Dataset Information

Automatic Grading of Stroke Symptoms for Rapid Assessment Using Optimized Machine Learning and 4-Limb Kinematics: Clinical Validation Study.

ABSTRACT: BACKGROUND:Subtle abnormal motor signs are indications of serious neurological diseases. Although neurological deficits require fast initiation of treatment in a restricted time, it is difficult for nonspecialists to detect and objectively assess the symptoms. In the clinical environment, diagnoses and decisions are based on clinical grading methods, including the National Institutes of Health Stroke Scale (NIHSS) score or the Medical Research Council (MRC) score, which have been used to measure motor weakness. Objective grading in various environments is necessitated for consistent agreement among patients, caregivers, paramedics, and medical staff to facilitate rapid diagnoses and dispatches to appropriate medical centers. OBJECTIVE:In this study, we aimed to develop an autonomous grading system for stroke patients. We investigated the feasibility of our new system to assess motor weakness and grade NIHSS and MRC scores of 4 limbs, similar to the clinical examinations performed by medical staff. METHODS:We implemented an automatic grading system composed of a measuring unit with wearable sensors and a grading unit with optimized machine learning. Inertial sensors were attached to measure subtle weaknesses caused by paralysis of upper and lower limbs. We collected 60 instances of data with kinematic features of motor disorders from neurological examination and demographic information of stroke patients with NIHSS 0 or 1 and MRC 7, 8, or 9 grades in a stroke unit. Training data with 240 instances were generated using a synthetic minority oversampling technique to complement the imbalanced number of data between classes and low number of training data. We trained 2 representative machine learning algorithms, an ensemble and a support vector machine (SVM), to implement auto-NIHSS and auto-MRC grading. The optimized algorithms performed a 5-fold cross-validation and were searched by Bayes optimization in 30 trials. The trained model was tested with the 60 original hold-out instances for performance evaluation in accuracy, sensitivity, specificity, and area under the receiver operating characteristics curve (AUC). RESULTS:The proposed system can grade NIHSS scores with an accuracy of 83.3% and an AUC of 0.912 using an optimized ensemble algorithm, and it can grade with an accuracy of 80.0% and an AUC of 0.860 using an optimized SVM algorithm. The auto-MRC grading achieved an accuracy of 76.7% and a mean AUC of 0.870 in SVM classification and an accuracy of 78.3% and a mean AUC of 0.877 in ensemble classification. CONCLUSIONS:The automatic grading system quantifies proximal weakness in real time and assesses symptoms through automatic grading. The pilot outcomes demonstrated the feasibility of remote monitoring of motor weakness caused by stroke. The system can facilitate consistent grading with instant assessment and expedite dispatches to appropriate hospitals and treatment initiation by sharing auto-MRC and auto-NIHSS scores between prehospital and hospital responses as an objective observation.

SUBMITTER: Park E

PROVIDER: S-EPMC7527905 | biostudies-literature | 2020 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Automatic Grading of Stroke Symptoms for Rapid Assessment Using Optimized Machine Learning and 4-Limb Kinematics: Clinical Validation Study.

Park Eunjeong E Lee Kijeong K Han Taehwa T Nam Hyo Suk HS

Journal of medical Internet research 20200916 9

<h4>Background</h4>Subtle abnormal motor signs are indications of serious neurological diseases. Although neurological deficits require fast initiation of treatment in a restricted time, it is difficult for nonspecialists to detect and objectively assess the symptoms. In the clinical environment, diagnoses and decisions are based on clinical grading methods, including the National Institutes of Health Stroke Scale (NIHSS) score or the Medical Research Council (MRC) score, which have been used to ...[more]

PMID: 32936079

Similar Datasets

Project description:BackgroundUpper-limb impairments in stroke patients are usually measured in clinical setting using standard clinical assessment. In addition, kinematic analysis using opto-electronic systems has been used in the laboratory setting to map arm recovery. Such kinematic measurements cannot capture the actual function of the upper extremity in daily life. The aim of this study is to longitudinally explore the complementarity of post-stroke upper-limb recovery measured by standard clinical assessments and daily-life recorded kinematics.MethodsThe study was designed as an observational, single-group study to evaluate rehabilitation progress in a clinical and home environment, with a full-body sensor system in stroke patients. Kinematic data were recorded with a full-body motion capture suit during clinical assessment and self-directed activities of daily living. The measurements were performed at three time points for 3 h: (1) 2 weeks before discharge of the rehabilitation clinic, (2) right after discharge, and (3) 4 weeks after discharge. The kinematic analysis of reaching movements uses the position and orientation of each body segment to derive the joint angles. Newly developed metrics for classifying activity and quality of upper extremity movement were applied.ResultsThe data of four stroke patients (three mildly impaired, one sever impaired) were included in this study. The arm motor function assessment improved during the inpatient rehabilitation, but declined in the first 4 weeks after discharge. A change in the data (kinematics and new metrics) from the daily-life recording was seen in in all patients. Despite this worsening patients increased the number of reaches they performed during daily life in their home environment.ConclusionIt is feasible to measure arm kinematics using Inertial Measurement Unit sensors during daily life in stroke patients at the different stages of rehabilitation. Our results from the daily-life recordings complemented the data from the clinical assessments and illustrate the potential to identify stroke patient characteristics, based on kinematics, reaching counts, and work area.Clinical trial registrationhttps://clinicaltrials.gov, identifier NCT02118363.

Project description:Background:Kinematic analysis using virtual reality (VR) environment provides quantitative assessment of upper limb movements. This technique has rarely been used in evaluating motor function in stroke despite its availability in stroke rehabilitation. Objective:To determine the discriminative validity of VR-based kinematics during target-to-target pointing task in individuals with mild or moderate arm impairment following stroke and in healthy controls. Methods:Sixty-seven participants with moderate (32-57 points) or mild (58-65 points) stroke impairment as assessed with Fugl-Meyer Assessment for Upper Extremity were included from the Stroke Arm Longitudinal study at the University of Gothenburg-SALGOT cohort of non-selected individuals within the first year of stroke. The stroke groups and 43 healthy controls performed the target-to-target pointing task, where 32 circular targets appear one after the other and disappear when pointed at by the haptic handheld stylus in a three-dimensional VR environment. The kinematic parameters captured by the stylus included movement time, velocities, and smoothness of movement. Results:The movement time, mean velocity, and peak velocity were discriminative between groups with moderate and mild stroke impairment and healthy controls. The movement time was longer and mean and peak velocity were lower for individuals with stroke. The number of velocity peaks, representing smoothness, was also discriminative and significantly higher in both stroke groups (mild, moderate) compared to controls. Movement trajectories in stroke more frequently showed clustering (spider's web) close to the target indicating deficits in movement precision. Conclusion:The target-to-target pointing task can provide valuable and specific information about sensorimotor impairment of the upper limb following stroke that might not be captured using traditional clinical scale. Trial registration details:The trial was registered with register number NCT01115348 at clinicaltrials.gov, on May 4, 2010. URL: https://clinicaltrials.gov/ct2/show/NCT01115348.

Dataset Information

Automatic Grading of Stroke Symptoms for Rapid Assessment Using Optimized Machine Learning and 4-Limb Kinematics: Clinical Validation Study.

Publications

Automatic Grading of Stroke Symptoms for Rapid Assessment Using Optimized Machine Learning and 4-Limb Kinematics: Clinical Validation Study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets