New generation machine learning tools and technologies that help companies in marketing could be used in the public healthcare system for efficient intervention in diseases like cancer, according to an American human genetics and genomics researcher who uses computational tools to understand complex traits such as cancer.

“These analytic tools are currently helping Amazon predict which books you want to purchase or a Google nerd to decide which ad to present to you. But it’s a shame that we are not tapping the same technology to figure out what cancer drugs work in which individuals? What therapies are best suited for which people? How do different backgrounds influence who would respond or not respond to a particular intervention,” said Ken Buetow, a professor at Arizona State University.

Buetow, who was in India recently to deliver a lecture at the invitation of NGO Open Health Systems Laboratory, was part of a team that designed Cell Studio, a virtual platform that allows 3D graphical simulation of cellular processes inside a human body.

India is projected to see 1.2 million new cancer cases this year and a 20 per cent increase is expected by 2020. And within the next 10-15 years, that number is going to be about 1.7 million.

That’s substantially larger than the burden faced by most Western countries.

According to Buetow, India, which is known for its IT capabilities and has a good network of cancer physicians, is well positioned to undertake this task. Besides, the country has no legacy infrastructure that has to be remade unlike the West and hence can implement the infrastructure from Day 1, leveraging the experience of other countries and communities.

Citing the example of the US, which was a pioneer in generating Electronic Health Records (EHRs) of patients, he said the digital health infrastructure in the US is of no use in patient care management. The system for which the US has so far spent $38 billion is mainly for billing. Moreover, the country is spending on an average $25,000 per physician annually as it pays doctors to key in the data, he said.

“So they didn’t consider a work flow and didn’t consider the information doctors would want,” he said When such a system is designed, physicians ought to have a prominent seat at the table.

Also, biomedical research and healthcare delivery should be partners, Buetow said.

“Biomedical research and health-care delivery are twins separated at birth, they should be joined together if you want a rapidly learning health system,” he said adding that such systems would learn from treatment of each patient and benefit the next patient.

“This would mean no patient’s pain goes in vain,” said the American professor.

Related Topics
comment COMMENT NOW