Data Science Bowl Winners Harness AI to Accelerate Life-Saving Medical Research
Imagine unleashing the power of artificial intelligence to automate a critical component of biomedical research, expediting life-saving research in the treatment of almost every disease from rare disorders to the common cold. This could soon be a reality, thanks to the fourth Data Science Bowl, a 90-day competition in which, for the very first time, participants trained deep learning models to examine images of cells and identify nuclei, regardless of the experimental setup—and without human intervention. Algorithms developed in this competition could save researchers hundreds of thousands of hours of effort per year.
This year, the competition brought together nearly 18,000 global participants, the most ever for the Data Science Bowl. Collectively, they submitted more than 68,000 algorithms and worked an estimated 288,000 hours to automate the vital, but time-consuming, process of nuclei detection.
Identifying nuclei allows researchers to find each individual cell in a sample and, by measuring how cells react to various treatments, understand the underlying biological processes at work. “By identifying nuclei quickly and accurately, the algorithms developed in this competition can free up biologists to focus on other aspects of their research, shortening the approximately 10 years it takes for each new drug to come to market and, ultimately, improving quality of life,” said Ray Hensberger, a Booz Allen Hamilton principal. “This year’s Data Science Bowl was fascinating because nuclei detection is crucial to biomedical research, but detection methods require time-consuming biologist oversight. Until now, there have never been any deep learning models available that can identify nuclei across multiple experimental setups and testing conditions.
2018 Data Science Bowl winners include:
First Place: Victor Durnov, Alexander Buslaev, and Selim Seferbekov, an international team from Russia and Germany who have competed against each other in previous crowdsourcing competitions but formed a team with the specific goal of winning this year’s competition.
Second Place: Minxi Jiang, who finished in the top one percent in last year’s Data Science Bowl.
Third Place: Angel Lopez-Urrutia, a marine ecologist in Spain who uses machine learning to automatically classify images of plankton, a challenge that was central to the inaugural Data Science Bowl.
Third place winner Angel Lopez-Urrutia said, “I have always been fascinated by the combination of biology and machine learning so, after reading about the Data Science Bowl for years, I decided to enter this year’s competition. I thought this competition would be a valuable opportunity to test my knowledge in a different field than my day job, but I never imagined I might place in the top three. I’m proud to have played a role in this competition, as it demonstrates that machine learning can make a difference in the scientific community by automating repetitive tasks so researchers can focus on more impactful work.”
“These solutions represent a paradigm shift in the way microscopy images are processed in biomedical research and will make research more accurate and efficient,” said Dr. Anne Carpenter, director of the Imaging Platform at Broad Institute of MIT and Harvard. The next step being investigated is to create a user-friendly and open-source software that biomedical researchers can begin using in their day-to-day work.
Read more about the Data Science Bowl and the lasting impact it has made, here.
Creators of the top algorithms will split $170,000 in cash and prizes, including NVIDIA GPU hardware for deep learning, such as the NVIDIA DGX Station and NVIDIA TITAN GPU. Researchers from all over the world use GPU computing to power their work. The competition receives sponsorship and support from several leading health and technology organizations, including the Broad Institute, NVIDIA, and PerkinElmer, Inc.
Kaggle is the world's largest online data science competition community. With over 1.3 million users across 194 countries, the Kaggle community uses its diverse set of academic backgrounds to solve complex data science problems. Working as individuals or in teams, the winning competitors are awarded prizes and industry recognition for their accomplishments. The top competitors are invited to work on the most interesting and sensitive business problems from some of the world’s biggest companies through Masters Competitions.