MR Imaging Radiomics Analysis Based on Lumbar Soft Tissue to Evaluate Lumbar Fascia Changes in Patients with Low Back Pain

• Published in: Academic radiology
• Main Authors: Song, Ming-Xin, Yang, Hui, Yang, He-Qi, Li, Shan-Shan, Qin, Jian, Xiao, Qiang
• Format: Journal Article
• Language: English
• Published: United States 01.11.2023
• Subjects: Low Back Pain; Fascia; Nomogram; Radiomics
• ISSN: 1878-4046
• DOI: 10.1016/j.acra.2023.02.038

Clinicians must precisely pinpoint the etiology of low back pain as the number of people suffering from it increases to provide targeted care. The purpose of this paper was to use MR imaging radiomics based on lumbar soft tissue to analyze changes in the lumbar fascia of patients with low back pain. We retrospectively analyzed the lumbar MRI of 197 patients with low back pain. Patients were randomly assigned to either the training (n = 138) or validation (n = 59) cohorts. Multivariate logistic regression analysis was used to create radiomics model and combined nomogram model and their predictive performance were evaluated using receiver operating characteristic curves. Seven radiomics features based on lumbar soft tissue MRI images were established, which performed well in distinguishing between low back pain patients with fascial changes and normal individuals demonstrated an excellent ability to identify differences, with an Area Under Curve (AUC) of 0.92 (95% CI, 0.88-0.96) in the training cohort and 0.84 (95% CI, 0.73-0.96) in the validation cohort, which performed better than the clinical model significantly only. The nomogram based on clinical features and radiomics features of MR images had a good predictive ability to differentiate fascial alterations in patients with low back pain from normal subjects. It had the potential to be used as a decision support tool to assist clinicians in determining the etiology of patients with lower back pain and managing patients promptly, particularly in the early stage of the fasciitis when significant abnormalities on imaging were difficult to detect.