Pulmonary Fibrosis Progression Prediction Using Image Processing and Machine Learning
Author | Aboeleneen A.E. |
Author | Patel M.K. |
Author | Al-Maadeed, Somaya |
Available date | 2022-05-19T10:23:08Z |
Publication Date | 2021 |
Publication Name | Advances in Science, Technology and Innovation |
Resource | Scopus |
Identifier | http://dx.doi.org/10.1007/978-3-030-14647-4_11 |
Abstract | The onset of COVID-19 has focused the attention of the research community on lung diseases and conditions. Idiopathic pulmonary fibrosis (IPF), in which internal scarring of the lung takes place, has gone undetected among the various populace. This condition has no known cure. So far, computer vision researchers, along with radiologists, have been successfully able to identify the IPF through lung CT-scans but have had difficulty in identifying the severity of IPF. In this research, we will investigate the use of image processing and machine learning techniques to identify the progression of the disease. For that, we will build two machine learning models and compare them. The first model uses patients biological indications and some histogram features of the CT scans. The second model uses the ensemble method of a convolution neural network (CNN) of patients CT scans and quantile regression of the patient's biological data for predicting the Forced Vital Capacity (FVC, an indicator of IPF severity). The results showed that by using the second model, we got a higher r2 value of 0.93 versus 0.89 using the first model and that the biological data had more importance than the CT scans for predicting the lung declination. |
Sponsor | Acknowledgements This publication was made possible by Qatar University Emergency Response Grant (QUERG-CENG-2020-1) from the Qatar University. The statements made herein are solely the responsibility of the authors. |
Language | en |
Publisher | Springer Nature |
Subject | Deep learning EfficientNet Ensemble learning Idiopathic pulmonary fibrosis Image processing Quantile regression Transfer learning |
Type | Book chapter |
Pagination | 159-177 |
Files in this item
Files | Size | Format | View |
---|---|---|---|
There are no files associated with this item. |
This item appears in the following Collection(s)
-
Computer Science & Engineering [2281 items ]