Histopathological evaluation of prostate biopsies using the Gleason scoring system is critical for prostate cancer diagnosis and treatment selection. However, grading variability among pathologists can lead to inconsistent assessments, risking inappropriate treatment. Similar challenges complicate the assessment of other prognostic features like cribriform cancer morphology and perineural invasion. Many pathology departments are also facing an increasingly unsustainable workload due to rising prostate cancer incidence and a decreasing pathologist workforce coinciding with increasing requirements for more complex assessments and reporting. Digital pathology and artificial intelligence (AI) algorithms for analysing whole slide images show promise in improving the accuracy and efficiency of histopathological assessments. Studies have demonstrated AI's capability to diagnose and grade prostate cancer comparably to expert pathologists. However, external validations on diverse data sets have been limited and often show reduced performance. Historically, there have been no well-established guidelines for AI study designs and validation methods. Diagnostic assessments of AI systems often lack preregistered protocols and rigorous external cohort sampling, essential for reliable evidence of their safety and accuracy. This study protocol covers the retrospective validation of an AI system for prostate biopsy assessment. The primary objective of the study is to develop a high-performing and robust AI model for diagnosis and Gleason scoring of prostate cancer in core needle biopsies, and at scale evaluate whether it can generalise to fully external data from independent patients, pathology laboratories and digitalisation platforms. The secondary objectives cover AI performance in estimating cancer extent and detecting cribriform prostate cancer and perineural invasion. This protocol outlines the steps for data collection, predefined partitioning of data cohorts for AI model training and validation, model development and predetermined statistical analyses, ensuring systematic development and comprehensive validation of the system. The protocol adheres to Transparent Reporting of a multivariable prediction model of Individual Prognosis Or Diagnosis+AI (TRIPOD+AI), Protocol Items for External Cohort Evaluation of a Deep Learning System in Cancer Diagnostics (PIECES), Checklist for AI in Medical Imaging (CLAIM) and other relevant best practices. Data collection and usage were approved by the respective ethical review boards of each participating clinical laboratory, and centralised anonymised data handling was approved by the Swedish Ethical Review Authority. The study will be conducted in agreement with the Helsinki Declaration. The findings will be disseminated in peer-reviewed publications (open access).

Artificial intelligence (AI) has significant potential in healthcare applications, but its training and deployment faces challenges due to healthcare's diverse data, complex tasks, and the need to preserve privacy. Foundation models that perform well on medical tasks and require less task-specific tuning data are critical to accelerate the development of healthcare AI applications. We introduce MedGemma, a collection of medical vision-language foundation models based on Gemma 3 4B and 27B. MedGemma demonstrates advanced medical understanding and reasoning on images and text, significantly exceeding the performance of similar-sized generative models and approaching the performance of task-specific models, while maintaining the general capabilities of the Gemma 3 base models. For out-of-distribution tasks, MedGemma achieves 2.6-10% improvement on medical multimodal question answering, 15.5-18.1% improvement on chest X-ray finding classification, and 10.8% improvement on agentic evaluations compared to the base models. Fine-tuning MedGemma further improves performance in subdomains, reducing errors in electronic health record information retrieval by 50% and reaching comparable performance to existing specialized state-of-the-art methods for pneumothorax classification and histopathology patch classification. We additionally introduce MedSigLIP, a medically-tuned vision encoder derived from SigLIP. MedSigLIP powers the visual understanding capabilities of MedGemma and as an encoder achieves comparable or better performance than specialized medical image encoders. Taken together, the MedGemma collection provides a strong foundation of medical image and text capabilities, with potential to significantly accelerate medical research and development of downstream applications. The MedGemma collection, including tutorials and model weights, can be found at https://goo.gle/medgemma.

Accurate detection of lymph node metastasis (LNM) in oral squamous cell carcinoma (OSCC) is crucial for treatment planning. This study developed a deep learning model using dual-energy CT to improve LNM detection. Preoperative dual-energy CT images (Iodine Map, Fat Map, monoenergetic 70 keV, and RHO/Z Map) and clinical data were collected from two centers. From the first center, 248 patients were divided into training (n=198) and internal validation (n=50) cohorts (8:2 ratio), while 106 patients from the second center comprised the external validation cohort. Region-of-interest images from all four sequences were stacked along the channel dimension to generate fused four-channel composite images. 16 deep learning models were developed as follows: three architectures (Crossformer, Densenet169, Squeezenet1_0) applied to each single-sequence/fused image, followed by MLP integration. Additionally, a Crossformer_Transformer model was constructed based on fused image. The top-performing model was compared against radiologists' assessments. Among the 16 deep learning models trained in this study, the Crossformer_Transformer model demonstrated the best diagnostic performance in predicting LNM in OSCC patients, with an AUC of 0.960 (95% CI: 0.9355-0.9842) on the training dataset, and 0.881 (95% CI: 0.7396-1.0000) and 0.881 (95% CI: 0.8033-0.9590) on the internal and external validation sets, respectively. The average AUC for radiologists across both validation cohorts (0.723-0.819) was lower than that of the model. The Crossformer_Transformer model, validated on multicenter data, shows strong potential for improving preoperative risk assessment and clinical decision-making in cervical LNM for OSCC patients.

