Search

One of the early manifestations of systemic atherosclerosis, which leads to blood circulation issues, is the enhanced arterial light reflex (EALR). Fundus images are commonly used for regular screening purposes to intervene and assess the severity of systemic atherosclerosis in a timely manner. However, there is a lack of automated methods that can meet the demands of large-scale population screening. Therefore, this study introduces a novel cross-scale transformer-based multi-instance learning method, named MIL-CT, for the detection of early arterial lesions (e.g., EALR) in fundus images. MIL-CT utilizes the cross-scale vision transformer to extract retinal features in a multi-granularity perceptual domain. It incorporates a multi-head cross-scale attention fusion module to enhance global perceptual capability and feature representation. By integrating information from different scales and minimizing information loss, the method significantly improves the performance of the EALR detection task. Furthermore, a multi-instance learning module is implemented to enable the model to better comprehend local details and features in fundus images, facilitating the classification of patch tokens related to retinal lesions. To effectively learn the features associated with retinal lesions, we utilize weights pre-trained on a large fundus image Kaggle dataset. Our validation and comparison experiments conducted on our collected EALR dataset demonstrate the effectiveness of the MIL-CT method in reducing generalization errors while maintaining efficient attention to retinal vascular details. Moreover, the method surpasses existing models in EALR detection, achieving an accuracy, precision, sensitivity, specificity, and F1 score of 97.62%, 97.63%, 97.05%, 96.48%, and 97.62%, respectively. These results exhibit the significant enhancement in diagnostic accuracy of fundus images brought about by the MIL-CT method. Thus, it holds potential for various applications, particularly in the early screening of cardiovascular diseases such as hypertension and atherosclerosis. [ABSTRACT FROM AUTHOR]

Retinopathy, a prevalent disease causing visual impairment and sometimes blindness, affects many individuals in the population. Early detection and treatment of the disease can be facilitated by monitoring the retina using fundus imaging. Nonetheless, the limited availability of fundus images and the imbalanced datasets warrant the development of more precise and efficient algorithms to enhance diagnostic performance. This study presents a novel online knowledge distillation framework, called CLRD, which employs a collaborative learning approach for detecting retinopathy. By combining student models with varying scales and architectures, the CLRD framework extracts crucial pathological information from fundus images. The transfer of knowledge is accomplished by developing distortion information particular to fundus images, thereby enhancing model invariance. Our selection of student models includes the Transformer-based BEiT and the CNN-based ConvNeXt, which achieve accuracies of 98.77% and 96.88%, respectively. Furthermore, the proposed method has 5.69–23.13%, 5.37–23.73%, 5.74–23.17%, 11.24–45.21%, and 5.87–24.96% higher accuracy, precision, recall, specificity, and F1 score, respectively, compared to the advanced visual model. The results of our study indicate that the CLRD framework can effectively minimize generalization errors without compromising independent predictions made by student models, offering novel directions for further investigations into detecting retinopathy. [ABSTRACT FROM AUTHOR]

The accurate segmentation of multiple types of lesions from adjacent tissues in medical images is significant in clinical practice. Convolutional neural networks (CNNs) based on the coarse-to-fine strategy have been widely used in this field. However, multi-lesion segmentation remains to be challenging due to the uncertainty in size, contrast, and high interclass similarity of tissues. In addition, the commonly adopted cascaded strategy is rather demanding in terms of hardware, which limits the potential of clinical deployment. To address the problems above, we propose a novel Prior Attention Network (PANet) that follows the coarse-to-fine strategy to perform multi-lesion segmentation in medical images. The proposed network achieves the two steps of segmentation in a single network by inserting a lesion-related spatial attention mechanism in the network. Further, we also propose the intermediate supervision strategy for generating lesion-related attention to acquire the regions of interest (ROIs), which accelerates the convergence and obviously improves the segmentation performance. We have investigated the proposed segmentation framework in two applications: 2D segmentation of multiple lung infections in lung CT slices and 3D segmentation of multiple lesions in brain MRIs. Experimental results show that in both 2D and 3D segmentation tasks our proposed network achieves better performance with less computational cost compared with cascaded networks. The proposed network can be regarded as a universal solution to multi-lesion segmentation in both 2D and 3D tasks. The source code is available at https://github.com/hsiangyuzhao/PANet [ABSTRACT FROM AUTHOR]

