Pulmonary Nodule Detection And Lung Cancer Prediction
Final year project currently in-progress at NIT Surat
Group:
- Mohith Damarapati
- Sachin Malepati
- Vatsal Sodha
- Konu Prithvi Raj
Guided By: Dr. Mukesh A Zaveri, Associate Professor at NIT Surat
Description:
Lung Cancer is one of the most commmon cancers and cause of cancer related deaths all over the world. Early diagnosis of lung cancer can significantly depreciate deaths due to it. Cancer screening using low dose CT scans is the most common and efficient diagnosis technique.
Lung Cancer demonstrates itself in the form of pulmonary nodules which are visible on CT scans. Nodule detection is an ardous task and highly error-proned. We plan to detect these nodules automatically and develop a CAD system that could assist radiologists or at some day, replace them.
Data:
We got the data-set from the Lung Image Database Consortium and Infectious Disease Research Institute (LIDC/IDRI), largest publicly available lung CT scan data.
Models:
Following are the models trained:
-
Classical Machine Learning algorithms like SVM, Naive Bayes, Random Forest, KNN etc based on 8 features chosen from a research paper.
-
Deep Nets based on 8 features.
-
Deep Nets based on pixels as features.
-
Convolutional Neural Nets.