An Evaluation of the Ambulatory Classification Levels and Reimbursement Methodologies of the Prosthetics Industry

Early onset scoliosis is a three-dimensional curvature of the spine that occurs in patients 10 years or younger. Physicians tend to pursue non-surgical options first. If these are unsuccessful, surgical intervention becomes necessary. In scoliosis cases, physicians use total lung capacity as the metric for determining the optimal time for surgery. Current standards of treatment do not offer an accurate way for physicians to measure total lung capacity for early onset scoliosis patients. This project offers a potential solution to this problem. The total lung capacity can be calculated by subtracting the mediastinum volume from the rib cage volume. A multivariate linear regression was used to create a predictive equation with patient demographics as input variables to predict the mediastinum volume. This equation predicted mediastinum volume at a higher accuracy compared to previous work and had a multiple R2 of 0.87. To calculate the rib cage volume, a convolutional neural network was built. Using X-ray images, a computer could train itself to identify the rib cage. Although the rib cage volume was not calculated, the convolutional neural network was able to identify the rib cage to a limited degree. The model built here can be further improved upon to calculate a volume. If successfully calculated, the mediastinum and rib cage volumes could be combined to calculate the total lung capacity.