Using machine learning to identify Parkinson’s disease severity subtypes with multimodal data

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: April 21, 2025

Background Classifying and predicting Parkinson's disease (PD) is challenging because of its diverse subtypes based on severity levels. Currently, identifying objective biomarkers associated with that can distinguish PD in clinical trials necessary. This study aims to address the applicability heterogeneity using classification digital biomarker development by combining multimodal data machine learning (ML) approaches. Methods We analyzed datasets combine characteristics, physical function lifestyle data, gait parameters motion analysis systems, wearable sensors collected from persons (n = 102) perform clustering for subtype classification. Results identified three subtypes, each exhibiting different patterns severity, increasing as it progressed clusters 1 3. found significant mutual information between all/single modalities unified rating scale scores, potential high feature importance ML. Among all modalities, principal components derived were most indicators severity. A model utilizing first component left right ankle achieved perfect an area under curve 1.0, accurately distinguishing clinically severe mild PD. These findings suggest features both ankles reflect asymmetry factors which contributes performance. Conclusions Digital obtained attached bilaterally body segments demonstrate classifying tracking progression. Our emphasized value sensor-based management, suggested integration into personalized monitoring systems therapeutic interventions

Language: Английский

---

A novel dynamic weighted prediction framework with stability-enhanced dynamic thresholding feature selection for neurodegenerative disease detection using gait features

Journal Of Big Data, Journal Year: 2025, Volume and Issue: 12(1)

Published: April 15, 2025

Language: Английский

---

Gait analysis for Parkinson’s disease using multiscale entropy

Neurodegenerative Disease Management, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 14

Published: May 26, 2025

Parkinson's disease (PD) is a progressive neurodegenerative disorder marked by motor dysfunction and complex gait abnormalities. Traditional linear methods often fail to capture the intricate movement patterns in PD. This review highlights Multiscale Entropy (MSE) as promising tool for assessing dynamics, offering deeper insights into variability across multiple temporal scales. MSE distinguishes healthy pathological patterns, enhancing early diagnosis monitoring. Advances wearable sensors, artificial intelligence, machine learning have boosted MSE's clinical relevance enabling real-time, personalized assessments. Despite these benefits, faces challenges such computational demands need high-resolution data. Addressing limitations through large-scale studies, standardized protocols, integration of emerging technologies may support broader adoption development robust normative database.

Language: Английский

---

A Systematic Review of Gait Analysis in the Context of Multimodal Sensing Fusion and AI

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2023, Volume and Issue: 31, P. 4189 - 4202

Published: Jan. 1, 2023

Background: Neurological diseases are a leading cause of disability and mortality. Gait, or human walking, is significant predictor quality life, morbidity, Gait patterns other kinematic, kinetic, balance gait features accurate powerful diagnostic prognostic tools. Objective: This review article focuses on the applicability analysis using fusion techniques artificial intelligence (AI) models. The aim to examine significance mixing several types wearable non-wearable sensor data impact this combination performance AI Method: In systematic review, 66 studies more than two modalities record analyze were identified. 40 incorporated multiple without use extract such as margin stability, temporal, spatial parameters, well cerebral activity. Similarly, 26 analyzed multimodal sensors algorithms. Results: research summarized here demonstrates that effectiveness models can both benefit from integration many sensors. Meanwhile, utilization EMG signals in especially advantageous. Conclusion: findings suggest smart, portable, wearable-based assessment system be developed sensing most cutting-edge, clinically relevant tools technology available. information presented may serve vital springboard for development.

Language: Английский

---

Effect of robot-assisted gait training on motor dysfunction in Parkinson’s patients:A systematic review and meta-analysis

Journal of Back and Musculoskeletal Rehabilitation, Journal Year: 2023, Volume and Issue: 37(2), P. 253 - 268

Published: Nov. 7, 2023

BACKGROUND: Robot-assisted gait training (RAGT) has been reported to treat motor dysfunction in patients with Parkinson’s disease (PD) the last few years. However, benefits of RAGT for treating PD are still unclear. OBJECTIVES: To investigate efficacy patients. METHODS: We searched PubMed, Web Science, Cochrane Library, Embase, CNKI, Wanfang, Chinese Biomedical Literature Database (CBM), and VIP randomized controlled trials investigating improve from databases’ inception dates until September 1, 2022. The following outcome indexes were employed evaluate dysfunction: Berg Balance Scale (BBS), Activities-specific Confidence (ABC), 10-Meter Walk Test speed (10-MWT), speed, stride length, cadence Unified Parkinson Disease Rating Part III (UPDRS III), 6-Minute (6MWT), Timed Up Go test (TUG). meta-analysis was performed using proper randomeffect model or fixed-effect difference between control groups. Risk Bias Tool used included studies Grading Recommendations, Assessment, Development, Evaluations (GRADE) interpret certainty results. RESULTS: results consisted 17 comprising a total 670 participants. Six hundred seven included: 335 group group. This established that when compared group, robot-assisted improved BBS (MD: 2.80, 95%CI: 2.11–3.49, P< 0.00001), ABC score 7.30, 5.08–9.52, 10-MWT 0.06, 0.03–0.10, P= 0.0009), 3.67, 2.58–4.76, length 5.53, 3.64–7.42, 4.52, 0.94–8.10, 0.01), UPDRS -2.16, -2.48–-1.83, 6MWT 13.87, 11.92–15.82, 0.00001). did not significantly TUG result (MD =-0.56, 95% CI: -1.12–0.00, 0.05). No safety concerns adverse reactions among observed. CONCLUSION: Even though can balance function, walking performance demonstrated positive several studies, there is currently insufficient compelling evidence suggest it all aspects lower function.

