Ultrasound‐based radiomics in the diagnosis of carpal tunnel syndrome: The influence of regions of interest delineation method on mode

Abstract

Background In the recent years, artificial intelligence (AI) algorithms have been used to accurately diagnose musculoskeletal diseases. However, it is not known whether the particular regions of interest (ROI) delineation method would affect the performance of the AI algorithm. Purpose The purpose of this study was to investigate the influence of ROI delineation methods on model performance and observer consistency. Methods In this retrospective analysis, ultrasound (US) measures of median nerves affected with carpal tunnel syndrome (CTS) were compared to median nerves in a control group without CTS. Two methods were used for delineation of the ROI: (1) the ROI along the hyperechoic medial edge of the median nerve but not including the epineurium (MN) (ROI1); and (2) the ROI including the hyperechoic epineurium (ROI2), respectively. The intra group correlation coefficient (ICC) was used to compare the observer consistency of ROI features (i.e. the corresponding radiomics parameters). Parameters α1 and α2 were obtained based on the ICC of ROI1 features and ROI2 features. The ROC analysis was used to determine the area under the curve (AUC) and evaluate the performance of the radiologists and network. In addition, four indices, namely sensitivity, specificity, positive prediction and negative prediction were analyzed too. Results A total of 136 wrists of 77 CTS group and 136 wrists of 74 control group were included in the study. Control group was matched to CTS group according to the age and sex. The observer consistency of ROI features delineated by the two schemes was different, and the consistency of ROI1 features was higher (α1 ˃ α2). The intra‐observer consistency was higher than the inter‐observer consistency regardless of the scheme, and the intra‐observer consistency was higher when chose scheme one. The performances of models based on the two ROI features were different, although the AUC of each model was greater than 0.8.The model performed better when the MN epineurium was included in the ROI. Among five artificial intelligence algorithms, the Forest models (model1 achieved an AUC of 0.921 in training datasets and 0.830 in testing datasets; model2 achieved an AUC of 0.967 in training datasets and 0.872 in testing datasets.) obtained the highest performance, followed by the support vector machine (SVM) models and the Logistic models. The performances of the models were significantly better than the inexperienced radiologist (Dr. B. Z. achieved an AUC of 0.702). Conclusion Different ROI delineation methods may affect the performance of the model and the consistency of observers. Model performance was better when the ROI contained the MN epineurium, and observer consistency was higher when the ROI was delineated along the hyperechoic medial border of the MN.

Ultrasound‐based measures in carpal tunnel syndrome: What to measure?

“…1 The inclusion of the epineurium in the ROI as reported by Lv et al is supported by studies of the pathophysiology of CTS. [1][2][3][4][5][6][7] The pathology of CTS is characterized by entrapment of the MN at the transverse carpal ligament, resulting markedly increased pressure in the carpal tunnel as well as decreased neural gliding (tethering) with motion causing traction and shear forces on the MN. [2][3][4][5][6][7] As the pressure in the carpal tunnel increases, there is also interference with fluid flow in the small blood and lymphatic vessels that pass through the epineurium to supply the MN (Figure 2A).…”

“…Lv et al have demonstrated that the epineurium should be included in the ROI for more consistent CTS measures 1 . Once reason that including the epineurium in the ROI provides more consistent results is likely due to the improved conspicuity of the epineurium‐extraneural connective tissue interface compared to more vague interfaces of the epineurium‐perineurium‐interfasicular interfaces (Figure 1).…”

“…In the current issue of the Journal of Clinical Ultrasound , Lv et al report on radiomic evaluation of ultrasound images of the median nerve (MN) in carpal tunnel syndrome (CTS) 1 . CTS is the most common peripheral entrapment neuropathy and is caused by compression of the MN at the level of transverse carpal ligament of the volar wrist 2–7 .…”

“…[6][7][8][9][10] Lv et al have demonstrated that the epineurium should be included in the ROI for more consistent CTS measures. 1 Once reason that including the epineurium in the ROI provides more consistent results is likely due to the improved conspicuity of the epineuriumextraneural connective tissue interface compared to more vague interfaces of the epineurium-perineurium-interfasicular interfaces (Figure 1). Thus, the specific echoreflectivity of the external surface of hypoechogenic epineurium compared to adjacent structures such as perineural adipose tissue provides greater contrast for segmentation so that there can be greater inclusion of altered pathologic tissues that provide ultrasound-specific features of CTS in the ROI.…”

“…1,6,7,[17][18][19][20] As Lv et al report, using artificial intelligence models and setting specific guidelines and automatic segmentation routines for ROI determination will help reduce operator variability and improve the accuracy and reproducibility of ultrasound for the diagnosis of CTS and other nerve entrapment syndromes. 1,7,17,18 F I G U R E 1 Carpal tunnel syndrome. (A) shows an ultrasound image of the median nerve with carpal tunnel syndrome.…”