| Summary: | Gripping is an important physical activity in daily routine. The capability of muscular force during gripping can be evaluated in terms of Hand grip Strength (HGS) and Hand grip Endurance (HGE). There are two types of movements that are associated with HGE which are dynamic or repetitive (HGEd) and static (HGE5) movements. In the literature, there are many studies which have been performed to investigate the relationship between demographics and hand anthropometric dimensions factors with HGS. These factors have been used as predictive factor for rehabilitation and recovery. However there is lack of studies showing the relationship combined of demographics and hand anthropometric dimensions to HGE which are important factors in hand rehabilitation and recovery. The aim of this project is to develop predictive model of young female HGS and HGE based on the demographic and hand anthropometric collected. Thus, the specific objectives of this study are; (1) to develop a optimal grip size electronic hand grip strength measuring system that records and analyze the HGS and HGE time series signals, (2) to determine the correlation between demographic and hand anthropometric dimensions, and the HGS as well as HGE of young Malaysian female, and (3) to develop an intelligent predictive model of HGS and HGE. There are three assessments in evaluating the HGS, HOEd and HGEs: single- repetition, 20-repetition and 30-
seconds static hold. In addition 6 demographics and 9 hand anthropometrics data are recorded from each volunteer in order to investigate the correlation
between HGS, HGEd and HGEs and these data. By using all the associated data, the predictive model of HGS, HOEd and HGEs are developed using Adaptive Neuro Fuzzy Inference System (ANFIS) model. In this study 45 females of the age group 18 to 30 years were recruited. The assessment of grip strength and endurance was measured using the fabricated hand grip measuring device and followed the America Society of Hand Therapy (ASHT) protocols of seating to maintain the consistency of each volunteer's measurement. By comparing with similar study performed on western population, the results show that the female HGS in this study is much higher probably due to optimal grip size of the fabricated measuring device.
Meanwhile for HGEd and HGEs, these measurements are lower and it is found that the hand dominant was significantly stronger than non-hand dominant for HGS, HGEd and HGEs. In addition the HGS was correlated with weight, Body Mass Index (BMI), hand breadth across thumb, wrist thickness and wrist circumference. Meanwhile HGEd and HGEs were correlated with age and occupation but not correlated with any of the hand anthropometric dimensions. Non-parametric predictive model based on ANFIS is used to develop the predictive HGS and HE model. In developing predictive ANFIS modeling, the input selection was executed and the most significant inputs with respect to
HGS, HGEd and HGEs for both hands are obtained. In ANFIS model, there is small discrepancy between actual and predicted average output for training and checking datasets. Nevertheless, this study has shown that ANFIS can be potentially used as an effective predictive model with larger dataset.
|
|---|