Search

Parkinson's Disease (PD) patients frequently transition between the 'ON' state, where medication is effective, and the 'OFF' state, affecting their quality of life. Monitoring these transitions is vital for personalized therapy. We introduced a framework based on Reinforcement Learning (RL) to detect transitions between medication states by learning from continuous movement data. Unlike traditional approaches that typically identify each state based on static data patterns, our approach focuses on understanding the dynamic patterns of change throughout the transitions, providing a more generalizable medication state monitoring method. We integrated a deep Long Short-Term Memory (LSTM) neural network and three one-class unsupervised classifiers to implement an RL-based adaptive classifier. We tested on two PD datasets: Dataset PD1 with 12 subjects (14-minute average recording) and Dataset PD2 with seven subjects (120-minute average recording). Data from wrist and ankle wearables captured transitions during 2 to 4-hour daily activities. The algorithm demonstrated its effectiveness in detecting medication states, achieving an average weighted F1-score of 82.94% when trained and tested on Dataset PD1. It performed well when trained on Dataset PD1 and tested on Dataset PD2, with a weighted F1-score of 76.67%. It surpassed other models, was resilient to severe PD symptoms, and performed well with imbalanced data. Notably, prior work has not addressed the generalizability from one dataset to another, essential for real-world applications with varied sensors. Our innovative framework revolutionizes PD monitoring, setting the stage for advanced therapeutic methods and greatly enhancing the life quality of PD patients.

The Unified Parkinson's Disease Rating Scale (UPDRS) is used to recognize patients with Parkinson's disease (PD) and rate its severity. The rating is crucial for disease progression monitoring and treatment adjustment. This study aims to advance the capabilities of PD management by developing an innovative framework that integrates deep learning with wearable sensor technology to enhance the precision of UPDRS assessments. We introduce a series of deep learning models to estimate UPDRS Part III scores, utilizing motion data from wearable sensors. Our approach leverages a novel Multi-shared-task Self-supervised Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) framework that processes raw gyroscope signals and their spectrogram representations. This technique aims to refine the estimation accuracy of PD severity during naturalistic human activities. Utilizing 526 min of data from 24 PD patients engaged in everyday activities, our methodology demonstrates a strong correlation of 0.89 between estimated and clinically assessed UPDRS-III scores. This model outperforms the benchmark set by single and multichannel CNN, LSTM, and CNN-LSTM models and establishes a new standard in UPDRS-III score estimation for free-body movements compared to recent state-of-the-art methods. These results signify a substantial step forward in bioengineering applications for PD monitoring, providing a robust framework for reliable and continuous assessment of PD symptoms in daily living settings.

This study explores the potential of smartphones to objectively assess balance, which is crucial for the elderly and individuals recovering from various medical conditions. We propose an innovative methodology to estimate the Modified Clinical Test of Sensory Interaction on Balance (m-CTSIB) scores using the accelerometer sensor of a smartphone coupled with machine learning techniques. Our dataset consists of 28 participants, aged 21 to 88 years. Notably, the XGBOOST algorithm demonstrates a strong correlation (0.92) with the ground truth balance scores. These ground truth scores are obtained using a force plate system collected simultaneously with the smartphone data, ensuring precise and reliable comparisons. This methodology offers an objective, accessible, and convenient means for balance assessment, greatly facilitating at-home monitoring and enhancing the potential for remote health monitoring. Our findings underscore the method's reliability and potential impact on telemedicine and patient care, offering notable improvements in the quality of life.

The Unified Parkinson's Disease Rating Scale (UP-DRS) is used to recognize patients with Parkinson's disease (PD) and rate its severity in clinical settings. Machine learning and wearables can reduce the need for clinical examinations and provide a reliable estimation of the severity of PD at home. This work introduces a multi-channel convolutional neural network to estimate UPDRS part III from motion data recorded by two wearable sensors, considering the gyroscope signals and their spectrogram representations. A novel shared-task self-supervised learning is then employed to leverage the knowledge extracted from the signal to improve the estimation during patients' free-body movements. We utilize 24 PD subjects' data performing daily activities. The estimated UPDRS-III showed an improved correlation with the clinical examinations from 0.67 to 0.81, reducing the mean absolute error from 7.75 to 6.96. Our investigation demonstrates the potential of our approach in providing a reliable estimation of PD severity scores during subjects' daily routines. It can also provide comprehensive information to help physicians manage the disease and adjust the dose and interval of PD medications.

