Multimodal medical imaging integrates diverse data types, such as structural and functional neuroimaging, to provide complementary insights that enhance deep learning predictions and improve outcomes. This study focuses on a neuroimaging prediction framework based on both structural and functional neuroimaging data. We propose a next-generation prediction model, \textbf{MultiViT2}, which combines a pretrained representative learning base model with a vision transformer backbone for prediction output. Additionally, we developed a data augmentation module based on the latent diffusion model that enriches input data by generating augmented neuroimaging samples, thereby enhancing predictive performance through reduced overfitting and improved generalizability. We show that MultiViT2 significantly outperforms the first-generation model in schizophrenia classification accuracy and demonstrates strong scalability and portability.

Foundation models (FMs), large neural networks pretrained on extensive and diverse datasets, have revolutionized artificial intelligence and shown significant promise in medical imaging by enabling robust performance with limited labeled data. Although numerous surveys have reviewed the application of FM in healthcare care, brain imaging remains underrepresented, despite its critical role in the diagnosis and treatment of neurological diseases using modalities such as MRI, CT, and PET. Existing reviews either marginalize brain imaging or lack depth on the unique challenges and requirements of FM in this domain, such as multimodal data integration, support for diverse clinical tasks, and handling of heterogeneous, fragmented datasets. To address this gap, we present the first comprehensive and curated review of FMs for brain imaging. We systematically analyze 161 brain imaging datasets and 86 FM architectures, providing information on key design choices, training paradigms, and optimizations driving recent advances. Our review highlights the leading models for various brain imaging tasks, summarizes their innovations, and critically examines current limitations and blind spots in the literature. We conclude by outlining future research directions to advance FM applications in brain imaging, with the aim of fostering progress in both clinical and research settings.

Chest radiography is widely used in diagnostic imaging. However, perceptual errors -- especially overlooked but visible abnormalities -- remain common and clinically significant. Current workflows and AI systems provide limited support for detecting such errors after interpretation and often lack meaningful human--AI collaboration. We introduce RADAR (Radiologist--AI Diagnostic Assistance and Review), a post-interpretation companion system. RADAR ingests finalized radiologist annotations and CXR images, then performs regional-level analysis to detect and refer potentially missed abnormal regions. The system supports a "second-look" workflow and offers suggested regions of interest (ROIs) rather than fixed labels to accommodate inter-observer variation. We evaluated RADAR on a simulated perceptual-error dataset derived from de-identified CXR cases, using F1 score and Intersection over Union (IoU) as primary metrics. RADAR achieved a recall of 0.78, precision of 0.44, and an F1 score of 0.56 in detecting missed abnormalities in the simulated perceptual-error dataset. Although precision is moderate, this reduces over-reliance on AI by encouraging radiologist oversight in human--AI collaboration. The median IoU was 0.78, with more than 90% of referrals exceeding 0.5 IoU, indicating accurate regional localization. RADAR effectively complements radiologist judgment, providing valuable post-read support for perceptual-error detection in CXR interpretation. Its flexible ROI suggestions and non-intrusive integration position it as a promising tool in real-world radiology workflows. To facilitate reproducibility and further evaluation, we release a fully open-source web implementation alongside a simulated error dataset. All code, data, demonstration videos, and the application are publicly available at https://github.com/avutukuri01/RADAR.

Despite the importance of memories in everyday life and the progress made in understanding how they are encoded and retrieved, the neural processes by which declarative memories are maintained or forgotten remain elusive. Part of the problem is that it is empirically difficult to measure the rate at which memories fade, even between repeated presentations of the source of the memory. Without such a ground-truth measure, it is hard to identify the corresponding neural correlates. This study addresses this problem by comparing individual patterns of functional connectivity against behavioral differences in forgetting speed derived from computational phenotyping. Specifically, the individual-specific values of the speed of forgetting in long-term memory (LTM) were estimated for 33 participants using a formal model fit to accuracy and response time data from an adaptive paired-associate learning task. Individual speeds of forgetting were then used to examine participant-specific patterns of resting-state fMRI connectivity, using machine learning techniques to identify the most predictive and generalizable features. Our results show that individual speeds of forgetting are associated with resting-state connectivity within the default mode network (DMN) as well as between the DMN and cortical sensory areas. Cross-validation showed that individual speeds of forgetting were predicted with high accuracy (r = .78) from these connectivity patterns alone. These results support the view that DMN activity and the associated sensory regions are actively involved in maintaining memories and preventing their decline, a view that can be seen as evidence for the hypothesis that forgetting is a result of storage degradation, rather than of retrieval failure.

