Trushdeep Agrawal, Founder at Association of Physicians & Engineers, shared on LinkedIn:
“Our research on Integration of MNN-GA and PCA for Lung Cancer Diagnosis has been accepted as an ePoster at the ESMO – European Society for Medical Oncology AI and Digital Oncology Conference!
Our project focused on the application of Principal Component Analysis (PCA) for decreasing the number of dimensions/input features in order to decrease the required computational strength and yield better results at a lower computational power and give a faster yield of results.
We used both public dataset as well as private hospital acquired dataset to work with around 156 CT Chest Scans.
We were able to achieve maximum accuracy, while simultaneously decrease the number of input features from 71 to 31.
This project was a true team effort, and I’m so grateful to my co-authors:
Dr. Vijay Agrawal (HVPM’s COET), Dr. R. S. Arora (Sujan Surgical Cancer Hospital and Amravati Cancer Foundation) and Dr. Aruni Ghose (Barts Cancer Institute (Queen Mary University of London), OncoFlow)
As the most junior member, I learned so much from their expertise and guidance. Their insights were invaluable in developing this novel Modular Neural Network (MNN) that combines Genetic Algorithm (GA) and Principal Component Analysis (PCA) to improve the accuracy of lung cancer diagnoses from chest CT scans.
Thank you for your tireless work and mentorship. Your contributions were instrumental in making this research a success. I’m excited to share our findings and represent our team at the conference.”
More posts about Lung Cancer.