According to the National Cancer Institute, in 2016, there were 1,685,210 newly diagnosed cases of cancer in America with 595,690 deaths.  With such staggering numbers, millions of dollars are poured into research into how to fight and treat cancer.  One of the newer innovations in cancer treatment is the increased use of machine learning and precision medicine.  Precision medicine is the act of specifically tailoring a patient’s treatment to his or her own genetic needs.  To do this, machine learning helps doctors provide the best care possible by using predictive analytics to recommend specific treatments.  This week’s AI spotlight will focus on how the Swedish Cancer Institute is partnering with the company GNS Healthcare to use machine learning to better treat cancer patients.

Machine learning has serious potential to aid in cancer treatment.  According to the Executive Director of the Swedish Cancer Institute, Thomas Brown, MD, it is rare that an oncologist has the “complete training or time necessary to decipher complex results of a tumor’s biological footprint.”  Machine learning, on the other hand, has the capability to do this. GNS is able to use reverse engineering and machine learning models to also support providers and give personalized treatment recommendations.

Personalized information in the area of cancer research is important because understanding genomic information plays a key role in treatment.  Each person’s genetic make-up is different, so cancer treatment methods may not work the same for one patient as they would for another.  By collaborating with GNS Healthcare, the Swedish Cancer Institute is able to share its data and expertise so that the engineers and data scientists of GNS Healthcare can aid oncologists with deciphering a tumor’s biological footprint.  The collaboration has specifically aided breast cancer patients.

This collaboration is not the only attempt at using machine learning to facilitate precision medicine and cancer treatment.  As we have written in a previous spotlight, IBM’s Watson has also aided in this process. In an IBM study, doctors agreed with IBM Watson for Genomics’s recommended treatment for cancer 96 percent of the time.  This study shows just how effective precision medicine can be.  Additionally, the U.S. Department of Veteran’s Affairs will use Watson for Genomics to offer cancer treatment to veterans who need it.

It is important to note that, in line with SIIA’s views, these machine learning tools for precision medicine serve as a supplement to enhance the ability for doctors to provide care.  They do not replace the need for human doctors.  Despite the tremendous accuracy of some of these precision medicine tools, humans are still necessary to improve upon the accuracy so that each patient can receive the care that they may need in comparison to what a different cancer patient needs.   The collaboration between the Swedish Cancer Institute and GNS Healthcare also highlights the need for human research and input to improve the algorithms by combining the input of cancer researchers with engineering and data scientists.

Precision medicine is still a fairly new and growing field in medicine.  Whether it be for cancer treatment or drug discovery, there is no doubt that machine learning in this area will continue to play an important role.  Rather than displacing the phenomenal work of human doctors, machine learning will help doctors in areas where there is a shortage of proper expertise or training, as is the case with the Swedish Cancer Institute’s collaboration with GNS Healthcare.  With so many thousands of people suffering from cancer, doctors need all the help they can get.  Precision medicine will continue to provide doctors with the tools they need.

Diane Pinto is the Public Policy Coordinator at SIIA. Follow the Policy team on Twitter @SIIAPolicy.