AI-powered medical devices have driven the need for real-time, on-device inference such as biomedical image classification. Deployment of deep learning models at the edge is now used for applications such as anomaly detection and classification in medical images. However, achieving this level of performance on edge devices remains challenging due to limitations in model size and computational capacity. To address this, we present MedMambaLite, a hardware-aware Mamba-based model optimized through knowledge distillation for medical image classification. We start with a powerful MedMamba model, integrating a Mamba structure for efficient feature extraction in medical imaging. We make the model lighter and faster in training and inference by modifying and reducing the redundancies in the architecture. We then distill its knowledge into a smaller student model by reducing the embedding dimensions. The optimized model achieves 94.5% overall accuracy on 10 MedMNIST datasets. It also reduces parameters 22.8x compared to MedMamba. Deployment on an NVIDIA Jetson Orin Nano achieves 35.6 GOPS/J energy per inference. This outperforms MedMamba by 63% improvement in energy per inference.

Sex differences in mental health become more evident across adolescence, with a two-fold increase of prevalence of mood disorders in females compared to males. The brain underpinnings remain understudied. Here, we investigated the role of age, puberty and mental health in determining the longitudinal development of sex differences in brain structure. We captured sex differences in limbic and non-limbic structures using machine learning models trained in cross-sectional brain imaging data of 1132 youths, yielding limbic and non-limbic estimates of brain sex. Applied to two independent longitudinal samples (total: 8184 youths), our models revealed pronounced sex differences in brain structure with increasing age. For females, brain sex was sensitive to pubertal development (menarche) over time and, for limbic structures, to mood-related mental health. Our findings highlight the limbic system as a key contributor to the development of sex differences in the brain and the potential of machine learning models for brain sex classification to investigate sex-specific processes relevant to mental health.

Artifacts pose a significant challenge in medical imaging, impacting diagnostic accuracy and downstream analysis. While image-based approaches for detecting artifacts can be effective, they often rely on preprocessing methods that can lead to information loss and high-memory-demand medical images, thereby limiting the scalability of classification models. In this work, we propose the use of implicit neural representations (INRs) for image quality assessment. INRs provide a compact and continuous representation of medical images, naturally handling variations in resolution and image size while reducing memory overhead. We develop deep weight space networks, graph neural networks, and relational attention transformers that operate on INRs to achieve image quality assessment. Our method is evaluated on the ACDC dataset with synthetically generated artifact patterns, demonstrating its effectiveness in assessing image quality while achieving similar performance with fewer parameters.

In recent years, Artificial Intelligence (AI) has gained prominence as a robust tool for clinical decision-making and diagnostics, owing to its capacity to process and analyze large datasets with high accuracy. More specifically, Deep Learning, and its subclasses, have shown significant potential in image processing, including medical imaging and histological analysis. In forensic pathology, AI has been employed for the interpretation of histopathological data, identifying conditions such as myocardial infarction, traumatic injuries, and heart rhythm abnormalities. This review aims to highlight key advances in AI's role, particularly machine learning (ML) and deep learning (DL) techniques, in forensic neuropathology, with a focus on its ability to interpret instrumental and histopathological data to support professional diagnostics. A systematic review of the literature regarding applications of Artificial Intelligence in forensic neuropathology was carried out according to the Preferred Reporting Item for Systematic Review (PRISMA) standards. We selected 34 articles regarding the main applications of AI in this field, dividing them into two categories: those addressing traumatic brain injury (TBI), including intracranial hemorrhage or cerebral microbleeds, and those focusing on epilepsy and SUDEP, including brain disorders and central nervous system neoplasms capable of inducing seizure activity. In both cases, the application of AI techniques demonstrated promising results in the forensic investigation of cerebral pathology, providing a valuable computer-assisted diagnostic tool to aid in post-mortem computed tomography (PMCT) assessments of cause of death and histopathological analyses. In conclusion, this paper presents a comprehensive overview of the key neuropathology areas where the application of artificial intelligence can be valuable in investigating causes of death.

