Breast ultrasound (BUS) computer-aided diagnosis (CAD) systems aims to perform two major steps: detecting lesions and classifying them as benign or malignant. However, the impact of combining both steps has not been previously addressed. Moreover, the specific method employed can influence the final outcome of the system. In this work, a comparison of the effects of using object detection, semantic segmentation and instance segmentation to detect lesions in BUS images was conducted. To this end, four approaches were examined: a) multi-class object detection, b) one-class object detection followed by localized region classification, c) multi-class segmentation, and d) one-class segmentation followed by segmented region classification. Additionally, a novel dataset for BUS segmentation, called BUS-UCLM, has been gathered, annotated and shared publicly. The evaluation of the methods proposed was carried out with this new dataset and four publicly available datasets: BUSI, OASBUD, RODTOOK and UDIAT. Among the four approaches compared, multi-class detection and multi-class segmentation achieved the best results when instance segmentation CNNs are used. The best results in detection were obtained with a multi-class Mask R-CNN with a COCO AP50 metric of 72.9%. In the multi-class segmentation scenario, Poolformer achieved the best results with a Dice score of 77.7%. The analysis of detection and segmentation models in BUS highlights several key challenges, emphasizing the complexity of accurately identifying and segmenting lesions. Among the methods evaluated, instance segmentation has proven to be the most effective for BUS images, offering superior performance in delineating individual lesions.

Retinal diseases recognition is still a challenging task. Many deep learning classification methods and their modifications have been developed for medical imaging. Recently, Vision Transformers (ViT) have been applied for classification of retinal diseases with great success. Therefore, in this study a novel method was proposed, the Residual Self-Attention Vision Transformer (RS-A ViT), for automatic detection of acquired vitelliform lesions (AVL), macular drusen as well as distinguishing them from healthy cases. The Residual Self-Attention module instead of Self-Attention was applied in order to improve model's performance. The new tool outperforms the classical deep learning methods, like EfficientNet, InceptionV3, ResNet50 and VGG16. The RS-A ViT method also exceeds the ViT algorithm, reaching 96.62%. For the purpose of this research a new dataset was created that combines AVL data gathered from two research centers and drusen as well as normal cases from the OCT dataset. The augmentation methods were applied in order to enlarge the samples. The Grad-CAM interpretability method indicated that this model analyses the appropriate areas in optical coherence tomography images in order to detect retinal diseases. The results proved that the presented RS-A ViT model has a great potential in classification retinal disorders with high accuracy and thus may be applied as a supportive tool for ophthalmologists.

Artificial intelligence (AI) tools for radiology are commonly unmonitored once deployed. The lack of real-time case-by-case assessments of AI prediction confidence requires users to independently distinguish between trustworthy and unreliable AI predictions, which increases cognitive burden, reduces productivity, and potentially leads to misdiagnoses. To address these challenges, we introduce Ensembled Monitoring Model (EMM), a framework inspired by clinical consensus practices using multiple expert reviews. Designed specifically for black-box commercial AI products, EMM operates independently without requiring access to internal AI components or intermediate outputs, while still providing robust confidence measurements. Using intracranial hemorrhage detection as our test case on a large, diverse dataset of 2919 studies, we demonstrate that EMM successfully categorizes confidence in the AI-generated prediction, suggesting different actions and helping improve the overall performance of AI tools to ultimately reduce cognitive burden. Importantly, we provide key technical considerations and best practices for successfully translating EMM into clinical settings.

Early and precise diagnosis of dyslexia is crucial for implementing timely intervention to reduce its effects. Timely identification can improve the individual's academic and cognitive performance. Traditional dyslexia detection (DD) relies on lengthy, subjective, restricted behavioral evaluations and interviews. Due to the limitations, deep learning (DL) models have been explored to improve DD by analyzing complex neurological, behavioral, and visual data. DL architectures, including convolutional neural networks (CNNs) and vision transformers (ViTs), encounter challenges in extracting meaningful patterns from cross-modality data. The lack of model interpretability and limited computational power restricts these models' generalizability across diverse datasets. To overcome these limitations, we propose an innovative model for DD using magnetic resonance imaging (MRI), electroencephalography (EEG), and handwriting images. We introduce a model, leveraging hybrid transformer-based feature extraction, including SWIN-Linformer for MRI, LeViT-Performer for handwriting images, and graph transformer networks (GTNs) with multi-attention mechanisms for EEG data. A multi-modal attention-based feature fusion network was used to fuse the extracted features in order to guarantee the integration of key multi-modal features. We enhance Dartbooster XGBoost (DXB)-based classification using Bayesian optimization with Hyperband (BOHB) algorithm. In order to reduce computational overhead, we employ a quantization-aware training technique. The local interpretable model-agnostic explanations (LIME) technique and gradient-weighted class activation mapping (Grad-CAM) were adopted to enable model interpretability. Five public repositories were used to train and test the proposed model. The experimental outcomes demonstrated that the proposed model achieves an accuracy of 99.8% with limited computational overhead, outperforming baseline models. It sets a novel standard for DD, offering potential for early identification and timely intervention. In the future, advanced feature fusion and quantization techniques can be utilized to achieve optimal results in resource-constrained environments.

