Despite its high negative predictive value (NPV) for clinically significant prostate cancer (csPCa), MRI suffers from a substantial number of false positives, especially for intermediate-risk cases. In this work, we determine whether a deep learning model trained with PI-RADS-guided representation learning can disambiguate the PI-RADS 3 classification, detect csPCa from bi-parametric prostate MR images, and avoid unnecessary benign biopsies. This study included 28,263 MR examinations and radiology reports from 21,938 men imaged for known or suspected prostate cancer between 2015 and 2023 at our institution (21 imaging locations with 34 readers), with 6352 subsequent biopsies. We trained a deep learning model, a representation learner (RL), to learn how radiologists interpret conventionally acquired T2-weighted and diffusion-weighted MR images, using exams in which the radiologists are confident in their risk assessments (PI-RADS 1 and 2 for the absence of csPCa vs. PI-RADS 4 and 5 for the presence of csPCa, n=21,465). We then trained biopsy-decision models to detect csPCa (Gleason score ≥7) using these learned image representations, and compared them to the performance of radiologists, and of models trained on other clinical variables (age, prostate volume, PSA, and PSA density) for treatment-naïve test cohorts consisting of only PI-RADS 3 (n=253, csPCa=103) and all PI-RADS (n=531, csPCa=300) cases. On the 2 test cohorts (PI-RADS-3-only, all-PI-RADS), RL-based biopsy-decision models consistently yielded higher AUCs in detecting csPCa (AUC=0.73 [0.66, 0.79], 0.88 [0.85, 0.91]) compared with radiologists (equivocal, AUC=0.79 [0.75, 0.83]) and the clinical model (AUCs=0.69 [0.62, 0.75], 0.78 [0.74, 0.82]). In the PIRADS-3-only cohort, all of whom would be biopsied using our institution's standard of care, the RL decision model avoided 41% (62/150) of benign biopsies compared with the clinical model (26%, P<0.001), and improved biopsy yield by 10% compared with the PI-RADS ≥3 decision strategy (0.50 vs. 0.40). Furthermore, on the all-PI-RADS cohort, RL decision model avoided 27% of additional benign biopsies (138/231) compared to radiologists (33%, P<0.001) with comparable sensitivity (93% vs. 92%), higher NPV (0.87 vs. 0.77), and biopsy yield (0.75 vs. 0.64). The combination of clinical and RL decision models further avoided benign biopsies (46% in PI-RADS-3-only and 62% in all-PI-RADS) while improving NPV (0.82, 0.88) and biopsy yields (0.52, 0.76) across the 2 test cohorts. Our PI-RADS-guided deep learning RL model learns summary representations from bi-parametric prostate MR images that can provide additional information to disambiguate intermediate-risk PI-RADS 3 assessments. The resulting RL-based biopsy decision models also outperformed radiologists in avoiding benign biopsies while maintaining comparable sensitivity to csPCa for the all-PI-RADS cohort. Such AI models can easily be integrated into clinical practice to supplement radiologists' reads in general and improve biopsy yield for any equivocal decisions.

Once deployed, medical image analysis methods are often faced with unexpected image corruptions and noise perturbations. These unknown covariate shifts present significant challenges to deep learning based methods trained on "clean" images. This often results in unreliable predictions and poorly calibrated confidence, hence hindering clinical applicability. While recent methods have been developed to address specific issues such as confidence calibration or adversarial robustness, no single framework effectively tackles all these challenges simultaneously. To bridge this gap, we propose LaDiNE, a novel ensemble learning method combining the robustness of Vision Transformers with diffusion-based generative models for improved reliability in medical image classification. Specifically, transformer encoder blocks are used as hierarchical feature extractors that learn invariant features from images for each ensemble member, resulting in features that are robust to input perturbations. In addition, diffusion models are used as flexible density estimators to estimate member densities conditioned on the invariant features, leading to improved modeling of complex data distributions while retaining properly calibrated confidence. Extensive experiments on tuberculosis chest X-rays and melanoma skin cancer datasets demonstrate that LaDiNE achieves superior performance compared to a wide range of state-of-the-art methods by simultaneously improving prediction accuracy and confidence calibration under unseen noise, adversarial perturbations, and resolution degradation.

Breast cancer is the leading cause of death among women worldwide, and early detection through the standardized BI-RADS framework helps physicians assess the risk of malignancy and guide appropriate diagnostic and treatment decisions. In this study, an interpretable deep learning model (BIScreener) was proposed for predicting BI-RADS classifications from breast ultrasound images, aiding in the accurate assessment of breast cancer risk and improving diagnostic efficiency. BIScreener utilizes the stacked generalization of three pretrained convolutional neural networks to analyze ultrasound images obtained from two specific instruments (Mindray R5 and HITACHI) used at local hospitals. BIScreener achieved a classification total accuracy of 90.0% and ROC-AUC value of 0.982 in the external test set for five BI-RADS categories. The proposed method achieved 83.8% classification total accuracy and 0.967 ROC-AUC value for seven BI-RADS categories. In addition, the model improved the diagnostic accuracy of two radiologists by more than 8.1% for five BI-RADS categories and by more than 4.8% for seven BI-RADS categories and reduced the explanation time by more than 19.0%, demonstrating its potential to accelerate and improve the breast cancer diagnosis process.

