Prediction of Pulmonary Fibrosis Progression using CNN and Regression
J Madhuri, Vikram Bhushan, Sai Kishore HR, Sharan Kumar G, J Shreyas · 2021 3rd International Conference on Advances in Computing, Communication Control and Networking (ICAC3N) · 2021
The term idiopathic is used to refer to disease caused by unknown causes. Idiopathic pulmonary fibrosis is a chronic, progressive disease which affects the lungs and causes scar tissues to develop within them. This prevents the patient’s lungs to transport oxygen into the bloodstream effectively. The daily routine of the patient is affected as the ease of breathing continues to decline. Early medical intervention and proper diagnosis can help keep the disease under control. The severity of the disease is measured used Forced Vital Capacity (FVC) values. In this paper we design an easy-to-use web application which collects the patient’s CT scans, characteristic data and the initial FVC measurement. We predict the FVC values for up to a period of 2.5-3.0 years which effectively provides the Pulmonologist the rate of decline upon which suitable medications can be provided to stall the decline. We use Regression Techniques and CNN architectures to predict FVC values.