Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference最新文献

This study evaluates the potential of a serious game, called Barn Ruins, as a spatial learning rehabilitation tool for older adults with mild to moderate cognitive impairment (MCI). The game's user navigates a maze environment on a laptop screen using a gaming controller (a joystick). It progresses through easy, medium, and hard routes, and has an error-based spatial learning score. The intervention spanned eight weeks, during which participants played the game for 30 minutes, three times a week. Pre-and post-intervention assessments were conducted using two independent and validated spatial orientation measures: VRNHouse as the primary and the Clock Orientation Test as the secondary outcome.Seven participants (86.3 ± 4.9 years, 2 males) completed the study. Although no statistically significant changes were observed in VRNHouse or Clock Orientation Test scores, 71.4% of participants improved or maintained their performance in the primary outcome measure, while 66.7% demonstrated improvement or stability in the secondary measure. Analysis of spatial learning scores within the Barn Ruins game revealed significant improvements over time (p = 0.0046, Kendall's W = 0.42), particularly in easy (p = 0.023) and hard (p = 0.01) routes. Performance on medium routes fluctuated, suggesting greater difficulty with these trials.Post-hoc comparisons revealed that by Weeks 7 and 8, participants' overall spatial learning scores were significantly higher compared to those in Week 1. Notably, easy routes exhibited a ceiling effect after Week 4, while harder routes showed consistent improvement after Week 5.Despite modest results in independent outcome measures, the game's significant performance gains suggest its utility in improving spatial skills. Future research with larger samples is needed to validate these findings.Clinical Relevance- These findings highlight the potential of the Barn Ruins game as a novel rehabilitation tool for enhancing spatial learning in older adults with MCI.

Patients with spasticity require intensive rehabilitation to regain motor function, but access is often limited by travel or healthcare constraints. Telerehabilitation provides a structured approach for home-based therapy, allowing patients to perform more exercises with minimal external assistance. This work presents a telerehabilitation platform for a compliant upper-limb robotized orthosis to support rehabilitation in both home and institutional settings with minimal setup required. Built using the OpenTera framework for rapid prototyping and modular service integration, the platform includes a patient user interface for guided exercises, progress tracking, and real-time feedback. A motor control module assists movement based on therapist-prescribed parameters, while role-based access control ensures secure data management for healthcare professionals.

Background: Latest placental epigenetic clocks (PlECs) were less accurate in earlier trimesters. We aimed to develop an R package of PlEC to estimate aging by DNA-methylation-based GA (DNAm-GA).

Methods: We developed a three-stage prediction model, including residual DNAm-GA for measuring placental aging. An R package was developed to simplify our scikit-learn models into a single function and to utilize DNAm-GA for placental aging study.

Results: Our PlEC achieved a lower root mean squared-error for preterm samples (0.558, 95% confidence interval [CI] 0.545, 0.570) compared to the two previous PlECs using the common dataset: (1) Lee et al. (1.696, 95% CI 1.667, 1.724); and (2) Mayne et al. (4.018, 95% CI 3.927, 4.108). We also provided a function to utilize DNAm-GA for placental aging study.

Conclusions: Our R package precisely estimated DNAm-GA and our analytical framework could utilize DNAm-GA for placental aging study.Clinical Relevance- Our placental epigenetic clock allows individual assessment of placental aging in clinical settings via the residual DNAm-GA.

This study investigates how various training strategies originally proposed in the context of image processing, can be used to improve the performance of a convolutional neural network designed for classification of seizures from EEG recordings.Random cropping of seizure segments, dropout, mixup and ensembling improved the performance of the baseline classifier, alone and in combination. The best results were obtained by a combination of random cropping, mixup and ensembling, improving the AUC from 0.957 to 0.981 and F1-score from 71.0% to 77.9%.This study shows the importance of optimizing the training of neural networks for seizure detection.

We propose a Topomap-based EEG decoding framework for distinguishing pictorial Imagination from Perception. By converting each trial's EEG signals into dense sequences of scalp voltage maps at short time intervals, our approach captures crucial spatiotemporal patterns that standard methods may overlook. We then apply a CNN with squeeze-and-excitation (SE) blocks to these Topomap "frames," enabling direct learning of both spatial topographies and rapid temporal fluctuations. Despite using only one trial per subject to simulate a data-scarce scenario, our model achieves 95.1% accuracy under a leave-one-subject-out (LOSO) cross-validation scheme. Results indicate clear neural distinctions between Imagination and Perception states, reflecting focused brain-region engagement during visual recall. In addition to confirming the viability of Topomaps as EEG feature representations, this study underscores their potential generalizability. We anticipate future extensions incorporating other modalities (orthographic, audio) and more advanced deep architectures will further expand the utility and robustness of this approach for brain-computer interface (BCI) applications.Clinical relevance- This framework offers a robust method for accurately distinguishing visual Imagination from Perception, even in data-scarce scenarios. It holds potential for enhancing diagnostic tools in cognitive disorders and refining BCI applications in clinical settings.

