Use of deep learning to predict response of treatment for late stage lung cancer patients.
In late stage lung cancer cases, oncologists make very complex and important decisions on the type and modality of treatment best suited for a patient. The decision depends on multiple factors, including diagnostic reports, image biomarkers, patient history, age and tumor characteristics.
In this project, we are studying the use of deep learning to predict the response of treatment on a lung cancer patient. We would study the importance and use of metrics from the screening chest x-ray images, biomarkers from the microscopic images of pathological biopsy, clinical information of the patient to predict the response to a treatment along with its prognosis.
1. Predict the response of different treatment modalities for lung cancer patients based on the following factors -
-- Screening chest x-ray images (size of nodules, location of nodue, etc)
-- Microscopic image of biopsy (tumor grading based on mitotic activity, # of tumor infiltrating lymphocytes, etc)
-- Clinical information, age, smoking history, survival time.
2. Predict the probability of relapse post treatment for lung cancer patients based on the above mentioned factors.
Dheeraj Mundhra, SGInnovate - Entrepreneur First