Reconstructing high-quality images from corrupted measurements remains a fundamental challenge in medical imaging. Recently, deep unrolling (DUN) methods have emerged as a promising solution, combining the interpretability of traditional iterative algorithms with the powerful representation capabilities of deep learning. However, their performance is often limited by weak information flow between iterative stages and a constrained ability to capture global features across stages-limitations that tend to worsen as the number of iterations increases.
Approach: In this work, we propose a memory-enhanced and multi-domain learning-based deep unrolling network for interpretable, high-fidelity medical image reconstruction. First, a memory-enhanced module is designed to adaptively integrate historical outputs across stages, reducing information loss. Second, we introduce a cross-stage spatial-domain learning transformer (CS-SLFormer) to extract both local and non-local features within and across stages, improving reconstruction performance. Finally, a frequency-domain consistency learning (FDCL) module ensures alignment between reconstructed and ground truth images in the frequency domain, recovering fine image details.
Main Results: Comprehensive experiments evaluated on three representative medical imaging modalities (PET, MRI, and CT) show that the proposed method consistently outperforms state-of-the-art (SOTA) approaches in both quantitative metrics and visual quality. Specifically, our method achieved a PSNR of 37.835 dB and an SSIM of 0.970 in 1 $\%$ dose PET reconstruction.
Significance: This study expands the use of model-driven deep learning in medical imaging, demonstrating the potential of memory-enhanced deep unrolling frameworks for high-quality reconstructions.

Objective&#xD;This study assesses rectum contours generated using a commercial deep learning auto-contouring model and compares them to clinician contours using geometry, changes in dosimetry and toxicity modelling. &#xD;Approach&#xD;This retrospective study involved 308 prostate cancer patients who were treated using 3D-conformal radiotherapy. Computed tomography images were input into Limbus Contour (v1.8.0b3) to generate auto-contour structures for each patient. Auto-contours were not edited after their generation.&#xD;Rectum auto-contours were compared to clinician contours geometrically and dosimetrically. Dice similarity coefficient (DSC), mean Hausdorff distance (HD) and volume difference were assessed. Dose-volume histogram (DVH) constraints (V41%-V100%) were compared, and a Wilcoxon signed rank test was used to evaluate statistical significance of differences. &#xD;Toxicity modelling to compare contours was carried out using equivalent uniform dose (EUD) and clinical factors of abdominal surgery and atrial fibrillation. Trained models were tested (80:20) in their prediction of grade 1 late rectal bleeding (ntotal=124) using area-under the receiver operating characteristic curve (AUC).&#xD;Main results&#xD;Median DSC (interquartile range (IQR)) was 0.85 (0.09), median HD was 1.38 mm (0.60 mm) and median volume difference was -1.73 cc (14.58 cc). Median DVH differences between contours were found to be small (<1.5%) for all constraints although systematically larger than clinician contours (p<0.05). However, an IQR up to 8.0% was seen for individual patients across all dose constraints.&#xD;Models using EUD alone derived from clinician or auto-contours had AUCs of 0.60 (0.10) and 0.60 (0.09). AUC for models involving clinical factors and dosimetry was 0.65 (0.09) and 0.66 (0.09) when using clinician contours and auto-contours.&#xD;Significance&#xD;Although median DVH metrics were similar, variation for individual patients highlights the importance of clinician review. Rectal bleeding prediction accuracy did not depend on the contour method for this cohort. The auto-contouring model used in this study shows promise in a supervised workflow.&#xD.

Magnetic resonance imaging (MRI) is commonly used in healthcare for its ability to generate diverse tissue contrasts without ionizing radiation. However, this flexibility complicates downstream analysis, as computational tools are often tailored to specific types of MRI and lack generalizability across the full spectrum of scans used in healthcare. Here, we introduce a versatile framework for the development and validation of AI models that can robustly process and analyze the full spectrum of scans achievable with MRI, enabling model deployment across scanner models, scan sequences, and age groups. Core to our framework is UltimateSynth, a technology that combines tissue physiology and MR physics in synthesizing realistic images across a comprehensive range of meaningful contrasts. This pan-contrast capability bolsters the AI development life cycle through efficient data labeling, generalizable model training, and thorough performance benchmarking. We showcase the effectiveness of UltimateSynth by training an off-the-shelf U-Net to generalize anatomical segmentation across any MR contrast. The U-Net yields highly robust tissue volume estimates, with variability under 4% across 150,000 unique-contrast images, 3.8% across 2,000+ low-field 0.3T scans, and 3.5% across 8,000+ images spanning the human lifespan from ages 0 to 100.

