Fair evaluation of federated learning algorithms for automated breast density classification: The results of the 2022 ACR-NCI-NVIDIA federated learning challenge.

The correct interpretation of breast density is important in the assessment of breast cancer risk. AI has been shown capable of accurately predicting breast density, however, due to the differences in imaging characteristics across mammography systems, models built using data from one system do not generalize well to other systems. Though federated learning (FL) has emerged as a way to improve the generalizability of AI without the need to share data, the best way to preserve features from all training data during FL is an active area of research. To explore FL methodology, the breast density classification FL challenge was hosted in partnership with the American College of Radiology, Harvard Medical Schools' Mass General Brigham, University of Colorado, NVIDIA, and the National Institutes of Health National Cancer Institute. Challenge participants were able to submit docker containers capable of implementing FL on three simulated medical facilities, each containing a unique large mammography dataset. The breast density FL challenge ran from June 15 to September 5, 2022, attracting seven finalists from around the world. The winning FL submission reached a linear kappa score of 0.653 on the challenge test data and 0.413 on an external testing dataset, scoring comparably to a model trained on the same data in a central location.

A Transformer-Based Knowledge Distillation Network for Cortical Cataract Grading.

Cortical cataract, a common type of cataract, is particularly difficult to be diagnosed automatically due to the complex features of the lesions. Recently, many methods based on edge detection or deep learning were proposed for automatic cataract grading. However, these methods suffer a large performance drop in cortical cataract grading due to the more complex cortical opacities and uncertain data. In this paper, we propose a novel Transformer-based Knowledge Distillation Network, called TKD-Net, for cortical cataract grading. To tackle the complex opacity problem, we first devise a zone decomposition strategy to extract more refined features and introduce special sub-scores to consider critical factors of clinical cortical opacity assessment (location, area, density) for comprehensive quantification. Next, we develop a multi-modal mix-attention Transformer to efficiently fuse sub-scores and image modality for complex feature learning. However, obtaining the sub-score modality is a challenge in the clinic, which could cause the modality missing problem instead. To simultaneously alleviate the issues of modality missing and uncertain data, we further design a Transformer-based knowledge distillation method, which uses a teacher model with perfect data to guide a student model with modality-missing and uncertain data. We conduct extensive experiments on a dataset of commonly-used slit-lamp images annotated by the LOCS III grading system to demonstrate that our TKD-Net outperforms state-of-the-art methods, as well as the effectiveness of its key components. Codes are available at https://github.com/wjh892521292/Cataract_TKD-Net.

Prediction of postoperative visual acuity in patients with age-related cataracts using macular optical coherence tomography-based deep learning method.

Background: To predict postoperative visual acuity (VA) in patients with age-related cataracts using macular optical coherence tomography-based deep learning method. Methods: A total of 2,051 eyes from 2,051 patients with age-related cataracts were included. Preoperative optical coherence tomography (OCT) images and best-corrected visual acuity (BCVA) were collected. Five novel models (I, II, III, IV, and V) were proposed to predict postoperative BCVA. The dataset was randomly divided into a training (n = 1,231), validation (n = 410), and test set (n = 410). The performance of the models in predicting exact postoperative BCVA was evaluated using mean absolute error (MAE) and root mean square error (RMSE). The performance of the models in predicting whether postoperative BCVA was improved by at least two lines in the visual chart (0.2LogMAR) was evaluated using precision, sensitivity, accuracy, F1 and area under curve (AUC). Results: Model V containing preoperative OCT images with horizontal and vertical B-scans, macular morphological feature indices, and preoperative BCVA had a better performance in predicting postoperative VA, with the lowest MAE (0.1250 and 0.1194LogMAR) and RMSE (0.2284 and 0.2362LogMAR), and the highest precision (90.7% and 91.7%), sensitivity (93.4% and 93.8%), accuracy (88% and 89%), F1 (92% and 92.7%) and AUCs (0.856 and 0.854) in the validation and test datasets, respectively. Conclusion: The model had a good performance in predicting postoperative VA, when the input information contained preoperative OCT scans, macular morphological feature indices, and preoperative BCVA. The preoperative BCVA and macular OCT indices were of great significance in predicting postoperative VA in patients with age-related cataracts.

FraudAuditor: A Visual Analytics Approach for Collusive Fraud in Health Insurance.

Collusive fraud, in which multiple fraudsters collude to defraud health insurance funds, threatens the operation of the healthcare system. However, existing statistical and machine learning-based methods have limited ability to detect fraud in the scenario of health insurance due to the high similarity of fraudulent behaviors to normal medical visits and the lack of labeled data. To ensure the accuracy of the detection results, expert knowledge needs to be integrated with the fraud detection process. By working closely with health insurance audit experts, we propose FraudAuditor, a three-stage visual analytics approach to collusive fraud detection in health insurance. Specifically, we first allow users to interactively construct a co-visit network to holistically model the visit relationships of different patients. Second, an improved community detection algorithm that considers the strength of fraud likelihood is designed to detect suspicious fraudulent groups. Finally, through our visual interface, users can compare, investigate, and verify suspicious patient behavior with tailored visualizations that support different time scales. We conducted case studies in a real-world healthcare scenario, i.e., to help locate the actual fraud group and exclude the false positive group. The results and expert feedback proved the effectiveness and usability of the approach.

