About 15 years ago, I was overjoyed to hold my beautiful newborn son, Aaron Matthew Kahan, in my arms. However, my joy was short-lived, as Aaron soon passed away of sudden infant death syndrome (SIDS). To this day, very little is known about SIDS, which is the leading cause of death for infants between the ages of one month and one year in the US and other developed countries.
Aaron’s death led me and my wife on a transformational journey that resulted in the formation of a global SIDS mission. It is peopled with data scientists, medical professionals and IT workers, all of whom have put their energies together to reduce infant mortality. But the beginning of this journey was equally serendipitous. Having seen a picture of Aaron on my desk, a data scientist colleague had asked me more about him.
An expectant father himself, the colleague was shocked to hear of Aaron’s passing, and scared to contemplate his unborn child’s future. He then decided to join forces with us to work out a solution for this condition. Our combined research pointed us to potential genetic markers and led us to begin building the first (cloud-based) genetic database ever to help researchers collaborate worldwide.
Today, a team of SIDS researchers, Microsoft data scientists, and TCS IT and genomics researchers together explore genetics, epidemiology, physiology, pathology and education/outreach to collaboratively address this disease. We believe that there are biological markers present in the genes of children who ultimately die of SIDS. With our infant genome database, parents can run sequencing tests on themselves and the potential siblings of the unborn child, determine risk factors and develop preventative measures.
In the US alone, nearly 60,000 children have died since Aaron’s passing 15 years ago. That’s equivalent to two busloads of children disappearing every week for the past 15 years, without anyone knowing why. Despite all the statistics available through the Center for Disease Control in the US, we face the challenge of inconsistent data, leading to a lot of what we call ‘noise’. Moreover, it does not give biological details of each child. This is why we built a genomic database that would employ machine learning to help us understand what is happening to these children.
The great news is that costs continue to drop as technology grows at an exponential rate, opening up possibilities and capabilities that we could not have imagined before. Just a year ago, the cost of sequencing a child’s genome was about $4,000. Today, at our scale, it’s about $1,700. Advances in data science and medical research will further bring down the cost of sequencing. Also, machine learning and AI capabilities are growing astronomically. As a result, our ability to understand large-scale data sets is much greater today.
As technology improves over the next few years, we will gain the ability to operate faster, and at scale. With partnerships between private and public sectors worldwide, we will break down antiquated barriers further. And with technology protecting individual privacy, we will be able to stop terrible diseases in their tracks to dramatically reduce infant mortality.
About the Author
Chief Data and Analytics Officer, Microsoft, AI for Good
John is responsible for using data science to address the world’s great challenges. Among the issues he is tackling head-on are promoting sustainable use of the planet’s resources, improving opportunities for people with disabilities, protecting human rights, strengthening humanitarian assistance in disaster response, addressing the needs of children, refugees and displaced people, protecting human rights and increasing the capabilities of the world’s NGOs.