Dynamic Pose Diagnosis with BlazePose and LSTM for Spinal Dysfunction Risk Estimation

The bodily movements of a person can be heavily restricted because of a spinal cord disorder. These disorders could be a result of various internal or external factors but it is believed that a preliminary diagnostic for the magnitude of the spinal dysfunction can be determined by the movement of the patient itself. The behavioural patterns of a person doing specific tasks with spinal dysfunction varies from someone who does the same tasks with a healthy back. In this contribution, it is demonstrated that this difference can be generalised and then implemented to predict the spinal health index of the patients with only but an input video of them performing predetermined tasks. A novel technique of abstracting dynamic heat-map as the sequential input to the custom LSTM model is featured allowing to outperform previous techniques by reaching 98.4% validation accuracy with a realtime computational complexity of ≤11 milliseconds.

I. Introduction

The spinal cord is responsible for electrochemical communication where the current travels across the nerves, sending signals to various parts of the body to establish a connection with the brain. A healthy spine always has a slight bend that assimilates the stress from the body movements. Recent studies have shown that spine curvature disorders like osteoporosis, spondylolisthesis, kyphosis, scoliosis, etc are increasing day by day in the world affecting the lives of many people disastrously. These disorders are responsible for the misalignment of the spinal cord causing it to bend more than normal. Most common spine disorder symptoms include swaybacks, pronounced buttock positions, problems in moving certain ways, a large gap between lower-back and floor when lying down, bending forward, hump / upper back curve more than usual, uneven shoulder blades and hip/waist positions, leaning towards one side, etc. depending on the severity of the conditions these symptoms vary from person to person.