Feasibility of artificial intelligence assisted quantitative muscle ultrasound in carpal tunnel syndrome

BMC Musculoskeletal Disorders

Research

Jung Sub Sim¹  Sunwoo Kim¹  Dongik Shin¹  Seungjun Baek¹  Joon Shik Yoon²  Sun Woong Kim²  Jae Hyeong Choi² 

[1] Department of Computer Science and Engineering, Korea University, 145, Anam-ro, 02841, Seoul, Republic of Korea;Department of Physical and Rehabilitation Medicine, Korea University Guro Hospital, 148 Gurodong-ro, Guro-gu, 08308, Seoul, Republic of Korea;

关键词: Muscle ultrasound;    Quantitative ultrasound;    Carpal tunnel syndrome;    Machine learning;    Artificial intelligence;   

DOI  :  10.1186/s12891-023-06623-3

 received in 2023-01-16, accepted in 2023-06-09,  发布年份 2023

来源: Springer

【 摘 要 】

BackgroundIn case of focal neuropathy, the muscle fibers innervated by the corresponding nerves are replaced with fat or fibrous tissue due to denervation, which results in increased echo intensity (EI) on ultrasonography. EI analysis can be conducted quantitatively using gray scale analysis. Mean value of pixel brightness of muscle image defined as EI. However, the accuracy achieved by using this parameter alone to differentiate between normal and abnormal muscles is limited. Recently, attempts have been made to increase the accuracy using artificial intelligence (AI) in the analysis of muscle ultrasound images. CTS is the most common disease among focal neuropathy. In this study, we aimed to verify the utility of AI assisted quantitative analysis of muscle ultrasound in CTS.MethodsThis is retrospective study that used data from adult who underwent ultrasonographic examination of hand muscles. The patient with CTS confirmed by electromyography and subjects without CTS were included. Ultrasound images of the unaffected hands of patients or subjects without CTS were used as controls. Ultrasonography was performed by one physician in same sonographic settings. Both conventional quantitative grayscale analysis and machine learning (ML) analysis were performed for comparison.ResultsA total of 47 hands with CTS and 27 control hands were analyzed. On conventional quantitative analysis, mean EI ratio (i.e. mean thenar EI/mean hypothenar EI ratio) were significantly higher in the patient group than in the control group, and the AUC was 0.76 in ROC analysis. In the analysis using machine learning, the AUC was the highest for the linear support vector classifier (AUC = 0.86). When recursive feature elimination was applied to the classifier, the AUC value improved to 0.89.ConclusionThis study showed a significant increase in diagnostic accuracy when AI was used for quantitative analysis of muscle ultrasonography. If an analysis protocol using machine learning can be established and mounted on an ultrasound machine, a noninvasive and non-time-consuming muscle ultrasound examination can be conducted as an ancillary tool for diagnosis.

【 授权许可】

CC BY   
© The Author(s) 2023

【 图 表 】

Fig. 6

Fig. 6

Fig. 1

Fig. 22

12936_2023_4634_Article_IEq2.gif

13690_2023_1130_Article_IEq59.gif