Biomechanical modelling and computational imaging to identify different causes of back pain in large epidemiological studies

Introduction

Chronic back pain is a major burden and source of disability worldwide. It is primarily attributed to different biomechanical factors but can also have inflammatory, neurological, or psychological causes. Clinical findings and conventional imaging cannot reliably distinguish different causes of back pain.

Biomechanical Models

In contrast, individual biomechanical models can quantify diverse (pathologic) loading patterns and thus could be used to distinguish different aetiologies of back pain. This approach aims to better understand individual pathophysiology and guide preventive strategies.

Project Overview

During my recent ERC-StG iBack, I developed quantitative imaging methods and deep-learning based image processing to automatically generate a fully individualized biomechanical model of the thoracolumbar spine.

Simultaneously, two large-scale epidemiologic studies collected clinical and high-resolution imaging data of the spine of more than 15,000 participants so far, aiming at more than 35,000 participants by mid-2022.

Objectives of iBack-epic

The high-level objective of iBack-epic is to use such novel image analysis techniques to identify different biomechanical and inflammatory causes of back pain in study participants.

Methodology

I will adopt and extend my recently developed deep-learning based spine labelling and segmentation algorithms to fully automatically calculate individual biomechanical, functional, and morphometric parameters of the spine.

In this large-scale population data, I will:

1. Identify different biomechanical loading patterns.
2. Use quantitative image-based parameters to discriminate normal ageing from pathologic degeneration.
3. Identify pathological conditions that are linked to back pain or subsequent development of chronic back pain.

Expected Outcomes

Such differentiation, for the first time based on quantitative image data, will allow for a better understanding of the underlying pathophysiology of back pain.

This will lead to:

• Improved risk stratification.
• Tailored investigation of genetic causes.
• Better guidance for preventive strategies.