47.2 ± 11.6 versus 44.1 ± 11.3 years ( P = 0.121); females 73.2% versus 90.8% ( P = 0.005); excess weight loss 53.8 ± 28.1% versus 57.4 ± 25.5% ( P = 0.422), follow-up duration 12.3 versus 7.4 months ( P = 0.503). Nuclear scintigraphy delineated bolus-induced deglutitive reflux events (29.6% vs 62.5%, P = 0.005) and postprandial reflux events [4 (IQR2) versus 4 (IQR 3) events, P = 0.356]. Total acid exposure was significantly elevated in the symptomatic population (7.7% vs 3.6%, P < 0.001), especially fasting acid exposure (6.0% vs 1.3%, P < 0.001). pH/manometry analysis demonstrated acute elevations of the gastro-esophageal pressure gradient (>10 mm Hg) underpinned most reflux events. Swallow-induced intragastric hyper-pressur-ization was associated with individual reflux events in most patients (90% in fasting state and 40% postprandial)., Conclusions: We found reflux to be strongly associated with SG and identified 3 unique categories. Bolus-induced deglutitive and postprandial reflux occurred in most patients. Elevated fasting esophageal acid exposure mediated symptoms. Frequent, significant elevation in the gastro-esophageal pressure gradient was the mechanism of reflux and seemed to relate to the noncompliant proximal stomach., Competing Interests: The authors report no conflicts of interest., (Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.)

Objective: To describe the evolution of respiratory antibiotic prescribing during the coronavirus disease 2019 (COVID-19) pandemic across 3 large hospitals that maintained antimicrobial stewardship services throughout the pandemic., Design: Retrospective interrupted time-series analysis., Setting: A multicenter study was conducted including medical and intensive care units (ICUs) from 3 hospitals within a Canadian epicenter for COVID-19., Methods: Interrupted time-series analysis was used to analyze rates of respiratory antibiotic utilization measured in days of therapy per 1,000 patient days (DOT/1,000 PD) in medical units and ICUs. Each of the first 3 waves of the pandemic were compared to the baseline., Results: Within the medical units, use of respiratory antibiotics increased during the first wave of the pandemic (rate ratio [RR], 1.76; 95% CI, 1.38-2.25) but returned to the baseline in waves 2 and 3 despite more COVID-19 admissions. In ICU, the use of respiratory antibiotics increased in wave 1 (RR, 1.30; 95% CI, 1.16-1.46) and wave 2 of the pandemic (RR, 1.21; 95% CI, 1.11-1.33) and returned to the baseline in the third wave, which had the most COVID-19 admissions., Conclusions: After an initial surge in respiratory antibiotic prescribing, we observed the normalization of prescribing trends at 3 large hospitals throughout the COVID-19 pandemic. This trend may have been due to the timely generation of new research and guidelines developed with frontline clinicians, allowing for the active application of new research to clinical practice., (© The Author(s) 2022.)

Objectives: Trauma chest radiographs may contain subtle and time-critical pathology. Artificial intelligence (AI) may aid in accurate reporting, timely identification and worklist prioritisation. However, few AI programs have been externally validated. This study aimed to evaluate the performance of a commercially available deep convolutional neural network - Annalise CXR V1.2 (Annalise.ai) - for detection of traumatic injuries on supine chest radiographs., Methods: Chest radiographs with a CT performed within 24 h in the setting of trauma were retrospectively identified at a level one adult trauma centre between January 2009 and June 2019. Annalise.ai assessment of the chest radiograph was compared to the radiologist report of the chest radiograph. Contemporaneous CT report was taken as the ground truth. Agreement with CT was measured using Cohen's κ and sensitivity/specificity for both AI and radiologists were calculated., Results: There were 1404 cases identified with a median age of 52 (IQR 33-69) years, 949 males. AI demonstrated superior performance compared to radiologists in identifying pneumothorax ( p = 0.007) and segmental collapse ( p = 0.012) on chest radiograph. Radiologists performed better than AI for clavicle fracture ( p = 0.002), humerus fracture ( p < 0.0015) and scapula fracture ( p = 0.014). No statistical difference was found for identification of rib fractures and pneumomediastinum., Conclusion: The evaluated AI performed comparably to radiologists in interpreting chest radiographs. Further evaluation of this AI program has the potential to enable it to be safely incorporated in clinical processes., Advances in Knowledge: Clinically useful AI programs represent promising decision support tools.

