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.

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.

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.

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).

Collateral circulation is a critical determinant of clinical outcomes in acute ischemic stroke (AIS) patients and plays a key role in patient selection for endovascular therapy. This study aimed to develop an automated method for assessing and quantifying collateral circulation on multi-phase CT angiography, aiming to reduce observer variability and improve diagnostic efficiency. This retrospective study included mCTA images from 420 AIS patients within 14 hours of stroke symptom onset. A deep learning-based classification method with a tailored preprocessing module was developed to assess collateral circulation status. Manual evaluations using the simplified Menon method served as the ground truth. Model performance was assessed through five-fold cross-validation using metrics including accuracy, F1 score, precision, sensitivity, specificity, and the area under the receiver operating characteristic curve. The median age of the 420 patients was 73 years (IQR: 64-80 years; 222 men), and the median time from symptom onset to mCTA acquisition was 123 minutes (IQR: 79-245.5 minutes). The proposed framework achieved an accuracy of 87.6% for three-class collateral scores (good, intermediate, poor), with F1 score (85.7%), precision (83.8%), sensitivity (89.3%), specificity (92.9%), AUC (93.7%), ICC (0.832), and Kappa (0.781). For two-class collateral scores, we obtained 94.0% accuracy for good vs. non-good scores (F1 score(94.4%), precision (95.9%), sensitivity (93.0%), specificity (94.1%), AUC (97.1%),ICC(0.882),kappa(0.881)) and 97.1% for poor vs. non-poor scores (F1 score (98.5%), precision (98.0%), sensitivity (99.0%), specificity (84.8%), AUC (95.6%), ICC(0.740), kappa(0.738)). Additional analyses demonstrated that multi-phase CTA showed improved performance over single or two-phase CTA in collateral assessment. The proposed deep learning framework demonstrated high accuracy and consistency with radiologist-assigned scores for evaluating collateral circulation on multi-phase CTA in AIS patients. This method may offer a useful tool to aid clinical decision-making, reducing variability and improving diagnostic workflow. AIS = Acute Ischemic Stroke; mCTA = multi-phase Computed Tomography Angiography; DL = deep learning; AUC = area under the receiver operating characteristic curve; IQR = interquartile range; ROC = receiver operating characteristic.

Lung and colon cancers (LCC) are among the foremost reasons for human death and disease. Early analysis of this disorder contains various tests, namely ultrasound (US), magnetic resonance imaging (MRI), and computed tomography (CT). Despite analytical imaging, histopathology is one of the effective methods that delivers cell-level imaging of tissue under inspection. These are mainly due to a restricted number of patients receiving final analysis and early healing. Furthermore, there are probabilities of inter-observer faults. Clinical informatics is an interdisciplinary field that integrates healthcare, information technology, and data analytics to improve patient care, clinical decision-making, and medical research. Recently, deep learning (DL) proved to be effective in the medical sector, and cancer diagnosis can be made automatically by utilizing the capabilities of artificial intelligence (AI), enabling faster analysis of more cases cost-effectively. On the other hand, with extensive technical developments, DL has arisen as an effective device in medical settings, mainly in medical imaging. This study presents an Enhanced Fusion of Transfer Learning Models and Optimization-Based Clinical Biomedical Imaging for Accurate Lung and Colon Cancer Diagnosis (FTLMO-BILCCD) model. The main objective of the FTLMO-BILCCD technique is to develop an efficient method for LCC detection using clinical biomedical imaging. Initially, the image pre-processing stage applies the median filter (MF) model to eliminate the unwanted noise from the input image data. Furthermore, fusion models such as CapsNet, EffcientNetV2, and MobileNet-V3 Large are employed for the feature extraction. The FTLMO-BILCCD technique implements a hybrid of temporal pattern attention and bidirectional gated recurrent unit (TPA-BiGRU) for classification. Finally, the beluga whale optimization (BWO) technique alters the hyperparameter range of the TPA-BiGRU model optimally and results in greater classification performance. The FTLMO-BILCCD approach is experimented with under the LCC-HI dataset. The performance validation of the FTLMO-BILCCD approach portrayed a superior accuracy value of 99.16% over existing models.