To develop an advanced deep learning model for accurate differentiation between benign and malignant masses in ultrasound breast cancer screening, addressing the challenges of noise, blur, and complex tissue structures in ultrasound imaging. We propose Dynamic Multiscale Fusion Transformer (DMFormer), a novel Transformer-based architecture featuring a dynamic multiscale feature fusion mechanism. The model integrates window attention for local feature interaction with grid attention for global context mixing, enabling comprehensive capture of both fine-grained tissue details and broader anatomical contexts. DMFormer was evaluated on two independent datasets and compared against state-of-the-art approaches, including convolutional neural networks, Transformer-based architectures, and hybrid models. The model achieved areas under the curve of 90.48% and 86.57% on the respective datasets, consistently outperforming all comparison models. DMFormer demonstrates superior performance in ultrasound breast cancer detection through its innovative dual-attention approach. The model's ability to effectively balance local and global feature processing while maintaining computational efficiency represents a significant advancement in medical image analysis. These results validate DMFormer's potential for enhancing the accuracy and reliability of breast cancer screening in clinical settings.

The success of machine learning algorithms is inherently related to the extraction of meaningful features, as they play a pivotal role in the performance of these algorithms. Central to this challenge is the quality of data representation. However, the ability to generalize and extract these features effectively from unseen datasets is also crucial. In light of this, we introduce a novel concept: the polyrepresentation. Polyrepresentation integrates multiple representations of the same modality extracted from distinct sources, for example, vector embeddings from the Siamese Network, self-supervised models, and interpretable radiomic features. This approach yields better performance metrics compared to relying on a single representation. Additionally, in the context of X-ray images, we demonstrate the transferability of the created polyrepresentation to a smaller dataset, underscoring its potential as a pragmatic and resource-efficient approach in various image-related solutions. It is worth noting that the concept of polyprepresentation on the example of medical data can also be applied to other domains, showcasing its versatility and broad potential impact.

Deep neural networks are increasingly applied for automated histopathology. Yet, whole-slide images (WSIs) are often acquired at gigapixel sizes, rendering it computationally infeasible to analyze them entirely at high resolution. Diagnostic labels are largely available only at the slide-level, because expert annotation of images at a finer (patch) level is both laborious and expensive. Moreover, regions with diagnostic information typically occupy only a small fraction of the WSI, making it inefficient to examine the entire slide at full resolution. Here, we propose SASHA -- {\it S}equential {\it A}ttention-based {\it S}ampling for {\it H}istopathological {\it A}nalysis -- a deep reinforcement learning approach for efficient analysis of histopathological images. First, SASHA learns informative features with a lightweight hierarchical, attention-based multiple instance learning (MIL) model. Second, SASHA samples intelligently and zooms selectively into a small fraction (10-20\%) of high-resolution patches, to achieve reliable diagnosis. We show that SASHA matches state-of-the-art methods that analyze the WSI fully at high-resolution, albeit at a fraction of their computational and memory costs. In addition, it significantly outperforms competing, sparse sampling methods. We propose SASHA as an intelligent sampling model for medical imaging challenges that involve automated diagnosis with exceptionally large images containing sparsely informative features.

Background and aimsAcute pancreatitis (AP) is a common gastrointestinal disease with rising global incidence. While most cases are mild, severe AP (SAP) carries high mortality. Early and accurate severity prediction is crucial for optimal management. However, existing severity prediction models, such as BISAP and mCTSI, have modest accuracy and often rely on data unavailable at admission. This study proposes a deep learning (DL) model to predict AP severity using abdominal contrast-enhanced CT (CECT) scans acquired within 24 hours of admission. MethodsWe collected 10,130 studies from 8,335 patients across a multi-site U.S. health system. The model was trained in two stages: (1) self-supervised pretraining on large-scale unlabeled CT studies and (2) fine-tuning on 550 labeled studies. Performance was evaluated against mCTSI and BISAP on a hold-out internal test set (n=100 patients) and externally validated on a Hungarian AP registry (n=518 patients). ResultsOn the internal test set, the model achieved AUROCs of 0.888 (95% CI: 0.800-0.960) for SAP and 0.888 (95% CI: 0.819-0.946) for mild AP (MAP), outperforming mCTSI (p = 0.002). External validation showed robust AUROCs of 0.887 (95% CI: 0.825-0.941) for SAP and 0.858 (95% CI: 0.826-0.888) for MAP, surpassing mCTSI (p = 0.024) and BISAP (p = 0.002). Retrospective simulation suggested the models potential to support admission triage and serve as a second reader during CECT interpretation. ConclusionsThe proposed DL model outperformed standard scoring systems for AP severity prediction, generalized well to external data, and shows promise for providing early clinical decision support and improving resource allocation.