Brain metastases are a common complication of systemic cancer, affecting over 20% of patients with primary malignancies. Longitudinal magnetic resonance imaging (MRI) is essential for diagnosing patients, tracking disease progression, assessing therapeutic response, and guiding treatment selection. However, the manual review of longitudinal imaging is time-intensive, especially for patients with multifocal disease. Artificial intelligence (AI) offers opportunities to streamline image evaluation, but developing robust AI models requires comprehensive training data representative of real-world imaging studies. Thus, there is an urgent necessity for a large dataset with heterogeneity in imaging protocols and disease presentation. To address this, we present an open-access dataset of 11,884 longitudinal brain MRI studies from 1,430 patients with clinically confirmed brain metastases, paired with clinical and image metadata. The provided dataset will facilitate the development of AI models to assist in the long-term management of patients with brain metastasis.

Explainability features are intended to provide insight into the internal mechanisms of an AI device, but there is a lack of evaluation techniques for assessing the quality of provided explanations. We propose a framework to assess and report explainable AI features. Our evaluation framework for AI explainability is based on four criteria: 1) Consistency quantifies the variability of explanations to similar inputs, 2) Plausibility estimates how close the explanation is to the ground truth, 3) Fidelity assesses the alignment between the explanation and the model internal mechanisms, and 4) Usefulness evaluates the impact on task performance of the explanation. Finally, we developed a scorecard for AI explainability methods that serves as a complete description and evaluation to accompany this type of algorithm. We describe these four criteria and give examples on how they can be evaluated. As a case study, we use Ablation CAM and Eigen CAM to illustrate the evaluation of explanation heatmaps on the detection of breast lesions on synthetic mammographies. The first three criteria are evaluated for clinically-relevant scenarios. Our proposed framework establishes criteria through which the quality of explanations provided by AI models can be evaluated. We intend for our framework to spark a dialogue regarding the value provided by explainability features and help improve the development and evaluation of AI-based medical devices.

Inter reader variability and cross site domain shift challenge the automatic segmentation of prostate anatomy using T2 weighted MRI images. This study investigates whether transformer models can retain precision amid such heterogeneity. We compare the performance of UNETR and SwinUNETR in prostate gland segmentation against our previous 3D UNet model [1], based on 546 MRI (T2weighted) volumes annotated by two independent experts. Three training strategies were analyzed: single cohort dataset, 5 fold cross validated mixed cohort, and gland size based dataset. Hyperparameters were tuned by Optuna. The test set, from an independent population of readers, served as the evaluation endpoint (Dice Similarity Coefficient). In single reader training, SwinUNETR achieved an average dice score of 0.816 for Reader#1 and 0.860 for Reader#2, while UNETR scored 0.8 and 0.833 for Readers #1 and #2, respectively, compared to the baseline UNets 0.825 for Reader #1 and 0.851 for Reader #2. SwinUNETR had an average dice score of 0.8583 for Reader#1 and 0.867 for Reader#2 in cross-validated mixed training. For the gland size-based dataset, SwinUNETR achieved an average dice score of 0.902 for Reader#1 subset and 0.894 for Reader#2, using the five-fold mixed training strategy (Reader#1, n=53; Reader#2, n=87) at larger gland size-based subsets, where UNETR performed poorly. Our findings demonstrate that global and shifted-window self-attention effectively reduces label noise and class imbalance sensitivity, resulting in improvements in the Dice score over CNNs by up to five points while maintaining computational efficiency. This contributes to the high robustness of SwinUNETR for clinical deployment.