Foundation models, pretrained on extensive datasets, have significantly advanced machine learning by providing robust and transferable embeddings applicable to various domains, including medical imaging diagnostics. This study evaluates the utility of embeddings derived from both general-purpose and medical domain-specific foundation models for training lightweight adapter models in multi-class radiography classification, focusing specifically on tube placement assessment. A dataset comprising 8842 radiographs classified into seven distinct categories was employed to extract embeddings using six foundation models: DenseNet121, BiomedCLIP, Med-Flamingo, MedImageInsight, Rad-DINO, and CXR-Foundation. Adapter models were subsequently trained using classical machine learning algorithms. Among these combinations, MedImageInsight embeddings paired with an support vector machine adapter yielded the highest mean area under the curve (mAUC) at 93.8%, followed closely by Rad-DINO (91.1%) and CXR-Foundation (89.0%). In comparison, BiomedCLIP and DenseNet121 exhibited moderate performance with mAUC scores of 83.0% and 81.8%, respectively, whereas Med-Flamingo delivered the lowest performance at 75.1%. Notably, most adapter models demonstrated computational efficiency, achieving training within one minute and inference within seconds on CPU, underscoring their practicality for clinical applications. Furthermore, fairness analyses on adapters trained on MedImageInsight-derived embeddings indicated minimal disparities, with gender differences in performance within 2% and standard deviations across age groups not exceeding 3%. These findings confirm that foundation model embeddings-especially those from MedImageInsight-facilitate accurate, computationally efficient, and equitable diagnostic classification using lightweight adapters for radiographic image analysis.

The automatic diagnosis of chest diseases is a popular and challenging task. Most current methods are based on convolutional neural networks (CNNs), which focus on local features while neglecting global features. Recently, self-attention mechanisms have been introduced into the field of computer vision, demonstrating superior performance. Therefore, this paper proposes an effective model, CheX-DS, for classifying long-tail multi-label data in the medical field of chest X-rays. The model is based on the excellent CNN model DenseNet for medical imaging and the newly popular Swin Transformer model, utilizing ensemble deep learning techniques to combine the two models and leverage the advantages of both CNNs and Transformers. The loss function of CheX-DS combines weighted binary cross-entropy loss with asymmetric loss, effectively addressing the issue of data imbalance. The NIH ChestX-ray14 dataset is selected to evaluate the model's effectiveness. The model outperforms previous studies with an excellent average AUC score of 83.76\%, demonstrating its superior performance.

To date, indications for a surgical approach of small pancreatic neuroendocrine tumours (PanNETs) remain controversial. This cohort study aimed to identify the pseudocapsule status preoperatively to estimate the rationality of enucleation and survival prognosis of PanNETs, particularly in small tumours. Clinicopathological data were collected from patients with PanNETs who underwent the first pancreatectomy at our hospital (n = 578) between February 2012 and September 2023. Kaplan-Meier curves were constructed to visualise prognostic differences. Five distinct tissue samples were obtained for single-cell RNA sequencing (scRNA-seq) to evaluate variations in the tumour microenvironment. Radiological features were extracted from preoperative arterial-phase contrast-enhanced computed tomography. The performance of the pseudocapsule radiomics model was assessed using the area under the curve (AUC) metric. 475 cases (mean [SD] age, 53.01 [12.20] years; female vs male, 1.24:1) were eligible for this study. The mean pathological diameter of tumour was 2.99 cm (median: 2.50 cm; interquartile range [IQR]: 1.50-4.00 cm). These cases were stratified into complete (223, 46.95%) and incomplete (252, 53.05%) pseudocapsule groups. A statistically significant difference in aggressive indicators was observed between the two groups (P < 0.001). Through scRNA-seq analysis, we identified that the incomplete group presented a markedly immunosuppressive microenvironment. Regarding the impact on recurrence-free survival, the 3-year and 5-year rates were 94.8% and 92.5%, respectively, for the complete pseudocapsule group, compared to 76.7% and 70.4% for the incomplete pseudocapsule group. The radiomics-predictive model has a significant discrimination for the state of the pseudocapsule, particularly in small tumours (AUC, 0.744; 95% CI, 0.652-0.837). By combining computed tomography-based radiomics and machine learning for preoperative identification of pseudocapsule status, the intact group is more likely to benefit from enucleation.