Nonalcoholic fatty liver disease (NAFLD) detection and deep learning in a Chinese community-based population.

OBJECTIVES: We aimed to develop and validate a deep learning system (DLS) by using an auxiliary section that extracts and outputs specific ultrasound diagnostic features to improve the explainable, clinical relevant utility of using DLS for detecting NAFLD. METHODS: In a community-based study of 4144 participants with abdominal ultrasound scan in Hangzhou, China, we sampled 928 (617 [66.5%] females, mean age: 56 years ± 13 [standard deviation]) participants (2 images per participant) to develop and validate DLS, a two-section neural network (2S-NNet). Radiologists' consensus diagnosis classified hepatic steatosis as none steatosis, mild, moderate, and severe. We also explored the NAFLD detection performance of six one-section neural network models and five fatty liver indices on our data set. We further evaluated the influence of participants' characteristics on the correctness of 2S-NNet by logistic regression. RESULTS: Area under the curve (AUROC) of 2S-NNet for hepatic steatosis was 0.90 for ≥ mild, 0.85 for ≥ moderate, and 0.93 for severe steatosis, and was 0.90 for NAFLD presence, 0.84 for moderate to severe NAFLD, and 0.93 for severe NAFLD. The AUROC of NAFLD severity was 0.88 for 2S-NNet, and 0.79-0.86 for one-section models. The AUROC of NAFLD presence was 0.90 for 2S-NNet, and 0.54-0.82 for fatty liver indices. Age, sex, body mass index, diabetes, fibrosis-4 index, android fat ratio, and skeletal muscle via dual-energy X-ray absorptiometry had no significant impact on the correctness of 2S-NNet (p > 0.05). CONCLUSIONS: By using two-section design, 2S-NNet had improved the performance for detecting NAFLD with more explainable, clinical relevant utility than using one-section design. KEY POINTS: • Based on the consensus review derived from radiologists, our DLS (2S-NNet) had an AUROC of 0.88 by using two-section design and yielded better performance for detecting NAFLD than using one-section design with more explainable, clinical relevant utility. • The 2S-NNet outperformed five fatty liver indices with the highest AUROCs (0.84-0.93 vs. 0.54-0.82) for different NAFLD severity screening, indicating screening utility of deep learning-based radiology may perform better than blood biomarker panels in epidemiology. • The correctness of 2S-NNet was not significantly influenced by individual's characteristics, including age, sex, body mass index, diabetes, fibrosis-4 index, android fat ratio, and skeletal muscle via dual-energy X-ray absorptiometry.

Mendelian randomization and clinical trial evidence supports TYK2 inhibition as a therapeutic target for autoimmune diseases.

BACKGROUND: To explore the associations of genetically proxied TYK2 inhibition with a wide range of disease outcomes and biomarkers to identify therapeutic repurposing opportunities, adverse effects, and biomarkers of efficacy. METHODS: The loss-of-function missense variant rs34536443 in TYK2 gene was used as a genetic instrument to proxy the effect of TYK2 inhibition. A phenome-wide Mendelian randomization (MR) study was conducted to explore the associations of genetically-proxied TYK2 inhibition with 1473 disease outcomes in UK Biobank (N = 339,197). Identified associations were examined for replication in FinnGen (N = 260,405). We further performed tissue-specific gene expression MR, colocalization analyses, and MR with 247 blood biomarkers. A systematic review of randomized controlled trials (RCTs) on TYK2 inhibitor was performed to complement the genetic evidence. FINDINGS: PheWAS-MR found that genetically-proxied TYK2 inhibition was associated with lower risk of a wide range of autoimmune diseases. The associations with hypothyroidism and psoriasis were confirmed in MR analysis of tissue-specific TYK2 gene expression and the associations with systemic lupus erythematosus, psoriasis, and rheumatoid arthritis were observed in colocalization analysis. There were nominal associations of genetically-proxied TYK2 inhibition with increased risk of prostate and breast cancer but not in tissue-specific expression MR or colocalization analyses. Thirty-seven blood biomarkers were associated with the TYK2 loss-of-function mutation. Evidence from RCTs confirmed the effectiveness of TYK2 inhibitors on plaque psoriasis and reported several adverse effects. INTERPRETATION: This study supports TYK2 inhibitor as a potential treatment for psoriasis and several other autoimmune diseases. Increased pharmacovigilance is warranted in relation to the potential adverse effects. FUNDING: None.

DPVisCreator: Incorporating Pattern Constraints to Privacy-preserving Visualizations via Differential Privacy.

