Alzheimer's Disease (AD) as one of the most prevalent neurodegenerative disorders worldwide, characterized by significant memory and cognitive decline in its later stages, severely impacting daily lives. Consequently, early diagnosis and accurate assessment are crucial for delaying disease progression. In recent years, multimodal imaging has gained widespread adoption in AD diagnosis and research, particularly the combined use of Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET). The complementarity of these modalities in structural and metabolic information offers a unique advantage for comprehensive disease understanding and precise diagnosis. With the rapid advancement of deep learning techniques, efficient fusion of MRI and PET multimodal data has emerged as a prominent research focus. This review systematically surveys the latest advancements in deep learning-based multimodal fusion of MRI and PET images for AD research, with a particular focus on studies published in the past five years (2021-2025). It first introduces the main sources of AD-related data, along with data preprocessing and feature extraction methods. Then, it summarizes performance metrics and multimodal fusion techniques. Next, it explores the application of various deep learning models and their variants in multimodal fusion tasks. Finally, it analyzes the key challenges currently faced in the field, including data scarcity and imbalance, inter-institutional data heterogeneity, etc., and discusses potential solutions and future research directions. This review aims to provide systematic guidance for researchers in the field of MRI and PET multimodal fusion, with the ultimate goal of advancing the development of early AD diagnosis and intervention strategies.

Medical imaging data employed in research frequently comprises sensitive Protected Health Information (PHI) and Personal Identifiable Information (PII), which is subject to rigorous legal frameworks such as the General Data Protection Regulation (GDPR) or the Health Insurance Portability and Accountability Act (HIPAA). Consequently, these types of data must be de-identified prior to utilization, which presents a significant challenge for many researchers. Given the vast array of medical imaging data, it is necessary to employ a variety of de-identification techniques. To facilitate the de-identification process for medical imaging data, we have developed an open-source tool that can be used to de-identify Digital Imaging and Communications in Medicine (DICOM) magnetic resonance images, computer tomography images, whole slide images and magnetic resonance twix raw data. Furthermore, the implementation of a neural network enables the removal of text within the images. The proposed tool reaches comparable results to current state-of-the-art algorithms at reduced computational time (up to × 265). The tool also manages to fully de-identify image data of various types, such as Neuroimaging Informatics Technology Initiative (NIfTI) or Whole Slide Image (WSI-)DICOMS. The proposed tool automates an elaborate de-identification pipeline for multiple types of inputs, reducing the need for additional tools used for de-identification of imaging data. Question How can researchers effectively de-identify sensitive medical imaging data while complying with legal frameworks to protect patient health information? Findings We developed an open-source tool that automates the de-identification of various medical imaging formats, enhancing the efficiency of de-identification processes. Clinical relevance This tool addresses the critical need for robust and user-friendly de-identification solutions in medical imaging, facilitating data exchange in research while safeguarding patient privacy.

The growing integration of vision-language models (VLMs) in medical applications offers promising support for diagnostic reasoning. However, current medical VLMs often face limitations in generalization, transparency, and computational efficiency-barriers that hinder deployment in real-world, resource-constrained settings. To address these challenges, we propose a Reasoning-Aware Reinforcement Learning framework, \textbf{RARL}, that enhances the reasoning capabilities of medical VLMs while remaining efficient and adaptable to low-resource environments. Our approach fine-tunes a lightweight base model, Qwen2-VL-2B-Instruct, using Low-Rank Adaptation and custom reward functions that jointly consider diagnostic accuracy and reasoning quality. Training is performed on a single NVIDIA A100-PCIE-40GB GPU, demonstrating the feasibility of deploying such models in constrained environments. We evaluate the model using an LLM-as-judge framework that scores both correctness and explanation quality. Experimental results show that RARL significantly improves VLM performance in medical image analysis and clinical reasoning, outperforming supervised fine-tuning on reasoning-focused tasks by approximately 7.78%, while requiring fewer computational resources. Additionally, we demonstrate the generalization capabilities of our approach on unseen datasets, achieving around 27% improved performance compared to supervised fine-tuning and about 4% over traditional RL fine-tuning. Our experiments also illustrate that diversity prompting during training and reasoning prompting during inference are crucial for enhancing VLM performance. Our findings highlight the potential of reasoning-guided learning and reasoning prompting to steer medical VLMs toward more transparent, accurate, and resource-efficient clinical decision-making. Code and data are publicly available.

