344 Income predicts low back pain but not lumbar disc height: Data for the UK Biobank Image Dataset

Abstract

Objectives/Goals: Here, we utilize deep learning to automate the analysis of dual X-ray absorptiometry (DEXA) scans in the UK Biobank (UKB) imaging dataset to enable a large-scale assessment of lumbar spine disc degeneration, low back pain, and socioeconomic status. Methods/Study Population: Study Population: The UKB is a biomedical database that includes lateral spine DEXA imaging for 50,000 participants. Deep Learning Model Development: A computer vision model was developed that receives a DEXA scan as input and outputs a quadrilateral that corresponds to the corners of 5 lumbar vertebral bodies. The model is a deep, fully convolutional, encoder–decoder network using DeepLabV3. Statistical Analysis: To determine our preliminary model accuracy, we used the intersection over union (IoU) metric.We analyzed data using an ordinal regression model to determine the relationship between income/ neighborhood level multiple deprivation index (MDI) and low back pain (LBP), as well as a mixed effects model to estimate the relationship between income/MDI and disc height index (DHI). Results/Anticipated Results: Our model predicted vertebral body quadrilaterals in training and unseen test data (train IoU  =  0.96, test IoU  = .91) and was used to infer data for 10,440 participants. Confirming previous studies, there were significant relationships (p0.05) between income or MDI and DHI (Figure 2). Discussion/Significance of Impact: Low back pain is the world’s leading cause of disability, and socioeconomic factors play an important role. We found no relationship between disc height index and socioeconomic status. Thus, disc degeneration may not be a factor in this low back pain phenotype.