The National Heart, Lung and Blood Institute (NHLBI) – a component of the National Institutes of Health (NIH) – is researching the use of artificial intelligence technologies in health care with the aim of improving screening and analytics capabilities.
Keyvan Farahani, the senior data science, imaging, and AI program director at NHLBI, said during a GovExec event on Aug. 7 that AI could be used to “improve prevention of disease or diagnosis and treatment” if data necessary to feed AI models is standardized and organized.
“A lot of imaging AI has to do with pattern recognition, which machines are really good at and sort of seeing patterns in data that are not visible to human eye,” said Farahani. “Machines can glean from data sets, and these can be used for earlier detection of diseases such as cancer, cardiovascular diseases, and neurological disorders.”
Some of the biggest challenges in integrating AI are inconsistent data standards in research and institutions, the presence of biased data, and training a workforce for the use of AI in the medical field, Farahani said.
The NHLBI’s work toward solving these problems and development has included providing research funding and grants for data curation, preparation and analytics. The NHLBI’s Catalyze Program – which provides funding and services for developing therapies and technologies while training a diverse biomedical workforce – helps to provide infrastructure for medical technology while the NIH’s Bridge to AI program focuses on biomedical data standardization and best practices.
Other ways NHLBI and NIH have supported AI advancements include public-private partnerships with tech companies and academic institutions to “accelerate development of AI techniques and technologies.”
Other government agencies have also been making the push to integrate AI into health care. Google and the Department of Defense’s (DoD) Defense Innovation Unit (DIU) recently announced their development of a “AI-powered microscope” for cancer detection.
This microscope, coined the Augmented Reality Microscope (ARM), is “deployed at military treatment facilities around the world” and uses AI algorithms to “analyze digitized tissue samples” and highlight “potential abnormalities,” according to a press release.
As AI technology becomes more widespread, it could also help “democratize” healthcare access for low-income communities, Farahani said.
“There’s going to be more refined responses to patients’ questions, but also on the health care provider side, we anticipate that there would be better summarization of the data about the patient, their health care records and potential treatments that would help provide a more accurate and efficient system for care of these patients,” said Farahani of the future of AI.