One of the most promising auxiliaries for screening breast cancer (BC) is ultrasound (US) radio-frequency (RF) time series. It has the superiority of not requiring any supplementary equipment over other methods. This article sought to propound a machine learning (ML) method for the automated categorization of breast lesions-categorized as benign, probably benign, suspicious, or malignant-using features extracted from the accumulated US RF time series. In this research, 220 data points of the categories as mentioned earlier, recorded from 118 patients, were analyzed. The RFTSBU dataset was registered by a SuperSonic Imagine Aixplorer® medical/research system fitted with a linear transducer. The expert radiologist manually selected regions of interest (ROIs) in B-mode images before extracting 283 features from each ROI in the ML approach, utilizing textural features such as Gabor filter (GF), gray-level co-occurrence matrix (GLCM), gray-level run-length matrix (GLRLM), gray-level size zone matrix (GLSZM), and gray-level dependence matrix (GLDM). Subsequently, the particle swarm optimization (PSO) narrowed the features to 131 highly effective ones. Ultimately, the features underwent classification using an innovative multi-origin method classification (MOMC), marking a significant leap in BC diagnosis. Employing 5-fold cross-validation, the study achieved notable accuracy rates of 98.57 ± 1.09%, 91.53 ± 0.89%, and 83.71 ± 1.30% for 2-, 3-, and 4-class classifications, respectively, using MOMC-SVM and MOMC-ensemble classifiers. This research introduces an innovative ML-based approach to differentiate between diverse breast lesion types using in vivo US RF time series data. The findings underscore its efficacy in enhancing classification accuracy, promising significant strides in computer-aided diagnosis (CAD) for BC screening.

This study aimed to investigate the role of volumetric and dissemination parameters obtained from pretreatment 18-fluorodeoxyglucose PET/computed tomography (18F-FDG PET/CT) in predicting progression/relapse in patients with diffuse large B-cell lymphoma (DLBCL) with machine learning algorithms. Patients diagnosed with DLBCL histopathologically, treated with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone, and followed for at least 1 year were reviewed retrospectively. Quantitative parameters such as tumor volume [total metabolic tumor volume (tMTV)], tumor burden [total lesion glycolysis (tTLG)], and the longest distance between two tumor foci (Dmax) were obtained from PET images with a standard uptake value threshold of 4.0. The MTV obtained from the volume of interest with the highest volume was noted as metabolic bulk volume (MBV). By analyzing the patients' PET parameters and clinical information with machine learning algorithms, models that attempt to predict progression/recurrence over 1 year were obtained. Of the 90 patients included, 16 had progression within 1 year. Significant differences were found in tMTV, tTLG, MBV, and Dmax values between patients with and without progression. The area under curve (AUC) of the model obtained with clinical data was 0.701. While a model with an AUC of 0.871 was obtained with a random forest algorithm using PET parameters, the model obtained with the Naive Bayes algorithm including clinical data in PET parameters had an AUC of 0.838. Using quantitative parameters derived from staging PET with machine learning algorithms may enable us to detect early progression in patients with DLBCL and improve early risk stratification and guide treatment decisions in these patients.

Thyroid ultrasound (US) is an essential tool for detecting and characterizing thyroid nodules. In this study, we propose an innovative approach to enhance thyroid nodule assessment by integrating Doppler US images with grayscale US images through weakly supervised data augmentation networks (WSDAN). Our method reduces background noise by replacing inefficient augmentation strategies, such as random cropping, with an advanced technique guided by bounding boxes derived from Doppler US images. This targeted augmentation significantly improves model performance in both classification and localization of thyroid nodules. The training dataset comprises 1288 paired grayscale and Doppler US images, with an additional 190 pairs used for three-fold cross-validation. To evaluate the model's efficacy, we tested it on a separate set of 190 grayscale US images. Compared to five state-of-the-art models and the original WSDAN, our Enhanced WSDAN model achieved superior performance. For classification, it reached an accuracy of 91%. For localization, it achieved Dice and Jaccard indices of 75% and 87%, respectively, demonstrating its potential as a valuable clinical tool.

Accurate interpretation of multi-view radiographs is crucial for diagnosing fractures, muscular injuries, and other anomalies. While significant advances have been made in AI-based analysis of single images, current methods often struggle to establish robust correspondences between different X-ray views, an essential capability for precise clinical evaluations. In this work, we present a novel self-supervised pipeline that eliminates the need for manual annotation by automatically generating a many-to-many correspondence matrix between synthetic X-ray views. This is achieved using digitally reconstructed radiographs (DRR), which are automatically derived from unannotated CT volumes. Our approach incorporates a transformer-based training phase to accurately predict correspondences across two or more X-ray views. Furthermore, we demonstrate that learning correspondences among synthetic X-ray views can be leveraged as a pretraining strategy to enhance automatic multi-view fracture detection on real data. Extensive evaluations on both synthetic and real X-ray datasets show that incorporating correspondences improves performance in multi-view fracture classification.

