IEEE Journal of Biomedical and Health Informatics最新文献

Photoplethysmography (PPG) is commonly used to gather health-related information but is highly affected by motion artifacts from daily activities. Inspired by the strong denoising capabilities and generalization of diffusion probabilistic models, this paper proposes a novel PPG denoising method using a diffusion probabilistic model to reduce the impact of these artifacts. While typical diffusion models handle Gaussian noises, motion artifacts often involve non-Gaussian noise. To address this, the proposed method incorporates noisy PPG signals into both the diffusion and reverse processes, allowing the model to adapt better to complex and non-Gaussian noises. A dataset with clean and noisy PPG signals from 15 subjects performing various motion tasks was collected for evaluation. The results show the proposed model significantly improves PPG signal quality, reducing the Peak-Rejection-Rate (PRR) from 0.24 to 0.03. It also enhances the accuracy of heart rate (HR) estimation and various heart rate variability (HRV) measures, showing robustness and good generalization across different tasks and subjects.

Brain atrophy emerges as a distinctive hallmark in various neurodegenerative diseases, demonstrating a progressive trajectory across diverse disease stages and concurrently manifesting in tandem with a discernible decline in cognitive abilities. Understanding the individualized patterns of brain atrophy is critical for precision medicine and the prognosis of neurodegenerative diseases. However, it is difficult to obtain longitudinal data to compare changes before and after the onset of diseases. In this study, we present a deep disentangled generative model (DDGM) for capturing individualized atrophy patterns via disentangling patient images into "realistic" healthy counterfactual images and abnormal residual maps. The proposed DDGM consists of four modules: normal MRI synthesis, residual map synthesis, input reconstruction module, and mutual information neural estimator (MINE). The MINE and adversarial learning strategy together ensure independence between disease-related features and features shared by both disease and healthy controls. In addition, we proposed a comprehensive evaluation of the effectiveness of synthetic pseudo-healthy images, focusing on both their healthiness and subject identity. The results indicated that the proposed DDGM effectively preserves these characteristics in the synthesized pseudo-healthy images, outperforming existing methods. The proposed method demonstrates robust generalization capabilities across two independent datasets from different races and sites. Analysis of the disease residual/saliency maps revealed specific atrophy patterns associated with Alzheimer's disease (AD), particularly in the hippocampus and amygdala regions. These accurate individualized atrophy patterns enhance the performance of AD classification tasks, resulting in an improvement in classification accuracy to 92.50 2.70%.

Computer-aided detection through deep learning is becoming a prevalent approach across various fields, including detection of anomalies in medical procedures. One such medical procedure involves the placement of medical tubes to provide nutrition or other medical procedures in critically ill patients. Medical tube placement can be highly complex and prone to subjective errors. Malposition of medical tubes is often observed, and associated with significant morbidity and mortality. In addition, continuous verification using manual procedures such as capnography, pH testing, auscultation, and visual inspection through chest X-ray (CXR) imaging is required. In this paper, we propose a Conv-MTD, a medical tube detection (MTD) model that detects the placement of medical tubes using CXR images, assisting radiologists with precise identification and categorizing the tubes into normal, abnormal, and borderline placement. Conv-MTD leverages the state-of-the-art EfficientNet-B7 architecture as its backbone, enhanced with auxiliary head in the intermediate layers to mitigate vanishing gradient issues common in deep neural networks. The Conv-MTD is further optimized using post-training 16-bit floating-point (FP16) quantization, which effectively reduces memory consumption and inference latency on resource-constrained devices. Conv-MTD provided the best performance, with an average area under the receiver-operator curve AUC-ROC of 0.95. The proposed Conv-MTD has the potential to operate on resource-constrained point-of-care devices, enabling low-cost and automated assessments in various healthcare settings.