Objectives: Around 30% of patients undergoing surgical resection for drug-resistant mesial temporal lobe epilepsy (MTLE) do not obtain seizure freedom. Success of anterior temporal lobe resection (ATLR) critically depends on the careful selection of surgical candidates, aiming at optimizing seizure freedom while minimizing postoperative morbidity. Structural MRI and FDG-PET neuroimaging are routinely used in presurgical assessment and guide the decision to proceed to surgery. In this study, we evaluate the potential of machine learning techniques applied to standard presurgical MRI and PET imaging features to provide enhanced prognostic value relative to current practice., Methods: Eighty two patients with drug resistant MTLE were scanned with FDG-PET pre-surgery and T1-weighted MRI pre- and postsurgery. From these images the following features of interest were derived: volume of temporal lobe (TL) hypometabolism, % of extratemporal hypometabolism, presence of contralateral TL hypometabolism, presence of hippocampal sclerosis, laterality of seizure onset volume of tissue resected and % of temporal lobe hypometabolism resected. These measures were used as predictor variables in logistic regression, support vector machines, random forests and artificial neural networks., Results: In the study cohort, 24 of 82 (28.3%) who underwent an ATLR for drug-resistant MTLE did not achieve Engel Class I (i.e., free of disabling seizures) outcome at a minimum of 2 years of postoperative follow-up. We found that machine learning approaches were able to predict up to 73% of the 24 ATLR surgical patients who did not achieve a Class I outcome, at the expense of incorrect prediction for up to 31% of patients who did achieve a Class I outcome. Overall accuracies ranged from 70% to 80%, with an area under the receiver operating characteristic curve (AUC) of .75-.81. We additionally found that information regarding overall extent of both total and significantly hypometabolic tissue resected was crucial to predictive performance, with AUC dropping to .59-.62 using presurgical information alone. Incorporating the laterality of seizure onset and the choice of machine learning algorithm did not significantly change predictive performance., Significance: Collectively, these results indicate that "acceptable" to "good" patient-specific prognostication for drug-resistant MTLE surgery is feasible with machine learning approaches utilizing commonly collected imaging modalities, but that information on the surgical resection region is critical for optimal prognostication., (© 2022 The Authors. Epilepsia published by Wiley Periodicals LLC on behalf of International League Against Epilepsy.)

Brain computed tomography (CTB) scans are widely used to evaluate intracranial pathology. The implementation and adoption of CTB has led to clinical improvements. However, interpretation errors occur and may have substantial morbidity and mortality implications for patients. Deep learning has shown promise for facilitating improved diagnostic accuracy and triage. This research charts the potential of deep learning applied to the analysis of CTB scans. It draws on the experience of practicing clinicians and technologists involved in development and implementation of deep learning-based clinical decision support systems. We consider the past, present and future of the CTB, along with limitations of existing systems as well as untapped beneficial use cases. Implementing deep learning CTB interpretation systems and effectively navigating development and implementation risks can deliver many benefits to clinicians and patients, ultimately improving efficiency and safety in healthcare., (Copyright © 2022 The Author(s). Published by Elsevier Ltd.. All rights reserved.)

Machine learning techniques, also known as artificial intelligence (AI), is about to dramatically change workflow and diagnostic capabilities in diagnostic radiology. The interest in AI in Interventional Radiology is rapidly gathering pace. With this early interest in AI in procedural medicine, IR could lead the way to AI research and clinical applications for all interventional medical fields. This review will address an overview of machine learning, radiomics and AI in the field of interventional radiology, enumerating the possible applications of such techniques, while also describing techniques to overcome the challenge of limited data when applying these techniques in interventional radiology. Lastly, this review will address common errors in research in this field and suggest pathways for those interested in learning and becoming involved about AI., (© 2022. The Author(s).)