Early Stage Lung Cancer Detection and Soil Spectroscopy
The goal of the project is to detect lung cancer early, from low dose CT Scans. National Cancer Institute in USA has released a data set containing high resolution low dose lung scans to motivate the researchers in detecting cancer as early as possible. We are aiming to use this data set so that we can create a method to detect lung cancer early in patients. The problem lies in designing a classification method that can detect which patient will suffer from lung cancer within 1 year, just from their CT scans. We hope to use the power of Machine Learning to unsurface the salient features that can be used to make such a classification. Machine Learning & Deep Learning has also proven to break bounds defined by previous works in the field of Medicine. Once we have successfully completed it IN our lab, we want to empower the medical facilities of Pakistan with this technology. At least, one third to 40 per cent of all cancer cases are preventable, if detected early. In Pakistan where medical facilities are scarce and most patients are diagnosed only in advanced stages of cancer, our project will not only can help save lives but also, reduce the strain of cancer patients on medical facilities. Successful delivery of this project will bring about cutting edge technology revolution in Medicine for Pakistan. This project will demonstrate the power of Machine Learning, its applications and inspire new project like ours to propel Pakistan towards Modernism and technological revolution. In agriculture we are working on predicting soil nutrients that are essential for soil health from spectroscopic measurements using advance machine learning techniques. Predicting soil properties is essential for management and assessment of soil quality that in return increase yield and improve quality of crops. Traditional soil testing methods through chemical labs are time consuming, expensive,complex and required skilled labor. Scarcity of chemical labs and lack of awareness of modern farming technologies knowledge among farmers add further complexity specially in developing country like Pakistan. We are developing novel state of the art algorithm in our lab to address this issue.