The unrelenting progression of Parkinson's disease (PD) leads to severely impaired quality of life, with considerable variability in progression rates among patients. Identifying biomarkers of PD progression could improve clinical monitoring and management. Radiomics, which facilitates data extraction from imaging for use in machine learning models, offers a promising approach to this challenge. This study investigated the use of multi-modality imaging, combining conventional magnetic resonance imaging (MRI) and dopamine transporter single photon emission computed tomography (DAT-SPECT), to predict motor progression in PD. Motor progression was measured by changes in the Movement Disorder Society Unified Parkinson's Disease Rating Scale (MDS-UPDRS) motor subscale scores. Radiomic features were selected from the midbrain region in MRI and caudate nucleus, putamen, and ventral striatum in DAT-SPECT. Patients were stratified into fast progression vs. slow progression based on change in MDS-UPDRS in follow-up. Various feature selection methods and machine learning classifiers were evaluated for each modality, and the best-performing models were combined into an ensemble. On the internal test set, the ensemble model, which integrated clinical information, T1WI, T2WI and DAT-SPECT achieved a ROC AUC of 0.93 (95% CI: 0.80-1.00), PR AUC of 0.88 (95%CI 0.61-1.00), accuracy of 0.85 (95% CI: 0.70-0.89), sensitivity of 0.72 (95% CI: 0.43-1.00), and specificity of 0.92 (95% CI: 0.77-1.00). On the external test set, the ensemble model outperformed single-modality models with a ROC AUC of 0.77 (95% CI: 0.53-0.93), PR AUC of 0.79 (95% CI: 0.56-0.95), accuracy of 0.68 (95% CI: 0.50-0.86), sensitivity of 0.53 (95% CI: 0.27-0.82), and specificity of 0.82 (95% CI: 0.55-1.00). In conclusion, this study developed an imaging-based model to identify baseline characteristics predictive of disease progression in PD patients. The findings highlight the strength of using multiple imaging modalities and integrating imaging data with clinical information to enhance the prediction of motor progression in PD.

The purpose of this study is to mainly develop a predictive model based on clinicoradiological and radiomics features from preoperative gadobenate-enhanced (Gd-BOPTA) magnetic resonance imaging (MRI) using multilayer perceptron (MLP) deep learning to predict vessels encapsulating tumor clusters (VETC) in hepatocellular carcinoma (HCC) patients. A total of 230 patients with histopathologically confirmed HCC who underwent preoperative Gd-BOPTA MRI before hepatectomy were retrospectively enrolled from three hospitals (144, 54, and 32 in training, test, and validation set, respectively). Univariate and multivariate logistic regression analyses were used to determine independent clinicoradiological predictors significantly associated with VETC, which then constituted the clinicoradiological model. Regions of interest (ROIs) included four modes, intratumoral (Tumor), peritumoral area ≤ 2 mm (Peri2mm), intratumoral + peritumoral area ≤ 2 mm (Tumor + Peri2mm) and intratumoral integrated with peritumoral ≤ 2 mm as a whole (TumorPeri2mm). A total of 7322 radiomics features were extracted respectively for ROI(Tumor), ROI(Peri2mm), ROI(TumorPeri2mm) and 14644 radiomics features for ROI(Tumor + Peri2mm). Least absolute shrinkage and selection operator (LASSO) and univariate logistic regression analysis were used to select the important features. Seven different machine learning classifiers respectively combined the radiomics signatures selected from four ROIs to constitute different models, and compare the performance between them in three sets and then select the optimal combination to become the radiomics model we need. Then a radiomics score (rad-score) was generated, which combined significant clinicoradiological predictors to constituted the fusion model through multivariate logistic regression analysis. After comparing the performance of the three models using area under receiver operating characteristic curve (AUC), integrated discrimination index (IDI) and net reclassification index (NRI), choose the optimal predictive model for VETC prediction. Arterial peritumoral enhancement and peritumoral hypointensity on hepatobiliary phase (HBP) were independent risk factors for VETC, and constituted the Radiology model, without any clinical variables. Arterial peritumoral enhancement defined as the enhancement outside the tumor boundary in the late stage of arterial phase or early stage of portal phase, extensive contact with the tumor edge, which becomes isointense during the DP. MLP deep learning algorithm integrated radiomics features selected from ROI TumorPeri2mm was the best combination, which constituted the radiomics model (MLP model). A MLP score (MLP_score) was calculated then, which combining the two radiology features composed the fusion model (Radiology MLP model), with AUCs of 0.871, 0.894, 0.918 in the training, test and validation sets. Compared with the two models aforementioned, the Radiology MLP model demonstrated a 33.4%-131.3% improvement in NRI and a 9.3%-50% improvement in IDI, showing better discrimination, calibration and clinical usefulness in three sets, which was selected as the optimal predictive model. We mainly developed a fusion model (Radiology MLP model) that integrated radiology and radiomics features using MLP deep learning algorithm to predict vessels encapsulating tumor clusters (VETC) in hepatocellular carcinoma (HCC) patients, which yield an incremental value over the radiology and the MLP model.