Essential tremor (ET) is one of the most common movement disorders in adults, and its early assessment and diagnosis are crucial for disease management in movement disorders. Nowadays, the severity of tremors can only be diagnosed and evaluated by laboratory tests. However, there are certain subjective factors in traditional assessment methods by the naked eye of a neurologist, which often leads to some biases. This study proposes a novel multi-modal signals-based automated quantitative assessment system for tremor severity. Specifically, we develop a two-stage framework that performs posture pattern recognition on the raw data, then extracts kinematic parameters to build an individualized model for each task. Besides, we established a strict clinical paradigm, including 121 ET patients, finely evaluated by a committee of neurologists to build a high-quality database. The models' performances showed that most of the kinematic parameters designed in this study could effectively map the severity of the tremor. The F1 score for classification of the posture task based on deep learning networks was 99.02%, and the quantification of symptom scores based on machine learning models ranged from 94.77 to 99.00%. These results demonstrate the effectiveness of the proposed framework can automatically provide objective and accurate scores for ET symptom assessment. [ABSTRACT FROM AUTHOR]

Objective. Fluorescence molecular tomography (FMT) is a promising non-invasive optical molecular imaging technology with strong specificity and sensitivity that has great potential for preclinical and clinical studies in tumor diagnosis, drug development and therapeutic evaluation. However, the strong scattering of photons and insufficient surface measurements make it very challenging to improve the quality of FMT image reconstruction and its practical application for early tumor detection. Therefore, continuous efforts have been made to explore more effective approaches or solutions in the pursuit of high-quality FMT reconstructions. Approach. This review takes a comprehensive overview of advances in imaging methodology for FMT, mainly focusing on two critical issues in FMT reconstructions: improving the accuracy of solving the forward physical model and mitigating the ill-posed nature of the inverse problem from a methodological point of view. More importantly, numerous impressive and practical strategies and methods for improving the quality of FMT reconstruction are summarized. Notably, deep learning methods are discussed in detail to illustrate their advantages in promoting the imaging performance of FMT thanks to large datasets, the emergence of optimized algorithms and the application of innovative networks. Main results. The results demonstrate that the imaging quality of FMT can be effectively promoted by improving the accuracy of optical parameter modeling, combined with prior knowledge, and reducing dimensionality. In addition, the traditional regularization-based methods and deep neural network-based methods, especially end-to-end deep networks, can enormously alleviate the ill-posedness of the inverse problem and improve the quality of FMT image reconstruction. Significance. This review aims to illustrate a variety of effective and practical methods for the reconstruction of FMT images that may benefit future research. Furthermore, it may provide some valuable research ideas and directions for FMT in the future, and could promote, to a certain extent, the development of FMT and other methods of optical tomography. [ABSTRACT FROM AUTHOR]

Essential tremor (ET) is a prevalent neurological disorder that necessitates using objective and non-invasive methods for assessing symptom severity. Traditional visual assessments are often limited by subjectivity, while other wearable sensors or visual markers may result in unnatural movements. This study presents a novel contact-free visual-based pipeline for ET assessment that integrates refined whole-body pose estimation with Transformer-based tremor detection to quantify tremor severity at a fine-grained level. The proposed pose estimation method combines the Transformer with HRNet, effectively capturing spatial-temporal complementary information from multiple body parts and enabling highly accurate tremor detection. The Transformer-based tremor detection is well-suited for modeling long-range dependencies and sequential tremor data extracted by the pose estimation model, further improving the performance of our proposed method. Our study collected data from 61 patients with ET, achieving an average accuracy, recall, and F1 score of 95.6%/95.6%, 89.2%/95.0%, and 83.0%/92.4% for classifying ET severity both during rest and postural tasks, respectively. Our proposed method outperforms the temporal convolutional network baseline, increasing F1 scores by 21.17% and 14.22% for rest and postural tasks, respectively. This high level of accuracy makes our method highly useful for clinical applications such as remote monitoring, diagnosis, and treatment evaluation. Our proposed technique has many advantages over traditional ET assessment techniques, including non-invasiveness, contact-free operation, and not requiring any wearable sensors or visual markers. Moreover, our method can be applied to other movement disorders requiring objective measurements of symptom severity. In summary, our contact-free visual-based pipeline for ET assessment represents a significant improvement over traditional ET assessment techniques, and our quantification results demonstrate its potential for use in clinical settings. • We propose an automated video-based quantitative evaluation method for disorders. • We propose a Transformer model named TDT that incorporates kinematic features. • We got robust pose estimation by inserting a Transformer-based encoder into HRNet. • Reduce quantification errors with DARK to obtain stable key point predictions. • Provides key point features that help quantify the severity of essential tremor. [ABSTRACT FROM AUTHOR]

