Presenter Name: Tyler Hutchinson
Description
Lower back pain (LBP) is a serious condition with a lifetime prevalence as high as 39% . The objective of this study is to identify dominant spinal movement patterns (i.e., phenotypes) among chronic lower back pain (cLBP) patients and interpret their significance for clinical applications. We hypothesize the findings from this study will provide clinicians with important information which will facilitate more personalized treatment paradigms and result in improved efficacy of treatment paradigms.
Data were collected from a group of 36 subjects with cLBP using an array of 16 viscoelastic sensors placed on each subject's lower back to detect skin stretch and spinal motion. Subjects were then instructed to perform 6 repetitions of 14 distinct spinal motions. Data were processed to detect the maximum change of resistance, a feature analogous to spinal range of motion.
The subjects were then clustered into phenotypes using a k-means clustering algorithm. The clustering algorithm divided the subjects into 3 phenotype groups for each exercise. These phenotypes were then tested for statistically significant differences among patient-reported outcomes, specifically using Oswestry Disability Index (ODI) scores, using one-way ANOVA and Student T-Tests.
This research shows that subjects suffering from cLBP can be clustered into distinct movement phenotypes. Cluster 1 patients demonstrated reduced right lumbar flexion and left lumbar extension ranges of motion (ROM); Cluster 2 demonstrated reduced upper right extension; and Cluster 3 were typified by slightly below average lower right flexion. Subjects in the phenotype clusters 1 and 2 exhibited higher ODI scores than subjects in cluster 3. These observations confirm that cLBP patients have different motion characteristics, and that these differences may result from different sources or mechanisms of cLBP, or from different coping mechanisms, which also influence the patients ODI scores. The study is an important first step to providing clinicians with the tools to improve prescribed treatment paradigms through greater personalization and tailoring.
Data were collected from a group of 36 subjects with cLBP using an array of 16 viscoelastic sensors placed on each subject's lower back to detect skin stretch and spinal motion. Subjects were then instructed to perform 6 repetitions of 14 distinct spinal motions. Data were processed to detect the maximum change of resistance, a feature analogous to spinal range of motion.
The subjects were then clustered into phenotypes using a k-means clustering algorithm. The clustering algorithm divided the subjects into 3 phenotype groups for each exercise. These phenotypes were then tested for statistically significant differences among patient-reported outcomes, specifically using Oswestry Disability Index (ODI) scores, using one-way ANOVA and Student T-Tests.
This research shows that subjects suffering from cLBP can be clustered into distinct movement phenotypes. Cluster 1 patients demonstrated reduced right lumbar flexion and left lumbar extension ranges of motion (ROM); Cluster 2 demonstrated reduced upper right extension; and Cluster 3 were typified by slightly below average lower right flexion. Subjects in the phenotype clusters 1 and 2 exhibited higher ODI scores than subjects in cluster 3. These observations confirm that cLBP patients have different motion characteristics, and that these differences may result from different sources or mechanisms of cLBP, or from different coping mechanisms, which also influence the patients ODI scores. The study is an important first step to providing clinicians with the tools to improve prescribed treatment paradigms through greater personalization and tailoring.
University / Institution: Brigham Young University
Type: Poster
Format: In Person
Presentation #B19
Area of Research: Engineering
Email: thutchn1@byu.edu
Faculty Mentor: David Fullwood