This study presents an unsupervised deep learning approach for computed tomography (CT) image reconstruction, leveraging the inherent similarities between deep neural network training and conventional iterative reconstruction methods. By incorporating forward and backward projection layers within the deep learning framework, we demonstrate the feasibility of reconstructing images from projection data without relying on ground-truth images. Our method is evaluated on the two-dimensional 2DeteCT dataset, showcasing superior performance in terms of mean squared error (MSE) and structural similarity index (SSIM) compared to traditional filtered backprojection (FBP) and maximum likelihood (ML) reconstruction techniques. Additionally, our approach significantly reduces reconstruction time, making it a promising alternative for real-time medical imaging applications. Future work will focus on extending this methodology to three-dimensional reconstructions and enhancing the adaptability of the projection geometry.

The current standard-of-care (SOC) practice for defining the clinical target volume (CTV) for radiation therapy (RT) in patients with glioblastoma still employs an isotropic 1-2 cm expansion of the T2-hyperintensity lesion, without considering the heterogeneous infiltrative nature of these tumors. This study aims to improve RT CTV definition in patients with glioblastoma by incorporating biologically relevant metabolic and physiologic imaging acquired before RT along with a deep learning model that can predict regions of subsequent tumor progression by either the presence of contrast-enhancement or T2-hyperintensity. The results were compared against two standard CTV definitions. Our multi-parametric deep learning model significantly outperformed the uniform 2 cm expansion of the T2-lesion CTV in terms of specificity (0.89 ± 0.05 vs 0.79 ± 0.11; p = 0.004), while also achieving comparable sensitivity (0.92 ± 0.11 vs 0.95 ± 0.08; p = 0.10), sparing more normal brain. Model performance was significantly enhanced by incorporating lesion size-weighted loss functions during training and including metabolic images as inputs.

Alzheimer's disease (AD) is the principal cause of dementia and requires the early diagnosis of people with mild cognitive impairment (MCI) who are at high risk of progressing. Early diagnosis is imperative for optimizing clinical management and selecting proper therapeutic interventions. Structural magnetic resonance imaging (MRI) markers have been widely investigated for predicting the conversion of MCI to AD, and recent advances in deep learning (DL) methods offer enhanced capabilities for identifying subtle neurodegenerative changes over time. We selected 228 MCI participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI) who had at least three T1-weighted MRI scans within 18 months of baseline. MRI volumes underwent bias correction, segmentation, and radiomics feature extraction. A 3D residual network (ResNet3D) was trained using a pairwise ranking loss to capture single-timepoint risk scores. Longitudinal analyses were performed by extracting deep convolutional neural network (CNN) embeddings and gray matter radiomics for each scan, which were put into a time-aware long short-term memory (LSTM) model with an attention mechanism. A single-timepoint ResNet3D model achieved modest performance (c-index ~ 0.70). Incorporating longitudinal MRI files or downstream survival models led to a pronounced prognostic improvement (c-index:0.80-0.90), but was not further improved by longitudinal radiomics data. Time-specific classification within two- and three-year and five-year windows after the last MRI acquisition showed high accuracy (AUC > 0.85). Several radiomics, including gray matter surface to volume and elongation, emerged as the most predictive features. Each SD change in the gray matter surface to volume change within the last visit was associated with an increased risk of developing AD (HR: 1.50; 95% CI: 1.25-1.79). These findings emphasize the value of structural MRI within the advanced DL architectures for predicting MCI-to-AD conversion. The approach may enable earlier risk stratification and targeted interventions for individuals most likely to progress. limitations in sample size and computational resources warrant larger, more diverse studies to confirm these observations and explore additional improvements.