Machine Learning Diagnostic Modeling for Classifying Fibromyalgia Using B-mode Ultrasound Images

Abstract

Fibromyalgia (FM) diagnosis remains a challenge for clinicians due to a lack of objective diagnostic tools. One proposed solution is the use of quantitative ultrasound (US) techniques, such as image texture analysis, which has demonstrated discriminatory capabilities with other chronic pain conditions. From this, we propose the use of image texture variables to construct and compare two machine learning models (support vector machine [SVM] and logistic regression) for differentiating between the trapezius muscle in healthy and FM patients. US videos of the right and left trapezius muscle were acquired from healthy ( n = 51) participants and those with FM ( n = 57). The videos were converted into 64,800 skeletal muscle regions of interest (ROIs) using MATLAB. The ROIs were filtered by an algorithm using the complex wavelet structural similarity index (CW-SSIM), which removed ROIs that were similar. Thirty-one texture variables were extracted from the ROIs, which were then used in nested cross-validation to construct SVM and elastic net regularized logistic regression models. The generalized performance accuracy of both models was estimated and confirmed with a final validation on a holdout test set. The predicted generalized performance accuracy of the SVM and logistic regression models was computed to be 83.9 ± 2.6% and 65.8 ± 1.7%, respectively. The models achieved accuracies of 84.1%, and 66.0% on the final holdout test set, validating performance estimates. Although both machine learning models differentiate between healthy trapezius muscle and that of patients with FM, only the SVM model demonstrated clinically relevant performance levels.

Pressure Algometer: An Effective Diagnostic Tool to Identify Latent Myofascial Trigger Points

“…Researchers state that a few diagnostic techniques, including magnetic resonance elastography, electromyography, pain pressure threshold, and ultra sonography, have been developed to locate and detect MTrPs as well as to identify and characterise their properties (Taheri et al, 2016). Studies reported that, TrPs were analysed through ultrasound elastography by doppler method whereas a handheld vibrator was used to induce vibrations.There is a decreased vibration amplitude associated with MTrPs, which appear as focal and hypoechoic nodules (Behr et al, 2020;Srbely et al, 2016).The effectiveness of ultra sonography and its many techniques in this area is supported by evidence. As a result, ultra sonography is regarded as a secure, accessible, and affordable imaging technique for impartial detection of MTrPs and assessment for efficacy of therapies (Sikdar et al, 2008).…”

A theoretical framework to improve the construct for chronic pain disorders using fibromyalgia as an example

“…A similar analysis was performed for FM. 42 Neurophysiological assessments have also been developed in chronic pain conditions like FM, headache and osteoarthritis. Current research suggests that central sensitisation may be an important part of the syndrome’s pathophysiology, to be included among the features of this construct.…”

“…Current research suggests that central sensitisation may be an important part of the syndrome’s pathophysiology, to be included among the features of this construct. 16 The Nociceptive Flexion Reflex (NFR) threshold has been proposed as a potential biomarker candidate that may assist uncovering the neurophysiological pain mechanism that could ultimately play an important role in a more homogeneous diagnostic categorization and an accurate treatment, 43 – 47 despite its well-known between-subject variability. 48 The NFR threshold has been described to be a marker not only of pain, 49 but also of the neuroanatomical reorganization at the spinal cord segments, specifically laminae II, III and IV of the dorsal horns.…”

“…36 Furthermore, recent advances in quantitative ultrasound of skeletal muscle have shown discriminative ability between healthy controls and persons with myofascial pain and FM. 37–40 Kumbhare et al used texture feature analysis to examine the B-mode ultrasound images of the trapezius muscle of subjects with myofascial pain, 41 with latent and active myofascial trigger points, and compared them with asymptomatic healthy controls, finding significant differences. A similar analysis was performed for FM.…”