肌肉骨骼疾患风险评估方法趋势研究

目的肌肉骨骼疾患是与工作姿势有关的主要职业病,不仅影响工人的健康,而且对经济造成巨大的损失。而不良的工作姿势是引发肌肉骨骼疾患的主要原因。因此评估工作姿势所造成的肌肉骨骼疾患并采取科学的纠正措施逐渐成为相应领域研究的热点问题。基于现有对肌肉骨骼疾患风险评估方法的研究现状和未来趋势进行分析与展望。方法通过对肌肉骨骼疾患、人因风险评估、人体姿态识别等核心概念的相关文献进行梳理和归纳,论述了肌肉骨骼疾患风险评估的主要方法,并重点分析了图像识别技术在肌肉骨骼疾患风险评估中的应用,结合人体骨架与神经网络算法模型对作业姿势进行识别,探讨了人工智能环境下,基于图像识别的评估方法中待解决的难点问题,对未来可能发展趋势进行预测。结论将肌肉骨骼风险评估方法总结为三大类,并分析其在现场评估应用过程中的优缺点;结合图像识别技术的发展,对肌肉骨骼风险评估提出了展望,即智能化自动化评估、多评估方法融合、多通道特征识别。

Trends of Risk Assessment Methods for Musculoskeletal Diseases

Musculoskeletal diseases are the main occupational diseases related to working postures, which not only affect the health of workers, but also cause huge losses to the economy. Poor working posture is the main cause of musculoskeletal diseases. Therefore, how to assess the musculoskeletal diseases caused by working posture and take scientific corrective measures has gradually become a hot issue in the corresponding field of research. Based on the existing research, the research status and trends of musculoskeletal disease risk assessment methods have been analyzed and prospected. By sorting out and summarizing the relevant literatures on the core concepts of musculoskeletal diseases, human-caused risk assessment, human posture recognition, etc., the main methods of musculoskeletal disease risk assessment were discussed and the application of image recognition technology in the risk assessment of muscu-loskeletal diseases was analyzed. Combined with human skeleton and neural network algorithm model, the working posture was identified, and the problems in the assessment method based on image recognition in AI environment were discussed, and the possible development trend in the future was predicted. The musculoskeletal risk assessment methods are summarized into three major categories, and their advantages and disadvantages in site assessment are analyzed. Combined with the development of image recognition technology, the prospect of musculoskeletal risk assessment is put forward, namely intelligent automated assessment, fusion of multiple assessment methods and multi-channel feature recognition.