This systematic review and meta-analysis aimed to assess the diagnostic accuracy of radiomics in risk stratification of gastrointestinal stromal tumors (GISTs). It focused on evaluating radiomic models as a non-invasive tool in clinical practice. A comprehensive search was conducted across PubMed, Web of Science, EMBASE, Scopus, and Cochrane Library up to May 17, 2025. Studies involving preoperative imaging and radiomics-based risk stratification of GISTs were included. Quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool and Radiomics Quality Score (RQS). Pooled sensitivity, specificity, and area under the curve (AUC) were calculated using bivariate random-effects models. Meta-regression and subgroup analyses were performed to explore heterogeneity. A total of 29 studies were included, with 22 (76 %) based on computed tomography scans, while 2 (7 %) were based on endoscopic ultrasound, 3 (10 %) on magnetic resonance imaging, and 2 (7 %) on ultrasound. Of these, 18 studies provided sufficient data for meta-analysis. Pooled sensitivity, specificity, and AUC for radiomics-based GIST risk stratification were 0.84, 0.86, and 0.90 for training cohorts, and 0.84, 0.80, and 0.89 for validation cohorts. QUADAS-2 indicated some bias due to insufficient pre-specified thresholds. The mean RQS score was 13.14 ± 3.19. Radiomics holds promise for non-invasive GIST risk stratification, particularly with advanced imaging techniques. However, radiomic models are still in the early stages of clinical adoption. Further research is needed to improve diagnostic accuracy and validate their role alongside conventional methods like biopsy or surgery.

The past decade has witnessed remarkable advancements in musculoskeletal radiology, driven by increasing demand for medical imaging and rapid technological innovations. Contrary to early concerns about artificial intelligence (AI) replacing radiologists, AI has instead enhanced imaging capabilities, aiding in automated abnormality detection and workflow efficiency. MRI has benefited from acceleration techniques that significantly reduce scan times while maintaining high-quality imaging. In addition, novel MRI methodologies now support precise anatomic and quantitative imaging across a broad spectrum of field strengths. In CT, dual-energy and photon-counting technologies have expanded diagnostic possibilities for musculoskeletal applications. This review explores these key developments, examining their impact on clinical practice and the future trajectory of musculoskeletal radiology.

In radiotherapy planning, acquiring both magnetic resonance (MR) and computed tomography (CT) images is crucial for comprehensive evaluation and treatment. However, simultaneous acquisition of MR and CT images is time-consuming, economically expensive, and involves ionizing radiation, which poses health risks to patients. The objective of this study is to generate CT images from radiation-free MR images using a novel quasi-supervised learning framework. In this work, we propose a quasi-supervised framework to explore the underlying relationship between unpaired MR and CT images. Normalized mutual information (NMI) is employed as a similarity metric to evaluate the correspondence between MR and CT scans. To establish optimal pairings, we compute an NMI matrix across the training set and apply the Hungarian algorithm for global matching. The resulting MR-CT pairs, along with their NMI scores, are treated as prior knowledge and integrated into the training process to guide the MR-to-CT image translation model. Experimental results indicate that the proposed method significantly outperforms existing unsupervised image synthesis methods in terms of both image quality and consistency of image features during the MR to CT image conversion process. The generated CT images show a higher degree of accuracy and fidelity to the original MR images, ensuring better preservation of anatomical details and structural integrity. This study proposes a quasi-supervised framework that converts unpaired MR and CT images into structurally consistent pseudo-pairs, providing informative priors to enhance cross-modality image synthesis. This strategy not only improves the accuracy and reliability of MR-CT conversion, but also reduces reliance on costly and scarce paired datasets. The proposed framework offers a practical 1 and scalable solution for real-world medical imaging applications, where paired annotations are often unavailable.

Vision-language models (VLMs) pre-trained at large scale have shown unprecedented transferability capabilities and are being progressively integrated into medical image analysis. Although its discriminative potential has been widely explored, its reliability aspect remains overlooked. This work investigates their behavior under the increasingly popular split conformal prediction (SCP) framework, which theoretically guarantees a given error level on output sets by leveraging a labeled calibration set. However, the zero-shot performance of VLMs is inherently limited, and common practice involves few-shot transfer learning pipelines, which cannot absorb the rigid exchangeability assumptions of SCP. To alleviate this issue, we propose full conformal adaptation, a novel setting for jointly adapting and conformalizing pre-trained foundation models, which operates transductively over each test data point using a few-shot adaptation set. Moreover, we complement this framework with SS-Text, a novel training-free linear probe solver for VLMs that alleviates the computational cost of such a transductive approach. We provide comprehensive experiments using 3 different modality-specialized medical VLMs and 9 adaptation tasks. Our framework requires exactly the same data as SCP, and provides consistent relative improvements of up to 27% on set efficiency while maintaining the same coverage guarantees.

