Machine Learning for Rapid Estimation of Lower Extremity Muscle and Joint Loading during Activities of Daily Living

The purpose of this study was to examine the test-retest reliability of hand-held dynamometer measurements in patients with dementia and determine if predictions about a patient’s ability to perform daily activities can be made from knee extension strength measurements. This study was composed of two rounds of data collection. Sixty patients with dementia were enrolled in the first round to assess the reliability of hand-held dynamometer measurements, and 54 patients with dementia were enrolled in the second round for predicting their ability to perform daily activities. Knee extensor strength was measured twice, separated by a three minute interval, with hand-held dynamometer. The authors also assessed daily activities related to the patient’s lower extremities, including dressing the lower body, using the toile, transferring to the bed/toilet/shower, and walking. Lower extremity activities of the Functional Independence Measure were assessed by the nursing home caregiver that had the most regular contact with each subject. When the Functional Independence Measure score of each lower extremity function was =6 points, the subject was considered to be independent. The intraclass correlation coefficient was 0.97. Bland-Altman plots showed the 95% difference value to be within 2 SDs of the mean. The curves of negative and positive predictive values revealed the following threshold levels: 0.8 Nm/kg was the best predictor for dressing the lower body and using the toilet; 1.2 Nm/kg was the best predictor for transferring to the bed/toilet/shower; and 0.6 Nm/kg was the best predictor for gait performance. Strength measurements taken with a hand-held dynamometer were reliable in patients with dementia, and normalized knee extensor strength was found to be a predictor of the ability to perform activities of daily living.

Download Full-text

Streamlining the KOOS Activities of Daily Living Subscale Using Machine Learning

Orthopaedic Journal of Sports Medicine ◽

10.1177/2325967120910447 ◽

2020 ◽

Vol 8 (3) ◽

pp. 232596712091044 ◽

Cited By ~ 1

Author(s):

Ashim Gupta ◽

Ajish S.R. Potty ◽

Deepak Ganta ◽

R. Justin Mistovich ◽

Sreeram Penna ◽

...

Keyword(s):

Machine Learning ◽

Activities Of Daily Living ◽

Functional Outcome ◽

Acl Reconstruction ◽

Daily Living ◽

High Accuracy ◽

Quality Data ◽

Specific Patient ◽

Patient Reported ◽

Essential Questions

Background: Functional outcome scores provide valuable data, yet they can be burdensome to patients and require significant resources to administer. The Knee injury and Osteoarthritis Outcome Score (KOOS) is a knee-specific patient-reported outcome measure (PROM) and is validated for anterior cruciate ligament (ACL) reconstruction outcomes. The KOOS requires 42 questions in 5 subscales. We utilized a machine learning (ML) algorithm to determine whether the number of questions and the resultant burden to complete the survey can be lowered in a subset (activities of daily living; ADL) of KOOS, yet still provide identical data. Hypothesis: Fewer questions than the 17 currently provided are actually needed to predict KOOS ADL subscale scores with high accuracy. Study Design: Cohort study (diagnosis); Level of evidence, 2. Methods: Pre- and postoperative patient-reported KOOS ADL scores were obtained from the Surgical Outcome System (SOS) data registry for patients who had ACL reconstruction. Categorical Boosting (CatBoost) ML models were built to analyze each question and its value in predicting the patient’s actual functional outcome (ie, KOOS ADL score). A streamlined set of minimal essential questions were then identified. Results: The SOS registry contained 6185 patients who underwent ACL reconstruction. A total of 2525 patients between the age of 16 and 50 years had completed KOOS ADL scores presurgically and 3 months postoperatively. The data set consisted of 51.84% male patients and 48.16% female patients, with a mean age of 29 years. The CatBoost model predicted KOOS ADL scores with high accuracy when only 6 questions were asked ( R2 = 0.95), similar to when all 17 questions of the subscale were asked ( R2 = 0.99). Conclusion: ML algorithms successfully identified the essential questions in the KOOS ADL questionnaire. Only 35% (6/17) of KOOS ADL questions (descending stairs, ascending stairs, standing, walking on flat surface, putting on socks/stockings, and getting on/off toilet) are needed to predict KOOS ADL scores with high accuracy after ACL reconstruction. ML can be utilized successfully to streamline the burden of patient data collection. This, in turn, can potentially lead to improved patient reporting, increased compliance, and increased utilization of PROMs while still providing quality data.