• The D2AFNet can exploit the channel-spatial and time series features to mine discriminative atrial fibrillation patterns. • The D2AFNet can profoundly explore the different contributions of spatial and temporal segments for excellent interpretation. • The D2AFNet method is tenfold cross-validated on the CPSC 2018 dataset and independently tested on the MIT-BIH dataset. • The D2AFNet method achieves the accuracies of 99.49% and 99.28% in the two-class and three-class AF detection tasks. Atrial fibrillation is one of the common and potentially dangerous persistent cardiac arrhythmias that are generally associated with the risk of stroke and heart failure. Manual electrocardiography diagnosis is the gold standard for the clinical detection of atrial fibrillation, but it has some drawbacks, such as being time-consuming and prone to misclassification due to inter-patient variability. Due to the powerful ability of deep learning to learn and extract rich features from huge datasets, end-to-end deep learning models are generally designed to detect abnormal atrial fibrillation signals automatically. However, these approaches usually ignore the key factors that feature maps from different channels and sequences may contribute differently to atrial fibrillation detection, making it challenging to implement accurate and interpretable models with better generalization performance. To tackle this challenge, we develop a dual-domain attention cascade D2AFNet for accurate and interpretable atrial fibrillation detection by cascading attention-based bidirectional gated recurrent units and densely connected networks embedded with channel-spatial information fusion modules. The D2AFNet can take full advantage of channel-spatial features to enhance the feature representation in the spatial domain, and then combine with the time series features in the temporal domain to form spatial–temporal fusion attention mechanisms to mine discriminative atrial fibrillation patterns. Besides, the D2AFNet can profoundly explore the different contributions of different spatial and temporal segments of feature maps for excellent interpretation. The proposed D2AFNet method is performed ten-fold cross-validation on the publicly available CPSC 2018 dataset, and achieves the accuracies of 99.49% and 99.28% in the two-class and three-class classification tasks, outperforming cutting-edge atrial fibrillation detection methods. In addition, the powerful generalization performance and inference efficiency of the D2AFNet method are also proved on another publicly available MIT-BIH dataset. The advantages of high performance and interpretability indicate that the D2AFNet method has huge potential in the computer-aided diagnosis of atrial fibrillation. [ABSTRACT FROM AUTHOR]

• An automated essential tremor assessment model based on a drawing task. • First database of 3 types of tremor tasks in patients with essential tremors. • The patient's diagnosis was independently scored by multiple neurologists. • Digital ink sequences were analyzed using a hybrid model of CNN and transformer. • The system incorporates both sequence features and kinematic handwriting features. Essential tremor and Parkinson's disease are common movement disorders, and early diagnosis and evaluation are critical to managing these diseases. Currently, laboratory tests are the only way to diagnose and assess tremor symptoms. Analysis of a patient's fine motor control, especially handwriting, is a powerful tool for disease assessment. However, traditional visual assessment methods by neurologists typically lead to biased diagnostic results due to some subjective factors. Therefore, it is necessary to automatically identify and quantify the captured motion events with the help of artificial intelligence in combination with the various dynamic attributes encapsulated in the digital ink features, such as pen pressure, stroke speed, handwriting variability, etc. In this paper, a novel Transformer deep-learning model is developed for sequence learning of electronic handwriting to effectively evaluate its potential in aiding the diagnosis of tremor symptoms. The one-dimensional convolution with an ingenious fusion attention mechanism is applied to the original pen sensor signal sequences and derived features are used as the embedding layer of the Transformer encoder part, and the global dynamic features are fused before the decision layer. Our proposed system performs excellent on private datasets and outperforms state-of-the-art methods on the PaHaW dataset. [ABSTRACT FROM AUTHOR]

• The effectiveness of multi-scale temporal fusion in sequence learning was proposed. • We use inter-patient paradigms to avoid skewing the classifications. • We tested the model on 3 datasets, showing strong robustness. • The model detected 5 arrhythmias with 99.57% F1-score, surpassing SOTA works. • Quarter memory consumption and more than 76.37% computation overhead reduction. Electrocardiogram automated arrhythmia detection plays a crucial role in the early prevention and diagnosis of cardiovascular diseases. However, previous research relies on noise removal algorithms and extracting solid features from raw ECGs. Besides, existing heartbeat classifiers ignore underlying complementary information of various scales, and intra-patient paradigms often lead to biased results. We constructed a novel end-to-end Multi-Scale Convolutional Neural Network-Sequence to Sequence architecture for heartbeat classification to address these issues. We have verified this approach on the clinical data collected by wearable devices and two heterogeneous datasets. The proposed model can effectively capture multi-granularity frequency and longitudinal temporal information by fusion representation and sequence learning. The overall F1 score of our approach was achieved at 99.57%, which exceeded the reference pure cascade model by 4.36%. [ABSTRACT FROM AUTHOR]