This study marks the first endeavor to utilize wearable technology combined with machine learning to objectively assess the Modified Clinical Test of Sensory Interaction on Balance (m-CTSIB). We focus on developing an affordable, easily accessible method for balance assessment, critical for adults at risk of falls and cognitive decline. Our novel approach uses a single inertial measurement unit sensor (APDM, INC.) to gather lumbar accelerometer and gyroscope data. This data is accompanied by ground truth scores obtained from m-CTSIB tests on a force plate (Falltrak II, MedTrak VNG, Inc.) from 34 participants aged 21 to 88. Using XGBOOST, we achieve a remarkable 0.94 correlation using accelerometer data and 0.90 with gyroscope data in the test dataset, demonstrating a strong correlation with actual scores in a subject-wise leave-one-out cross-validation. Offering objectivity, affordability, and potential for remote monitoring, our innovative approach holds promise for enhancing the diagnosis and management of balance disorders in adults, thereby improving their quality of life and independence.

Smartwatches play an increasingly vital role in health monitoring, offering practical solutions for patients to manage their conditions in daily life. This is particularly significant for Parkinson's Disease (PD) patients, as tremors, a notable symptom, significantly impact their quality of life. Our research addresses unanswered questions about the optimal configurations for these devices, considering factors such as the placement of the smartwatch, the type of sensor used (accelerometer vs. gyroscope), and the sampling frequency. Employing a gradient tree-boosting approach, we analyze data from the wrist and ankle sensors of 24 PD subjects engaged in daily living activities. The findings highlight a cost-effective, practical, and resource-friendly configuration: utilizing accelerometer wrist data sampled at 32 Hz demonstrates a robust correlation between the estimations of total tremors (0.82) and rest tremors (0.84) compared to their clinical ground-truth values. This underlines the significant potential of employing smartwatches in a natural environment for clinicians to monitor their patients efficiently, minimizing the necessity for costly, inconvenient in-clinic assessments.

Accurate balance assessment is important in healthcare for identifying and managing conditions affecting stability and coordination. It plays a key role in preventing falls, understanding movement disorders, and designing appropriate therapeutic interventions across various age groups and medical conditions. However, traditional balance assessment methods often suffer from subjectivity, lack of comprehensive balance assessments and remote assessment capabilities, and reliance on specialized equipment and expert analysis. In response to these challenges, our study introduces an innovative approach for estimating scores on the Modified Clinical Test of Sensory Interaction on Balance (m-CTSIB). Utilizing wearable sensors and advanced machine learning algorithms, we offer an objective, accessible, and efficient method for balance assessment. We collected comprehensive movement data from 34 participants under four different sensory conditions using an array of inertial measurement unit (IMU) sensors coupled with a specialized system to evaluate ground truth m-CTSIB balance scores for our analysis. This data was then preprocessed, and an extensive array of features was extracted for analysis. To estimate the m-CTSIB scores, we applied Multiple Linear Regression (MLR), Support Vector Regression (SVR), and XGBOOST algorithms. Our subject-wise Leave-One-Out and 5-Fold cross-validation analysis demonstrated high accuracy and a strong correlation with ground truth balance scores, validating the effectiveness and reliability of our approach. Key insights were gained regarding the significance of specific movements, feature selection, and sensor placement in balance estimation. Notably, the XGBOOST model, utilizing the lumbar sensor data, achieved outstanding results in both methods, with Leave-One-Out cross-validation showing a correlation of 0.96 and a Mean Absolute Error (MAE) of 0.23 and 5-fold cross-validation showing comparable results with a correlation of 0.92 and an MAE of 0.23, confirming the model's consistent performance. This finding underlines the potential of our method to revolutionize balance assessment practices, particularly in settings where traditional methods are impractical or inaccessible., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (© 2024 Nassajpour, Shuqair, Rosenfeld, Tolea, Galvin and Ghoraani.)