RESUMEN
BACKGROUND: Using proper transfer technique can help to reduce forces and prevent secondary injuries. However, current assessment tools rely on the ability to subjectively identify harmful movement patterns. OBJECTIVES: The purpose of the study was to determine the accuracy of using a low-cost markerless motion capture camera and machine learning methods to evaluate the quality of independent wheelchair sitting pivot transfers. We hypothesized that the algorithms would be able to discern proper (low risk) and improper (high risk) wheelchair transfer techniques in accordance with component items on the Transfer Assessment Instrument (TAI). METHODS: Transfer motions of 91 full-time wheelchair users were recorded and used to develop machine learning classifiers that could be used to discern proper from improper technique. The data were labeled using the TAI item scores. Eleven out of 18 TAI items were evaluated by the classifiers. Motion variables from the Kinect were inputted as the features. Random forests and k-nearest neighbors algorithms were chosen as the classifiers. Eighty percent of the data were used for model training and hyperparameter turning. The validation process was performed using 20% of the data as the test set. RESULTS: The area under the receiver operating characteristic curve of the test set for each item was over 0.79. After adjusting the decision threshold, the precisions of the models were over 0.87, and the model accuracies were over 71%. CONCLUSION: The results show promise for the objective assessment of the transfer technique using a low cost camera and machine learning classifiers.
Asunto(s)
Aprendizaje Automático , Sedestación , Traumatismos de la Médula Espinal/rehabilitación , Análisis y Desempeño de Tareas , Silla de Ruedas , Adulto , Fenómenos Biomecánicos , Femenino , Humanos , Masculino , Persona de Mediana EdadRESUMEN
OBJECTIVE: To evaluate the immediate effects of transfer training based on the Transfer Assessment Instrument (TAI) on the upper limb biomechanics during transfers. DESIGN: Pre-post intervention. SETTING: Biomechanics laboratory. PARTICIPANTS: Full-time manual wheelchair users (N=24) performed 5 transfers to a level height bench, while their natural transfer skills were scored using the TAI, and their biomechanical data were recorded. INTERVENTION: Participants with 2 or more component skill deficits were invited to return to receive personalized transfer training. MAIN OUTCOME MEASURES: TAI part 1 summary scores and biomechanical variables calculated at the shoulder, elbow, and wrist joints were compared before and immediately after transfer training. RESULTS: Sixteen of the 24 manual wheelchair users met the criteria for training, and 11 manual wheelchair users came back for the revisit. Their TAI part 1 summary scores improved from 6.31±.98 to 9.92±.25. They had significantly smaller elbow range of motion, shoulder resultant moment, and rates of rise of elbow and wrist resultant forces on their trailing side during transfers after training (P<.05). On the leading side, shoulder maximum internal rotation and elevation angles, and shoulder resultant moments and rates of rise of shoulder resultant force and moment decreased after training (P<.04). CONCLUSIONS: The TAI-based training showed short-term beneficial biomechanical effects on wheelchair users' upper limbs, such as better shoulder positioning and lower joint loadings. If the skills are practiced longer-term, they may help protect the upper limbs from developing pain and injuries.
Asunto(s)
Educación del Paciente como Asunto/métodos , Modalidades de Fisioterapia , Traumatismos de la Médula Espinal/rehabilitación , Extremidad Superior/fisiopatología , Silla de Ruedas , Adulto , Fenómenos Biomecánicos , Femenino , Humanos , Masculino , Persona de Mediana Edad , Hombro/fisiopatología , Lesiones del Hombro/prevención & controlRESUMEN
OBJECTIVES: To refine the Transfer Assessment Instrument (TAI 2.0), develop a training program for the TAI, and analyze the basic psychometric properties of the TAI 3.0, including reliability, standard error of measurement (SEM), minimal detectable change (MDC), and construct validity. DESIGN: Repeated measures. SETTING: A winter sports clinic for disabled veterans. PARTICIPANTS: Wheelchair users (N=41) who perform sitting-pivot or standing-pivot transfers. INTERVENTION: Not applicable. MAIN OUTCOME MEASURES: TAI version 3.0, intraclass correlation coefficients, SEMs, and MDCs for reliable measurement of raters' responses. Spearman correlation coefficient, 1-way analysis of variance, and independent t tests to evaluate construct validity. RESULTS: TAI 3.0 had acceptable to high levels of reliability (range, .74-.88). The SEMs for part 1, part 2, and final scores ranged from .45 to .75. The MDC was 1.5 points on the 10-point scale for the final score. There were weak correlations (ρ range, -.13 to .25; P>.11) between TAI final scores and subjects' characteristics (eg, sex, body mass index, age, type of disability, length of wheelchair use, grip and elbow strength, sitting balance). CONCLUSIONS: With comprehensive training, the refined TAI 3.0 yields high reliability among raters of different clinical backgrounds and experience. TAI 3.0 was unbiased toward certain physical characteristics that may influence transfer. TAI fills a void in the field by providing a quantitative measurement of transfers and a tool that can be used to detect problems and guide transfer training.