Posted on April 26, 2017

For anyone that has attended a school science fair, you might expect the typical volcano projects or even the usual Mentos and soda displays. However, some of today’s young scientists exemplify levels that far outpace those types of situations. Sarasota County Florida has had some of the top science projects in the country, with students that have gone on to compete in both the Intel and Google International Science Fairs…and win. Thus is the story of a young lady by the name of Brittany Wenger, who created a non-invasive method to detect breast cancer.

At 17 years of age, this teen from Lakewood Ranch, FL blended the fields of computer science and biology to create an app that assists doctors in breast cancer diagnosis. It is called the ‘neural network’ and as described by Wenger, it is designed to mimic the human brain aka an artificial ‘brain’ that can learn and adapt in detecting complex patterns that can be a diagnostic tool to diagnose breast cancer.

Brittany’s program uses data from ‘fine needle aspirates’, which is a minimally invasive process. Wenger knew that this is sadly one of the least precise methods used for diagnosis and her plan was to make use of today’s technologies to change that. Her app that makes use of the cloud, has resulted in correctly identifying ninety nine percent of malignant tumors.

A full description of Brittany’s project can be found on the Google Science Fair page. Brittany’s project won the local Sarasota County science fair, went on to win the state science fair and when she entered it into the International Google Science Fair, she was the grand prize winner. This latter achievement is no easy task as students from all over the world compete in this astounding fair.

To get an idea of how revolutionary her project was and the genius of this young girl, you need only read a part of the Google page project description:  The successfully implemented custom network is tested with 6,800 trials.  To assure maximum training, each sample is run through ten trials evaluated by different networks trained against all other samples.  The custom neural network achieved predictive success of 97.4% with 99.1% sensitivity to malignancy – substantially better than the evaluated commercial products.  Out of the commercial products, two experienced consistent success while the third experienced erratic success. The sensitivity to malignancy for the custom network was 5% higher than the best commercial network’s sensitivity. This experiment demonstrates modern neural networks can handle outliers and work with unmodified datasets to identify patterns. In addition, when all data is used for training, the custom network achieves 100% success with only 4 inconclusive samples, proving the network is more effective with more samples.  Additionally, 7.6 million trials were run using different training sample sizes to demonstrate the sensitivity and predictive success improves as the network receives more training samples.

 The Global Neural Network Cloud Service for Breast Cancer may be ready to diagnose actual patients – more global participation is required to confirm the findings and increase the predictive success on blind samples.