Data privacy is an essential issue in publishing data visualizations. However, it is challenging to represent multiple data patterns in privacy-preserving visualizations. The prior approaches target specific chart types or perform an anonymization model uniformly without considering the importance of data patterns in visualizations. In this paper, we propose a visual analytics approach that facilitates data custodians to generate multiple private charts while maintaining user-preferred patterns. To this end, we introduce pattern constraints to model users' preferences over data patterns in the dataset and incorporate them into the proposed Bayesian network-based Differential Privacy (DP) model PriVis. A prototype system, DPVisCreator, is developed to assist data custodians in implementing our approach. The effectiveness of our approach is demonstrated with quantitative evaluation of pattern utility under the different levels of privacy protection, case studies, and semi-structured expert interviews.

A Task Decomposing and Cell Comparing Method for Cervical Lesion Cell Detection.

Automatic detection of cervical lesion cells or cell clumps using cervical cytology images is critical to computer-aided diagnosis (CAD) for accurate, objective, and efficient cervical cancer screening. Recently, many methods based on modern object detectors were proposed and showed great potential for automatic cervical lesion detection. Although effective, several issues still hinder further performance improvement of such known methods, such as large appearance variances between single-cell and multi-cell lesion regions, neglecting normal cells, and visual similarity among abnormal cells. To tackle these issues, we propose a new task decomposing and cell comparing network, called TDCC-Net, for cervical lesion cell detection. Specifically, our task decomposing scheme decomposes the original detection task into two subtasks and models them separately, which aims to learn more efficient and useful feature representations for specific cell structures and then improve the detection performance of the original task. Our cell comparing scheme imitates clinical diagnosis of experts and performs cell comparison with a dynamic comparing module (normal-abnormal cells comparing) and an instance contrastive loss (abnormal-abnormal cells comparing). Comprehensive experiments on a large cervical cytology image dataset confirm the superiority of our method over state-of-the-art methods.

Discriminative Cervical Lesion Detection in Colposcopic Images With Global Class Activation and Local Bin Excitation.

Accurate cervical lesion detection (CLD) methods using colposcopic images are highly demanded in computer-aided diagnosis (CAD) for automatic diagnosis of High-grade Squamous Intraepithelial Lesions (HSIL). However, compared to natural scene images, the specific characteristics of colposcopic images, such as low contrast, visual similarity, and ambiguous lesion boundaries, pose difficulties to accurately locating HSIL regions and also significantly impede the performance improvement of existing CLD approaches. To tackle these difficulties and better capture cervical lesions, we develop novel feature enhancing mechanisms from both global and local perspectives, and propose a new discriminative CLD framework, called CervixNet, with a Global Class Activation (GCA) module and a Local Bin Excitation (LBE) module. Specifically, the GCA module learns discriminative features by introducing an auxiliary classifier, and guides our model to focus on HSIL regions while ignoring noisy regions. It globally facilitates the feature extraction process and helps boost feature discriminability. Further, our LBE module excites lesion features in a local manner, and allows the lesion regions to be more fine-grained enhanced by explicitly modelling the inter-dependencies among bins of proposal feature. Extensive experiments on a number of 9888 clinical colposcopic images verify the superiority of our method (AP .75 = 20.45) over state-of-the-art models on four widely used metrics.

Full Scale Attention for Automated COVID-19 Diagnosis from CT Images.

The wide spread of coronavirus pneumonia (COVID-19) has been a severe threat to global health since 2019. Apart from the nucleic acid detection, medical imaging examination is a vital diagnostic modality to confirm and treat the disease. Thus, implementing the automatic diagnosis of the COVID-19 bears particular significance. However, the limitations of data quality and size strongly hinder the clas-sification and segmentation performance and it also result in high misdiagnosis rate. To this end, we propose a novel full scale attention mechanism (FUSA) to capture more contextual dependencies of features, which enables the model easier to classify positive cases and improve the sensitivity. Specifically, FUSA parallelly extracts the information of channel domain and spatial domain, and fuses them together. The experimental study shows FUSA can significantly improve the COVID-19 automated diagnosis performance and eliminate false negative cases compared with other state-of-the-art ones.

A Transfer Learning Based Super-Resolution Microscopy for Biopsy Slice Images: The Joint Methods Perspective.

Higher-resolution biopsy slice images reveal many details, which are widely used in medical practice. However, taking high-resolution slice images is more costly than taking low-resolution ones. In this paper, we propose a joint framework containing a novel transfer learning strategy and a deep super-resolution framework to generate high-resolution slice images from low-resolution ones. The super-resolution framework called SRFBN+ is proposed by modifying a state-of-the-art framework SRFBN. Specifically, the structure of the feedback block of SRFBN was modified to be more flexible. Besides, it is challenging to use typical transfer learning strategies directly for the tasks on slice images, as the patterns on different types of biopsy slice images are varying. To this end, we propose a novel transfer learning strategy, called Channel Fusion Transfer Learning (CF-Trans). CF-Trans builds a middle domain by fusing the data manifolds of the source domain and the target domain, serving as a springboard for knowledge transfer. Thus, in the transfer learning setting, SRFBN+ can be trained on the source domain and then the middle domain and finally the target domain. Experiments on biopsy slice images validate SRFBN+ works well in generating super-resolution slice images, and CF-Trans is an efficient transfer learning strategy.