Synthesizing high quality CT projection data remains a significant challenge due to the limited availability of annotated data and the complex nature of CT imaging. In this work, we present PRO, a projection domain synthesis foundation model for CT imaging. To the best of our knowledge, this is the first study that performs CT synthesis in the projection domain. Unlike previous approaches that operate in the image domain, PRO learns rich structural representations from raw projection data and leverages anatomical text prompts for controllable synthesis. This projection domain strategy enables more faithful modeling of underlying imaging physics and anatomical structures. Moreover, PRO functions as a foundation model, capable of generalizing across diverse downstream tasks by adjusting its generative behavior via prompt inputs. Experimental results demonstrated that incorporating our synthesized data significantly improves performance across multiple downstream tasks, including low-dose and sparse-view reconstruction. These findings underscore the versatility and scalability of PRO in data generation for various CT applications. These results highlight the potential of projection domain synthesis as a powerful tool for data augmentation and robust CT imaging. Our source code is publicly available at: https://github.com/yqx7150/PRO.

Kernel size selection in Convolutional Neural Networks (CNNs) is a critical but often overlooked design decision that affects receptive field, feature extraction, computational cost, and model accuracy. This paper proposes the Best Kernel Size Estimation Function (BKSEF), a mathematically grounded and empirically validated framework for optimal, layer-wise kernel size determination. BKSEF balances information gain, computational efficiency, and accuracy improvements by integrating principles from information theory, signal processing, and learning theory. Extensive experiments on CIFAR-10, CIFAR-100, ImageNet-lite, ChestX-ray14, and GTSRB datasets demonstrate that BKSEF-guided architectures achieve up to 3.1 percent accuracy improvement and 42.8 percent reduction in FLOPs compared to traditional models using uniform 3x3 kernels. Two real-world case studies further validate the approach: one for medical image classification in a cloud-based setup, and another for traffic sign recognition on edge devices. The former achieved enhanced interpretability and accuracy, while the latter reduced latency and model size significantly, with minimal accuracy trade-off. These results show that kernel size can be an active, optimizable parameter rather than a fixed heuristic. BKSEF provides practical heuristics and theoretical support for researchers and developers seeking efficient and application-aware CNN designs. It is suitable for integration into neural architecture search pipelines and real-time systems, offering a new perspective on CNN optimization.

Background and purposeAccurate estimation of Overall Survival (OS) in Non-Small Cell Lung Cancer (NSCLC) patients provides critical insights for treatment planning. While previous studies showed that radiomics and Deep Learning (DL) features increased prediction accuracy, this study aimed to examine whether a model that combines the radiomics and DL features with the clinical and dosimetric features outperformed other models. Materials and methodsWe collected pre-treatment lung CT scans and clinical data for 225 NSCLC patients from the Maastro Clinic: 180 for training and 45 for testing. Radiomics features were extracted using the Python radiomics feature extractor, and DL features were obtained using a 3D ResNet model. An ensemble model comprising XGB and NN classifiers was developed using: (1) clinical features only; (2) clinical and radiomics features; (3) clinical and DL features; and (4) clinical, radiomics, and DL features. The performance metrics were evaluated for the test and K-fold cross-validation data sets. ResultsThe prediction model utilizing only clinical variables provided an Area Under the Receiver Operating Characteristic Curve (AUC) of 0.64 and a test accuracy of 77.55%. The best performance came from combining clinical, radiomics, and DL features (AUC: 0.84, accuracy: 85.71%). The prediction improvement of this model was statistically significant compared to models trained with clinical features alone or with a combination of clinical and radiomics features. ConclusionIntegrating radiomics and DL features with clinical characteristics improved the prediction of OS after radiotherapy for NSCLC patients. The increased accuracy of our integrated model enables personalized, risk-based treatment planning, guiding clinicians toward more effective interventions, improved patient outcomes and enhanced quality of life.