Most agricultural workers are exposed to musculoskeletal disorders due to the characteristics of agricultural work performed manually. As observational methods to prevent musculoskeletal disorders, a cube method has been proposed that considers the risk factors of posture, time and
force workload simultaneously. However, force workload could evaluate using the weight of an object or qualitative measurement to prevent interfering with a worker’s occupation. The purpose of this study is to propose a novel method for evaluating quantitatively the risk factor of force
in agricultural field using insole system and artificial neural network model. Agricultural simulated experiments were performed on ten healthy adult males and six observers were recruited to evaluate the risk factors of force for the experiments. The model was constructed using the signals
measured in the insole system and the consensus among observers about evaluation results. To verify the performance of the model, the performance measurement was calculated using 10-fold cross-validation. The results of the proposed method are compared with those of the observers to verify
reproducibility and usefulness. The model showed more than 97% prediction accuracy in all risk levels, and the proposed method showed 1.59%, 0.99 and 0.98 in the coefficient of variation, proportion agreement index, Cohen’s kappa coefficient, and high reproducibility and usefulness when
compared with the observers’ evaluation. The method of quantitatively evaluating the risk factor of force proposed in this study is possible to be applied to various agricultural works using observational methods.
An Integrative Second-Order Adaptive Network Model for the Effect of L. Reuteri Probiotics in the Gut on ASD Symptoms