Large language models (LLMs) have shown promising potential in analyzing complex textual data, including radiological reports. These models can assist clinicians, particularly those with limited experience, by integrating and presenting diagnostic criteria within radiological classifications. However, before clinical adoption, LLMs must be rigorously validated by medical professionals to ensure accuracy, especially in the context of advanced radiological classification systems. This study evaluates the performance of four LLMs-ChatGPT-4o, AmbossGPT, Claude 3.5 Sonnet, and Gemini 2.0 Flash-in classifying fractures based on the AO classification system using CT reports. A dataset of 292 fictitious physician-generated CT reports, representing 310 fractures, was used to assess the accuracy of each LLM in AO fracture classification retrospectively. Performance was evaluated by comparing the models' classifications to ground truth labels, with accuracy rates analyzed across different fracture types and subtypes. ChatGPT-4o and AmbossGPT achieved the highest overall accuracy (74.6 and 74.3%, respectively), outperforming Claude 3.5 Sonnet (69.5%) and Gemini 2.0 Flash (62.7%). Statistically significant differences were observed in fracture type classification, particularly between ChatGPT-4o and Gemini 2.0 Flash (Δ12%, p < 0.001). While all models demonstrated strong bone recognition rates (90-99%), their accuracy in fracture subtype classification remained lower (71-77%), indicating limitations in nuanced diagnostic categorization. LLMs show potential in assisting radiologists with initial fracture classification, particularly in high-volume or resource-limited settings. However, their performance remains inconsistent for detailed subtype classification, highlighting the need for further refinement and validation before clinical integration in advanced diagnostic workflows.

Quantitative imaging (QI) is demonstrating strong promise across multiple clinical applications. For clinical translation of QI methods, objective evaluation on clinically relevant tasks is essential. To address this need, multiple evaluation strategies are being developed. In this paper, based on previous literature, we outline four emerging frameworks to perform evaluation studies of QI methods. We first discuss the use of virtual imaging trials (VITs) to evaluate QI methods. Next, we outline a no-gold-standard evaluation framework to clinically evaluate QI methods without ground truth. Third, a framework to evaluate QI methods for joint detection and quantification tasks is outlined. Finally, we outline a framework to evaluate QI methods that output multi-dimensional parameters, such as radiomic features. We review these frameworks, discussing their utilities and limitations. Further, we examine future research areas in evaluation of QI methods. Given the recent advancements in PET, including long axial field-of-view scanners and the development of artificial-intelligence algorithms, we present these frameworks in the context of PET.