Download Full-text

Effects of Massage in the Lower Extremity on Range of Motion and Activities of Daily Living in Spasticity Cerebral Palsy with Children

10.17817/2015.07.09.246 ◽

2015 ◽

Author(s):

Roh Lyun

Keyword(s):

Cerebral Palsy ◽

Lower Extremity ◽

Activities Of Daily Living ◽

Range Of Motion ◽

Daily Living

Download Full-text

Effectiveness of a Novel Contralaterally Controlled Neuromuscular Electrical Stimulation for Restoring Lower Limb Motor Performance and Activities of Daily Living in Stroke Survivors: A Randomized Controlled Trial

Neural Plasticity ◽

10.1155/2022/5771634 ◽

2022 ◽

Vol 2022 ◽

pp. 1-9

Author(s):

Ying Shen ◽

Lan Chen ◽

Li Zhang ◽

Shugang Hu ◽

Bin Su ◽

...

Keyword(s):

Electrical Stimulation ◽

Lower Extremity ◽

Activities Of Daily Living ◽

Motor Function ◽

Daily Living ◽

Neuromuscular Electrical Stimulation ◽

Lower Extremity Function ◽

Stroke Patients ◽

Extremity Function ◽

Subacute Stroke

Background. Contralaterally controlled neuromuscular electrical stimulation (CCNMES) is a novel electrical stimulation treatment for stroke; however, reports on the efficacy of CCNMES on lower extremity function after stroke are scarce. Objective. To compare the effects of CCNMES versus NMES on lower extremity function and activities of daily living (ADL) in subacute stroke patients. Methods. Forty-four patients with a history of subacute stroke were randomly assigned to a CCNMES group and a NMES group ( n = 22 per group). Twenty-one patients in each group completed the study per protocol, with one subject lost in follow-up in each group. The CCNMES group received CCNMES to the tibialis anterior (TA) and the peroneus longus and brevis muscles to induce ankle dorsiflexion motion, whereas the NMES group received NMES. The stimulus current was a biphasic waveform with a pulse duration of 200 μs and a frequency of 60 Hz. Patients in both groups underwent five 15 min sessions of electrical stimulation per week for three weeks. Indicators of motor function and ADL were measured pre- and posttreatment, including the Fugl–Meyer assessment of the lower extremity (FMA-LE) and modified Barthel index (MBI). Surface electromyography (sEMG) assessments included average electromyography (aEMG), integrated electromyography (iEMG), and root mean square (RMS) of the paretic TA muscle. Results. Values for the FMA-LE, MBI, aEMG, iEMG, and RMS of the affected TA muscle were significantly increased in both groups after treatment ( p < 0.01 ). Patients in the CCNMES group showed significant improvements in all the measurements compared with the NMES group after treatment. Within-group differences in all post- and pretreatment indicators were significantly greater in the CCNMES group than in the NMES group ( p < 0.05 ). Conclusion. CCNMES improved motor function and ADL ability to a greater extent than the conventional NMES in subacute stroke patients.

Download Full-text

Detecting hand washing activity among activities of daily living and classification of WHO hand washing techniques using wearable devices and machine learning algorithms

Healthcare Technology Letters ◽

10.1049/htl2.12018 ◽

2021 ◽

Author(s):

Yiyuan Zhang ◽

Tianwei Xue ◽

Zhenjie Liu ◽

Wei Chen ◽

Bart Vanrumste

Keyword(s):

Machine Learning ◽

Activities Of Daily Living ◽

Daily Living ◽

Learning Algorithms ◽

Wearable Devices ◽

Machine Learning Algorithms ◽

Hand Washing

Download Full-text

Range of Motion: AMA Guides, Sixth Edition

Guides Newsletter ◽

10.1001/amaguidesnewsletters.2015.mayjun01 ◽

2015 ◽

Vol 20 (3) ◽

pp. 3-5

Author(s):

James B. Talmage ◽

Jay Blaisdell

Keyword(s):

Lower Extremity ◽

Physical Examination ◽

Activities Of Daily Living ◽

Range Of Motion ◽

Daily Living ◽

Lower Extremities ◽

Upper Extremities ◽

Sixth Edition

Abstract In the AMA Guides to the Evaluation of Permanent Impairment (AMA Guides), Sixth Edition, range of motion (ROM) is used to calculate the physical examination modifier when the diagnosis-based impairment (DBI) method is used, ie, Table 15-8, Physical Examination Adjustment: Upper Extremities, and Table 16-7, Physical Examination Adjustment: Lower Extremities. The DBI method is preferable for calculating upper and lower extremity ratings, but in instances specified by the regional grid, the evaluator may choose to use the ROM method, typically because it yields a higher impairment rating. The article outlines the steps for measuring ROM. Invalid results on the day of testing are declared after three consecutive efforts if the three measurements for a given plane of motion vary by more than 10 degrees from the average (mean) of these three measurements. The process can be tried again another day, or the DBI method can be used. All ROM measurements should be rounded to the nearest number ending in zero, and a figure provides examples of suitable annotations. Adjustments for functional history can be made if ROM is the only method used for rating, if results are deemed reliable and consistent with results from an activities of daily living questionnaire or other valid functional report, and if the current ROM impairment does not adequately capture the full impairment.

Download Full-text