Science Fair Rock Star: Anika Cheerla

This is the second in an ongoing series highlighting some of the amazing young women participating in this year's Google Science Fair. (Here is the first post.) Each has survived the rigorous first pass and have moved into the first regional round. The second set of winners will be announced on August 11, narrowing the field of candidates for the final round. You can be sure I'll be watching the celebration event on September 27, 2016. 

Looking at the list of Regional Finalists for the 2016 Google Science Fair, one name jumps out at me right away. Anika Cheerla was one of the 2015 Finalists as well! Last year she created a neural net program that could be used to detect brain damage when presented with MRI scans and clinical features and aid in diagnosing Alzheimer's disease in patients. Her Alzheimer's screening program has a remarkable 95% accuracy rate, and dramatically cut down on the "wait and see" time doctors often employ for patients whose scans are not as clear to the human eye. Oh, and at the time she was only 13.

She's back with another brilliant bioinformatics project: Automated Prediction of Future Breast Cancer Occurrence from Non-Cancerous Mammograms. This time she's using her love for research and coding to predict the likelihood of a patient developing breast cancer by analyzing the data from previous mammograms. And, again (no surprise!), her method works.

She took images from over 400 mammograms of healthy breast tissue and ran them through a series of algorithms to teach her code how to recognize changes that eventually lead to cancerous growths, creating a program that is 35% better at predicting cancer than previous methods employed by medical staff.

Accurate prediction of an individual's future cancer risk helps doctors and patients alike: enabling the early detection of breast cancer if it does arise and allowing patients to get possible interventions to lower their cancer risk. Hopefully, this research is a step towards reducing the fatality rate of breast cancer. My system also can significantly reduce burden on on our health care systems and patients by reducing the false positive rates of the mammograms, which often leads to expensive MRI scans and invasive biopsies.

Until now, the purpose of regular mammogram screening has been to detect existing cancer. But what if it could predict cancer before it develops? How great would it be to have more infrequent screenings that were able to give doctors and patients a better idea of who is more likely to develop cancer based on changes in breast tissue, relieving the financial strain on medical facilities and saving many patients undue worry while giving others a better chance through pre-treatment routines?

Perhaps by the time Anika Cheerla is able to fine tune her algorithm through further study, this may become a reality. Did I also mention she's only 14? With brilliant young scientists like this, the future is bright for all of us.

If you appreciate the work I do here on SRPS, please support me!