Google's Deep Learning AI Program to Detect Lung Cancer

Image:BBC

We all know that continuous development in technological advancement has played a major role in the field of healthcare. There are lots and lots of technological instruments and machines which are introduced since few years, and now Google engineers have come up with something. Early diagnosis of lung cancer by doctors is quite impossible, as symptoms of lung cancer are very common and cannot be specified easily. So there is necessity of some kind of technique which can detect the symptoms of lung cancer at first stage. Google engineers have developed that technique: an Artificial Intelligence program which would be able of diagnosing lung cancer, that too more accurately.

As per the journal “Nature Medicine” study titled :End-to-end lung cancer screening with three dimensional deep learning on low-dose chest computed tomography”, team of researchers at Google are planning to use deep learning program which will be able of detecting lung cancer in patients with an accuracy of approximate 95%. This plan of Google engineers is under construction.

Chest Radiography is one of the first investigative steps if a person reports symptoms that may be suggestive of lung cancer. But Google Artificial Intelligence program can detect the lung cancer with more accuracy compares to Radiography.

This new model of AI program comprises of three fundamental components:

1. Construction of 3D CNN (Convolutional neural networks) model that performs end to end analysis of all CT (Computed Tomography) volumes.
2. Training of CNN region of interest detection model to detect 3D cancer candidate regions in CT volume.
3. Prediction model which provides output.

This deep learning model for detecting malignancy risk in lung cancer is prepared by feeding lot of data from NLST(National Lung Cancer Screening Trial) which comprises of 42290 number of CT cases from 14851 patients. This dataset will help the new AI deep learning program to identify the medical complications easily, which is quite difficult for doctors to detect.

“We have some of the biggest computers in the world. We started wanting to push the boundaries of basic science to find interesting and cool applications to work on”, said Dr. Daniel Tse , a researcher at Google.