Domain shifts between samples acquired with different instruments are one of the major challenges in accurate segmentation of Optical Coherence Tomography (OCT) images. Given that OCT images may be acquired with different devices in different clinical centers, this study presents astyle and structure data augmentation (SSDA) method to improve the adaptability of segmentation models. Inspired by our initial analysis of OCT domain differences, we propose an innovative hypothesis that domain shifts are primarily due to differences in image style and anatomical structure, which further guides the design of our method. By designing a modality-specific NURBS curve for style enhancement and implementing global and local elastic deformation fields, SSDA addresses both stylistic and structural variations in OCT data. Global deformations simulate changes in retinal curvature, while local deformations model layer-specific changes observed in OCT images. We validate our hypothesis through a comprehensive evaluation conducted on five OCT data domains, each differing in device type and imaging conditions. We train models on each of these domains for single-domain generalisation experiments and evaluate performance on the remaining unseen domains. The results show that SSDA outperforms existing methods when segmenting OCT images from different sources with different requirements for retinal layer segmentation. Specifically, across five different source domain generalisation experiments, SSDA achieves approximately 1.6% higher Dice and 2.6% improved MIOU, underscoring its superior segmentation accuracy and robust generalisation across all evaluated unseen domains. The source code can be found at: https://github.com/iMED-Lab/SSDA-OCTSeg.

Balance and gait impairments play a key role in falls among the elderly. Traditional clinical scales such as the Berg Balance Scale (BBS) to assess fall risk are often subjective, time consuming, and does not assess gait performance. Shorter assessments such as Timed Up and Go (TUG) are available, but most clinicians only look into the completion time. This study aimed to develop a fast, low-cost, and automated framework for balance function assessment and comprehensive gait analysis by enhancing the traditional TUG test with a markerless motion capture (MoCap) system and machine learning models. In total, we included TUG datasets of 70 participants with varying degrees of fall risk based on the BBS scores. We segmented TUG trials into five phases automatically using data from the MoCap system and extracted features from the phases. These features were then analyzed to identify those that significantly discriminate between high and low fall risk groups. Using the identified features, various machine learning models were tested to estimate the BBS scores. The markers obtained from the markerless MoCap system were used for detailed gait analysis, and lower limb kinematics were compared between the markerless and marker-based methods. Our findings indicate that individuals at high risk of falling had longer completion times, lower performance velocities, and smaller ranges of motion in lower-limb joints. Among the tested machine learning models, random forest demonstrated the best performance in predicting BBS scores (RMSE: 0.98, : 0.94). Additionally, our markerless MoCap system showed comparable accuracy to state-of-the-art systems, eliminating the need to attach markers or sensors. The findings could help develop a quick and objective tool for balance and gait assessment in older adults, providing quantitative data to improve screening and intervention planning.

Current camera-based infant monitoring mainly focuses on physiological measurement, overlooking its important semantic analysis potential for detecting accidental suffocation caused by oronasal occlusion during sleep. However, developing a robust infant suffocation risk detection model typically requires substantial labeled data, which is very difficult to obtain in real-world scenarios. To address this, we utilized the text-to-image diffusion model to generate diverse infant images depicting oronasal occlusion and non-occlusion scenarios controlled by text prompts. To ease the process of labeling, self- and semi-supervised learning algorithms are leveraged to learn the semantic information from unlabeled data with the support of minimal labeled data to train different model architectures. To evaluate the feasibility of this solution, we conducted a clinical trial in the neonatology department, which collected video data from 22 infants under various oronasal occlusion scenarios using breathable covers (e.g. clinical tissue). The clinical evaluation shows that most models trained on 25,000 generated images achieved over 90% performance on metrics of accuracy, recall, and F1-score, outperforming conventional approaches that pre-train and fine-tune the model using over 90,000 labeled task-related online images. This demonstrates the feasibility of leveraging text-to-image generated data to achieve robust camera-based infant suffocation risk detection, so as to secure the sleep safety of infants. More importantly, it beacons the potential of using text-based large-scale model to solve the general issue of scarcity of human data in artificial intelligence-based healthcare or clinical applications.