The CARDIOCARE project combines psycho-emotional and behavioral interventions into a mobile health (mHealth) application for the elderly breast cancer patients who have a risk for therapy-induced cardiac toxicity. The mHealth application includes psycho-emotional modules such as Expressive Writing, the ABCDE Model, and Best Possible Self alongside behavioral interventions like prompted voiding and pelvic floor exercises, which are focused on both improving psychological resilience and physical health. By focusing on both physical and psychological needs, the app aims to improve patient adherence to treatment and to alleviate healthcare burden. Preliminary results based on data analysis collected from 67 patients from six clinical centers show promising trends: 45% of patients initially expressed denial in their first entries using the ABCDE module, which later shifted towards acceptance and active coping strategies. These findings show the potential of the CARDIOCARE interventions to enhance the well-being in elderly cancer patients. Ongoing trials are expected to provide a more comprehensive understanding of these interventions and their impact on improving psychological well-being and overall quality of life of cancer patients.Clinical Relevance- This study investigates the potential of the CARDIOCARE mHealth application to address both psychological and physical needs in elderly cancer patients with therapy-induced cardiac toxicity. Preliminary results from six clinical centers indicate that the CARDIOCARE mHealth application can support elderly cancer patients by helping them to express their emotions, cope with their illness, and adopt healthy routines. These interventions could help clinicians enhance patient care by providing personalized support and remote monitoring, resulting in better quality of life outcomes.

Idiopathic Parkinson's disease (PD) is a progressive neurodegenerative disorder that affects the human nervous system. PD is the second most common neurodegenerative disorder worldwide, affecting more than 10 million people. Despite its high prevalence, the diagnosis still relies on subjective assessments that examine motor performance through clinical scales. The use of sensor-based technologies offers a promising opportunity to improve diagnosis processes as it might make the disease progression quantifiable and more objective. This paper presents a methodology for the automatic detection of PD using kinematic data in a novel database that is being recorded for the analysis of this disorder using motion capture technologies. Experiments are carried out with 37 patients with PD and 15 healthy subjects wearing Inertial Measurement Units (IMU) and photogrammetry systems while walking on flat terrain. Five different classification systems, including a novel transformer-based foundational model for tabular data (TabFPN), were used to discriminate between healthy controls and patients with PD and compared to more classical and tabular-based classification algorithms. The results indicate the abilities of TabFPN in automatically discriminating between PD and controls, reaching an accuracy of up to 78% and a ROC-AUC of 89%.

Epilepsy, a prevalent neurological disease, often requires accurate identification of the seizure onset zone (SOZ) in the brain for successful surgical removal of this region. SOZ identification is a lengthy process that traditionally relies on the knowledge and experience of the neurosurgeon. Advancements in artificial intelligence have opened the door to automatic SOZ localization. This study proposes the application of autoencoder-based anomaly detection in SOZ electrode classification for the first time. We trained the autoencoder in a leave-one-patient-out manner with electrocorticography (ECoG) signals from intact channels. The anomaly feature was determined by the maximum error between original and reconstructed signals for each channel of the test patient. We investigated the usefulness of anomaly features along with the interictal epileptiform discharge (IED) biomarker feature. A linear support vector machine classifier achieved 70.49% accuracy with the anomaly feature, 64.71% accuracy with IED feature, and 74.44% accuracy with combined anomaly and IED features. The study demonstrates the effectiveness of anomaly detection in the direct localization of SOZs and suggests that multiple biomarker features can enhance automatic SOZ localization performance.Clinical Relevance- This study demonstrates the clinical relevance of using anomaly features from ECoG data for efficient SOZ localization. It highlights the effectiveness of combining anomaly features with IED biomarkers to enhance the automatic SOZ classification performance and improve surgical planning in epilepsy treatment.

Vagus Nerve Stimulation (VNS) is an effective therapy for drug-resistant epilepsy (DRE), but its efficacy varies across individuals. This study uses the weighted Phase Lag Index (wPLI) across multiple frequency bands to examine the dose-dependent regional brain synchronization effect at an individual level. EEG data from four DRE patients (two responders and two non-responders) were analyzed under controlled VNS intensities. Results show that non-responders exhibit an increased synchrony at higher intensity levels, particularly in the beta band and the broadband. Findings highlight individualized neural responses to VNS, underscoring the possibility for personalized stimulation strategies to optimize therapy.Clinical Relevance- Refining the VNS titration process through dose-dependent brain connectivity analysis could accelerate optimization, improving therapeutic outcomes and personalization for drug-resistant epilepsy patients.

This manuscript introduces a deep learning algorithm designed for spatial and temporal source reconstruction based on signals captured by MEG devices. Estimating brain signals at the source level is a significant challenge in magnetoencephalographic (MEG) data processing. Traditional algorithms excel in temporal resolution but face limitations in spatial resolution due to the inherently ill-posed nature of the problem. However, precise localization of pathological tissues is often crucial for providing reliable information to clinicians, which makes this a key area for improvement. Deep learning solutions have emerged as promising candidates for high-resolution signal estimation in this context. The proposed approach, called 'Deep-MEG', utilizes a hybrid neural network architecture capable of extracting both temporal and spatial information from MEG sensor signals. Unlike traditional methods, the algorithm can handle the entire brain, making it suitable for imaging not just cortical sources but also subcortical ones. To validate its performance, the authors conducted simulations with multiple active sources using a realistic forward model and compared the results with those from various state-of-the-art reconstruction algorithms.Clinical relevance- This study represent a first approach for accurate deep source localization and reconstruction leading to diagnosis support to clinicians.