How AI is Used in Cancer Research
Machine learning for cancer prediction isn’t a new thing. Decision trees, Convolutional neural networks have all been used in predicting cancer. According to PubMed stats, more than 1500 papers have been the subject of cancer prediction.
Now, there is a difference between diagnosis and prognosis. Prognosis is about predicting before diseases happens and diagnosis is about trying to identify the diseases. AI can be used in 3 ways to predict cancer: 1) the prediction of cancer for the 1st time, 2) the prediction of cancer recurrence 3.) the prediction of cancer survivability.
Before, Cancer prediction involves many physicians from different specialities, they take a look at the patient and also the stats from a macro level. Even for skilled physicians, this is not easy to do.
Overall, cancer prediction was hard before machine learning, even the best physicians find it hard. However, AI has changed the cancer physicians job rapidly and AI has helped many cancer patients. Things like CNNs, Decision Trees, and Linear Classification has changed the way humans predict cancer.