<b>Background:</b> Chest X-rays are rapidly gaining prominence as a prevalent diagnostic tool, as recognized by the World Health Organization (WHO). However, interpreting chest X-rays can be demanding and time-consuming, even for experienced radiologists, leading to potential misinterpretations and delays in treatment. <b>Method:</b> The purpose of this research is the development of a RadAI model. The RadAI model can accurately detect four types of lung abnormalities in chest X-rays and generate a report on each identified abnormality. Moreover, deep learning algorithms, particularly convolutional neural networks (CNNs), have demonstrated remarkable potential in automating medical image analysis, including chest X-rays. This work addresses the challenge of chest X-ray interpretation by fine tuning the following three advanced deep learning models: Feature-selective and Spatial Receptive Fields Network (FSRFNet50), ResNext50, and ResNet50. These models are compared based on accuracy, precision, recall, and F1-score. <b>Results:</b> The outstanding performance of RadAI shows its potential to assist radiologists to interpret the detected chest abnormalities accurately. <b>Conclusions:</b> RadAI is beneficial in enhancing the accuracy and efficiency of chest X-ray interpretation, ultimately supporting the timely and reliable diagnosis of lung abnormalities.

<b>Background:</b> Accurate upfront risk stratification in suspected clinically significant prostate cancer (csPCa) may reduce unnecessary prostate biopsies. Integrating clinical and Magnetic Resonance Imaging (MRI) variables using deep learning could improve prediction. <b>Methods:</b> We retrospectively analysed 538 men who underwent MRI and biopsy between April 2019-September 2024. A fully connected neural network was trained using 5-fold cross-validation. Model 1 included clinical features (age, prostate-specific antigen [PSA], PSA density, digital rectal examination, family history, prior negative biopsy, and ongoing therapy). Model 2 used MRI-derived Prostate Imaging Reporting and Data System (PI-RADS) categories. Model 3 used all previous variables as well as lesion size, location, and prostate volume as determined on MRI. <b>Results:</b> Model 3 achieved the highest area under the receiver operating characteristic curve (AUC = 0.822), followed by Model 2 (AUC = 0.778) and Model 1 (AUC = 0.716). Sensitivities for detecting clinically significant prostate cancer (csPCa) were 87.4%, 91.6%, and 86.8% for Models 1, 2, and 3, respectively. Although Model 3 had slightly lower sensitivity than Model 2, it showed higher specificity, reducing false positives and avoiding 43.4% and 21.2% more biopsies compared to Models 1 and 2. Decision curve analysis showed M2 had the highest net benefit at risk thresholds ≤ 20%, while M3 was superior above 20%. <b>Conclusions:</b> Model 3 improved csPCa risk stratification, particularly in biopsy-averse settings, while Model 2 was more effective in cancer-averse scenarios. These models support personalized, context-sensitive biopsy decisions.

Early recurrence (ER) following curative-intent surgery remains a major obstacle to improving long-term outcomes in patients with pancreatic cancer (PC). The accurate preoperative prediction of ER could significantly aid clinical decision-making and guide postoperative management. A retrospective cohort of 493 patients with histologically confirmed PC who underwent resection was analyzed. Contrast-enhanced computed tomography (CT) images were used for tumor segmentation, followed by radiomics and deep learning feature extraction. In total, four distinct feature selection algorithms were employed. Predictive models were constructed using random forest (RF) and support vector machine (SVM) classifiers. The model performance was evaluated by the area under the receiver operating characteristic curve (AUC). A comprehensive nomogram integrating feature scores and clinical factors was developed and validated. Among all of the constructed models, the Inte-SVM demonstrated superior classification performance. The nomogram, incorporating the Inte-feature score, CT-assessed lymph node status, and carbohydrate antigen 19-9 (CA19-9), yielded excellent predictive accuracy in the validation cohort (AUC = 0.920). Calibration curves showed strong agreement between predicted and observed outcomes, and decision curve analysis confirmed the clinical utility of the nomogram. A CT-based deep learning radiomics nomogram enabled the accurate preoperative prediction of early recurrence in patients with pancreatic cancer. This model may serve as a valuable tool to assist clinicians in tailoring postoperative strategies and promoting personalized therapeutic approaches.