Language: Английский

---

Automated Parkinson's Disease Detection: A Review of Techniques, Datasets, Modalities, and Open Challenges

International Journal on Smart Sensing and Intelligent Systems, Journal Year: 2024, Volume and Issue: 17(1)

Published: Jan. 1, 2024

Abstract Parkinson's disease (PsD) is a prevalent neurodegenerative malady, which keeps intensifying with age. It acquired by the progressive demise of dopaminergic neurons existing in substantia nigra pars compacta region human brain. In absence single accurate test, and due to dependency on doctors, intensive research being carried out automate early detection predict severity also. this study, detailed review various artificial intelligence (AI) models applied different datasets across modalities has been presented. The emotional (EI) modality, can be used for help maintaining comfortable lifestyle, identified. EI predominant, emerging technology that detect PsD at initial stages enhance socialization patients their attendants. Challenges possibilities assist bridging differences between fast-growing technologies meant actual implementation automated model are presented research. This highlights prominence using support vector machine (SVM) classifier achieving an accuracy about 99% many such as magnetic resonance imaging (MRI), speech, electroencephalogram (EEG). A 100% achieved EEG handwriting modality convolutional neural network (CNN) optimized crow search algorithm (OCSA), respectively. Also, 95% progression Bagged Tree, (ANN), SVM. maximum attained K-nearest Neighbors (KNN) Naïve Bayes classifiers signals EI. most widely dataset identified Progression Markers Initiative (PPMI) database.

Language: Английский

---

Assisted technology in Parkinson's disease gait: what's up?

Arquivos de Neuro-Psiquiatria, Journal Year: 2024, Volume and Issue: 82(06), P. 001 - 010

Published: Feb. 23, 2024

Abstract Background Gait disturbances are prevalent and debilitating symptoms, diminishing mobility quality of life for Parkinson's disease (PD) individuals. While traditional treatments offer partial relief, there is a growing interest in alternative interventions to address this challenge. Recently, remarkable surge assisted technology (AT) development was witnessed aid individuals with PD. Objective To explore the burgeoning landscape AT tailored alleviate PD-related gait impairments describe current research related such aim. Methods In review, we searched on PubMed papers published English (2018-2023). Additionally, abstract each study read ensure inclusion. Four researchers independently, including studies according our inclusion exclusion criteria. Results We included that met all identified key trends assistive parameters analysis These encompass wearable sensors, analysis, real-time feedback cueing techniques, virtual reality, robotics. Conclusion This review provides resource guiding future research, informing clinical decisions, fostering collaboration among researchers, clinicians, policymakers. By delineating rapidly evolving field's contours, it aims inspire further innovation, ultimately improving lives PD patients through more effective personalized interventions.

Language: Английский

---

Spatiotemporal gait parameter fluctuations in older adults affected by mild cognitive impairment: comparisons among three cognitive dual-task tests

BMC Geriatrics, Journal Year: 2023, Volume and Issue: 23(1)

Published: Sept. 27, 2023

Gait disorder is associated with cognitive functional impairment, and this disturbance more pronouncedly when performing additional tasks. Our study aimed to characterize gait disorders in mild impairment (MCI) under three dual tasks determine the association between performance function.A total of 260 participants were enrolled cross-sectional divided into MCI cognitively normal control. Spatiotemporal kinematic parameters (31 items) single task (serial 100-7, naming animals words recall) measured using a wearable sensor. Baseline characteristics two groups balanced propensity score matching. Important features filtered random forest method LASSO regression further described logistic analysis.After matching, 106 controls recruited. Top 5 4 ~ 6 important selected. Robust variables associating function temporal parameters. Participants exhibited decreased swing time terminal swing, increased mid stance variability stride length compared Subjects walked slower an extra task. In tasks, recall test pronounced impact on regularity, velocity, cost than other tests.Gait assessment conditions, particularly test, portable sensors could be useful as complementary strategy for early detection MCI.

Language: Английский

---

Integrating Abnormal Gait Detection with Activities of Daily Living Monitoring in Ambient Assisted Living: A 3D Vision Approach

Sensors, Journal Year: 2023, Volume and Issue: 24(1), P. 82 - 82

Published: Dec. 23, 2023

Gait analysis plays a crucial role in detecting and monitoring various neurological musculoskeletal disorders early. This paper presents comprehensive study of the automatic detection abnormal gait using 3D vision, with focus on non-invasive practical data acquisition methods suitable for everyday environments. We explore configurations, including multi-camera setups placed at different distances angles, as well performing daily activities directions. An integral component our involves combining living (ADLs), given paramount relevance this integration context Ambient Assisted Living. To achieve this, we investigate cutting-edge Deep Neural Network approaches, such Temporal Convolutional Network, Gated Recurrent Unit, Long Short-Term Memory Autoencoder. Additionally, scrutinize representation formats, Euclidean-based representations, angular adjacency matrices, rotation matrices. Our system’s performance evaluation leverages both publicly available datasets collected ourselves while accounting individual variations environmental factors. The results underscore effectiveness proposed configurations accurately classifying gait, thus shedding light optimal setup efficient collection.

Language: Английский

---

The inertial-based gait normalcy index of dual task cost during turning quantifies gait automaticity improvement in early-stage Parkinson’s rehabilitation

Journal of NeuroEngineering and Rehabilitation, Journal Year: 2024, Volume and Issue: 21(1)

Published: Sept. 19, 2024

Language: Английский

---