Current artificial intelligence models for medical imaging are predominantly single modality and single disease. Attempts to create multimodal and multi-disease models have resulted in inconsistent clinical accuracy. Furthermore, training these models typically requires large, labour-intensive, well-labelled datasets. We developed MerMED-FM, a state-of-the-art multimodal, multi-specialty foundation model trained using self-supervised learning and a memory module. MerMED-FM was trained on 3.3 million medical images from over ten specialties and seven modalities, including computed tomography (CT), chest X-rays (CXR), ultrasound (US), pathology patches, color fundus photography (CFP), optical coherence tomography (OCT) and dermatology images. MerMED-FM was evaluated across multiple diseases and compared against existing foundational models. Strong performance was achieved across all modalities, with AUROCs of 0.988 (OCT); 0.982 (pathology); 0.951 (US); 0.943 (CT); 0.931 (skin); 0.894 (CFP); 0.858 (CXR). MerMED-FM has the potential to be a highly adaptable, versatile, cross-specialty foundation model that enables robust medical imaging interpretation across diverse medical disciplines.

Whole-heart coronary magnetic resonance angiography (CMRA) enables noninvasive and accurate detection of coronary artery stenosis. Nevertheless, the visual interpretation of CMRA is constrained by the observer's experience, necessitating substantial training. The purposes of this study were to develop a deep learning (DL) algorithm using a deep convolutional neural network to accurately detect significant coronary artery stenosis in CMRA and to investigate the effectiveness of this DL algorithm as a tool for assisting in accurate detection of coronary artery stenosis. Nine hundred and fifty-one coronary segments from 75 patients who underwent both CMRA and invasive coronary angiography (ICA) were studied. Significant stenosis was defined as a reduction in luminal diameter of >50% on quantitative ICA. A DL algorithm was proposed to classify CMRA segments into those with and without significant stenosis. A 4-fold cross-validation method was used to train and test the DL algorithm. An observer study was then conducted using 40 segments with stenosis and 40 segments without stenosis. Three radiology experts and 3 radiology trainees independently rated the likelihood of the presence of stenosis in each coronary segment with a continuous scale from 0 to 1, first without the support of the DL algorithm, then using the DL algorithm. Significant stenosis was observed in 84 (8.8%) of the 951 coronary segments. Using the DL algorithm trained by the 4-fold cross-validation method, the area under the receiver operating characteristic curve (AUC) for the detection of segments with significant coronary artery stenosis was 0.890, with 83.3% sensitivity, 83.6% specificity and 83.6% accuracy. In the observer study, the average AUC of trainees was significantly improved using the DL algorithm (0.898) compared to that without the algorithm (0.821, p<0.001). The average AUC of experts tended to be higher with the DL algorithm (0.897), but not significantly different from that without the algorithm (0.879, p=0.082). We developed a DL algorithm offering high diagnostic accuracy for detecting significant coronary artery stenosis on CMRA. Our proposed DL algorithm appears to be an effective tool for assisting inexperienced observers to accurately detect coronary artery stenosis in whole-heart CMRA.

Pneumonia is a respiratory condition characterized by inflammation of the alveolar sacs in the lungs, which disrupts normal oxygen exchange. This disease disproportionately impacts vulnerable populations, including young children (under five years of age) and elderly individuals (over 65 years), primarily due to their compromised immune systems. The mortality rate associated with pneumonia remains alarmingly high, particularly in low-resource settings where healthcare access is limited. Although effective prevention strategies exist, pneumonia continues to claim the lives of approximately one million children each year, earning its reputation as a "silent killer." Globally, an estimated 500 million cases are documented annually, underscoring its widespread public health burden. This study explores the design and evaluation of the CNN-based Computer-Aided Diagnostic (CAD) systems with an aim of carrying out competent as well as resourceful classification and categorization of chest radiographs into binary classes (Normal, Pneumonia). An augmented Kaggle dataset of 18,200 chest radiographs, split between normal and pneumonia cases, was utilized. This study conducts a series of experiments to evaluate lightweight CNN models--ShuffleNet, NASNet-Mobile, and EfficientNet-b0--using transfer learning that achieved accuracy of 90%, 88% and 89%, prompting the task for deep feature extraction from each of the networks and applying feature fusion to further pair it with SVM classifier and XGBoost classifier, achieving an accuracy of 97% and 98% resepectively. The proposed research emphasizes the crucial role of CAD systems in advancing radiological diagnostics, delivering effective solutions to aid radiologists in distinguishing between diagnoses by applying feature fusion, feature selection along with various machine learning algorithms and deep learning architectures.