To investigate lung changes in patients with polymyositis/dermatomyositis-associated interstitial lung disease (PM/DM-ILD) using quantitative CT and to construct a diagnostic model to evaluate the application of quantitative CT and machine learning in diagnosing PM/DM-ILD. Chest CT images from 348 PM/DM individuals were quantitatively analyzed to obtain the lung volume (LV), mean lung density (MLD), and intrapulmonary vascular volume (IPVV) of the whole lung and each lung lobe. The percentage of high attenuation area (HAA %) was determined using the lung density histogram. Patients hospitalized from 2016 to 2021 were used as the training set (n=258), and from 2022 to 2023 were used as the temporal test set (n=90). Seven classification models were established, and their performance was evaluated through ROC analysis, decision curve analysis, calibration, and precision-recall curve. The optimal model was selected and interpreted with Python's SHAP model interpretation package. Compared to the non-ILD group, the mean lung density and percentage of high attenuation area in the whole lung and each lung lobe were significantly increased, and the lung volume and intrapulmonary vessel volume were significantly decreased in the ILD group. The Random Forest (RF) model demonstrated superior performance with the test set area under the curve of 0.843 (95% CI: 0.821-0.865), accuracy of 0.778, sensitivity of 0.784, and specificity of 0.750. Quantitative CT serves as an objective and precise method to assess pulmonary changes in PM/DM-ILD patients. The RF model based on CT quantitative parameters displayed strong diagnostic efficiency in identifying ILD, offering a new and convenient approach for evaluating and diagnosing PM/DM-ILD patients.

Clinical assessment of lymphedema, particularly for lymphedema severity and fluid-fibrotic lesions, remains challenging with traditional methods. We aimed to develop and validate a deep learning segmentation tool for automated tissue component analysis in lower extremity CT scans. For development datasets, lower extremity CT venography scans were collected in 118 patients with gynecologic cancers for algorithm training. Reference standards were created by segmentation of fat, muscle, and fluid-fibrotic tissue components using 3D slicer. A deep learning model based on the Unet++ architecture with an EfficientNet-B7 encoder was developed and trained. Segmentation accuracy of the deep learning model was validated in an internal validation set (n = 10) and an external validation set (n = 10) using Dice similarity coefficient (DSC) and volumetric similarity (VS). A graphical user interface (GUI) tool was developed for the visualization of the segmentation results. Our deep learning algorithm achieved high segmentation accuracy. Mean DSCs for each component and all components ranged from 0.945 to 0.999 in the internal validation set and 0.946 to 0.999 in the external validation set. Similar performance was observed in the VS, with mean VSs for all components ranging from 0.97 to 0.999. In volumetric analysis, mean volumes of the entire leg and each component did not differ significantly between reference standard and deep learning measurements (p > 0.05). Our GUI displays lymphedema mapping, highlighting segmented fat, muscle, and fluid-fibrotic components in the entire leg. Our deep learning algorithm provides an automated segmentation tool enabling accurate segmentation, volume measurement of tissue component, and lymphedema mapping. Question Clinical assessment of lymphedema remains challenging, particularly for tissue segmentation and quantitative severity evaluation. Findings A deep learning algorithm achieved DSCs > 0.95 and VS > 0.97 for fat, muscle, and fluid-fibrotic components in internal and external validation datasets. Clinical relevance The developed deep learning tool accurately segments and quantifies lower extremity tissue components on CT scans, enabling automated lymphedema evaluation and mapping with high segmentation accuracy.

Trauma-induced rib fractures are a common injury. The number and characteristics of these fractures influence whether a patient is treated nonoperatively or surgically. Rib fractures are typically diagnosed using CT scans, yet 19.2-26.8% of fractures are still missed during assessment. Another challenge in managing rib fractures is the interobserver variability in their classification. Purpose of this study was to develop and assess an automated method that detects rib fractures in CT scans, and classifies them according to the Chest Wall Injury Society (CWIS) classification. 198 CT scans were collected, of which 170 were used for training and internal validation, and 28 for external validation. Fractures and their classifications were manually annotated in each of the scans. A detection and classification network was trained for each of the three components of the CWIS classifications. In addition, a rib number labeling network was trained for obtaining the rib number of a fracture. Experiments were performed to assess the method performance. On the internal test set, the method achieved a detection sensitivity of 80%, at a precision of 87%, and an F1-score of 83%, with a mean number of FPPS (false positives per scan) of 1.11. Classification sensitivity varied, with the lowest being 25% for complex fractures and the highest being 97% for posterior fractures. The correct rib number was assigned to 94% of the detected fractures. The custom-trained nnU-Net correctly labeled 95.5% of all ribs and 98.4% of fractured ribs in 30 patients. The detection and classification performance on the external validation dataset was slightly better, with a fracture detection sensitivity of 84%, precision of 85%, F1-score of 84%, FPPS of 0.96 and 95% of the fractures were assigned the correct rib number. The method developed is able to accurately detect and classify rib fractures in CT scans, there is room for improvement in the (rare and) underrepresented classes in the training set.