Medical segmentation plays an important role in clinical applications like radiation therapy and surgical guidance, but acquiring clinically acceptable results is difficult. In recent years, progress has been witnessed with the success of utilizing transformer-like models, such as combining the attention mechanism with CNN. In particular, transformer-based segmentation models can extract global information more effectively, compensating for the drawbacks of CNN modules that focus on local features. However, utilizing transformer architecture is not easy, because training transformer-based models can be resource-demanding. Moreover, due to the distinct characteristics in the medical field, especially when encountering mid-sized and small organs with compact regions, their results often seem unsatisfactory. For example, using ViT to segment medical images directly only gives a DSC of less than 50\%, which is far lower than the clinically acceptable score of 80\%. In this paper, we used Mask2Former with deformable attention to reduce computation and proposed offset adjustment strategies to encourage sampling points within the same organs during attention weights computation, thereby integrating compact foreground information better. Additionally, we utilized the 4th feature map in Mask2Former to provide a coarse location of organs, and employed an FCN-based auxiliary head to help train Mask2Former more quickly using Dice loss. We show that our model achieves SOTA (State-of-the-Art) performance on the HaNSeg and SegRap2023 datasets, especially on mid-sized and small organs.Our code is available at link https://github.com/earis/Offsetadjustment\_Background-location\_Decoder\_Mask2former.

Stroke ranks among the leading causes of disability and death worldwide. Timely detection can reduce its impact. Machine learning delivers powerful tools for image‑based diagnosis. This study introduces StrokeNeXt, a lightweight convolutional neural network (CNN) for computed tomography (CT) and magnetic resonance (MR) scans, and couples it with deep feature engineering (DFE) to improve accuracy and facilitate clinical deployment. We assembled a multimodal dataset of CT and MR images, each labeled as stroke or control. StrokeNeXt employs a ConvNeXt‑inspired block and a squeeze‑and‑excitation (SE) unit across four stages: stem, StrokeNeXt block, downsampling, and output. In the DFE pipeline, StrokeNeXt extracts features from fixed‑size patches, iterative neighborhood component analysis (INCA) selects the top features, and a t algorithm-based k-nearest neighbors (tkNN) classifier has been utilized for classification. StrokeNeXt achieved 93.67% test accuracy on the assembled dataset. Integrating DFE raised accuracy to 97.06%. This combined approach outperformed StrokeNeXt alone and reduced classification time. StrokeNeXt paired with DFE offers an effective solution for stroke detection on CT and MR images. Its high accuracy and fewer learnable parameters make it lightweight and it is suitable for integration into clinical workflows. This research lays a foundation for real‑time decision support in emergency and radiology settings.

Automatic segmentation of cervical tumors is important in quantitative analysis and radiotherapy planning. A parallel encoder U-Net (PEU-Net) integrating the multi-modality information of PET/MRI was proposed to segment cervical tumor, which consisted of two parallel encoders with the same structure for PET and MR images. The features of the two modalities were extracted separately and fused at each layer of the decoder. Res2Net module on skip connection aggregated the features of various scales and refined the segmentation performance. PET/MRI images of 165 patients with cervical cancer were included in this study. U-Net, TransUNet, and nnU-Net with single or multi-modality (PET or/and T2WI) input were used for comparison. The Dice similarity coefficient (DSC) with volume data, DSC and the 95th percentile of Hausdorff distance (HD95) with tumor slices were calculated to evaluate the performance. The proposed PEU-Net exhibited the best performance (DSC_3d: 0.726 ± 0.204, HD₉₅: 4.603 ± 4.579 mm), DSC_2d (0.871 ± 0.113) was comparable to the best result of TransUNet with PET/MRI (0.873 ± 0.125). The networks with multi-modality input outperformed those with single-modality images as input. The results showed that the proposed PEU-Net could use multi-modality information more effectively through the redesigned structure and achieved competitive performance.