Exploiting multi-modal magnetic resonance imaging complementary information for brain tumor segmentation is still a challenging task. Existing methods are usually inclined to learn the joint representation of all tumor regions indiscriminately, thus salient sub-region or healthy tissue would be dominant during the training procedure, which leads to a biased and limited representation performance. In this study, a novel transformer-based multi-modal brain tumor segmentation approach is developed by decoupling and coupling strategy. First, Anatomy-induced Region Decoupler decouples the representation of the tumor scattered in different semantic sub-regions following anatomical view, which forces the model to fully learn intra-region representation separately with multiple modalities context. Additionally, we introduce the collaborative decoupling of the corresponding sub-region edge to serve auxiliary cues. We then design the Edge-supported Intra-region Coupler to separately couple edge and object learning within each anatomical sub-region structure. Lastly, the Mutual Cross-region Coupler is further applied to implement mutual improvement by coupling complementary gains among the above decoupled sub-regions. Extensive experiments clearly demonstrate that our method outperforms current state-of-the-arts for brain tumor segmentation on BRATS2018, BRATS2020, MSD, and BRATS2021 benchmarks while retaining high efficiency in the learning procedure.

While Electronic Prescription Systems (EPS) adoption varies across EU Member States, there's a lack of comprehensive comparative analysis. Existing studies focus on single EPSs, employ diverse methodologies, and lack up-to-date data. This study fills this gap by providing a comprehensive overview of EPS development, functionalities, and usage statistics in each EU Member State. Most EU Member States widely adopted EPS by 2022, with exceptions including Germany, France, and Luxembourg, where pilot projects or just plans existed at that time. Out of the 27 EPSs, 25 employ a similar design featuring a central server and end-user software or web-based applications. Among these, 22 are structured as single national systems. The fundamental technical solution is remarkably similar across the EU. Despite these similarities, functionalities, authentication methods, prescription validity, and medication coverage differ significantly among EPSs. A multinational team, including co-authors from each EU Member State, collected data using a structured questionnaire. The study underscores the need for standardized methodologies in EPS research and emphasizes the importance of comprehensive comparative analysis to inform healthcare policies and digitalization efforts.

Fast virtual stenting (FVS) is a promising preoperative planning aid for thoracic endovascular aortic repair (TEVAR) of aortic dissection. It aims at digitally predicting the reshaped aortic true lumen (TL) under specific operation plans (stent-graft deployment region and radius) to assess and avoid reoperation risk, but has not yet been applied clinically due to the difficulty in achieving accurate and time-dependent predictions. In this work, we propose a deep-learning-based model for FVS to solve the above problems. It models the FVS task as a time-dependent prediction of inner wall (TL surface) deformation and leverages outer wall (entire aortic surface) to improve it. Two point clouds (PC_iw and PC_ow) are generated to represent the walls, where patient information, operation plan, and post-operative time are set as the attributes of PC_iw. Afterwards, graphs are constructed based on the PCs and processed by a graph deep network to predict a point-wise inner wall deformation for generating the time-dependent reshaped TL. Our model successfully perceives and utilizes the virtual setting of operation plan and achieves the time-dependent predictions for 108 patients (269 real follow-up visits). Compared with the existing rule-based FVS model, it predicts the long-term reshaped TL with 9%, 5%, and 2% lower mean relative error of volume, surface area, and centerline length, respectively, and supports more accurate clinical measurements of poor outcome risk factors. Overall, our model may be of great significance for predicting reoperation risk, optimizing operation plan, and eventually improving the effectiveness and safety of TEVAR.

The exponential growth of sensitive patient information and diagnostic records in digital healthcare systems has increased the complexity of data protection, while frequent medical data breaches severely compromise system security and reliability. Existing privacy protection techniques often lack robustness and real-time capabilities in high-noise, high-packet-loss, and dynamic network environments, limiting their effectiveness in detecting healthcare data leaks. To address these challenges, we propose a Swarm Intelligence-Based Network Watermarking (SIBW) method for real-time privacy data leakage detection in digital healthcare systems. SIBW integrates fountain codes with outer error correction codes and employs a Multi-Phase Synergistic Swarm Optimization Algorithm (MPSSOA) to dynamically optimize encoding parameters, significantly enhancing the robustness and interference resistance of watermark detection. Additionally, a reliable synchronization sequence and lightweight embedding mechanism are designed to ensure adaptability to complex, dynamic networks. Experimental results demonstrate that SIBW achieves over 90% detection accuracy under high latency jitter and packet loss conditions, surpassing existing methods in both robustness and efficiency. With a compact design of only 3.7 MB, SIBW is particularly suited for rapid deployment in resource-constrained digital healthcare systems.