Artificial intelligence is coming into its own in medical applications at an increasingly rapid pace. We’ve seen reports of impressive results from deep-learning machine language diagnostic studies from research groups worldwide. Most recently we wrote about a joint project involving MIT, Harvard Medical School, and Massachusetts General Hospital using AI to assess breast density: a breast cancer flag. We wrote about Google AI predicting clinical outcomes in another joint study, this one with hospitals in California and Chicago. An international study found a machine-language trained neural network outperformed a group of expert dermatologists from 17 countries in diagnosing melanoma from images of skin cancers. AI’s future in medicine appears to be a sure bet not only due to the speed with which machines can analyze data, but also the high degree of accuracy demonstrated in the observations and decisions gleaned from the research.
Scripps Research Translational Institute and NVIDIA are major entities in medical research and machine learning. The two companies recently announced a partnership to establish a center of excellence for artificial intelligence in genomics and medical science. The overall purpose of the joint effort is to advance the development and use of machine learning with genomic and health data. The pair intends to establish an infrastructure including tools and best practices to handle the exploding quantity of health data. Faster, more affordable genome sequencing gear and digital health sensors such as smartwatches, blood pressure cuffs, and glucose monitors generate such massive data output that artificial intelligence applications provide the most effective tools to analyze the mountains of information. According to Scripps, the growth in genomics data led to a 40-fold increase in deep learning approaches in genomic research studies in the last four years. The center of excellence, which as yet as no formal name, intends to provide a platform to leverage the skills and tools of diverse research efforts worldwide.
Big data by definition is too large for traditional data processing applications. Big data’s quantity, complexity, and rapid growth makes it challenging to reveal the answers hidden in masses of information. This new partnership could lead to new tools and new understanding that will create new ways to find valuable insights that could transform healthcare.
