NIH devoted $130M to developing AI in medicine

Artificial Intelligence in the biomedical industry has already contributed to speedy drug discoveries and advanced diagnostic tools that have led to new information regarding the human genome. However, a gap in information still exists and researchers are unable to reap the full benefits of these advancements because current AI is not compatible to read diverse types of data.

Experts explain that present data sets are not only disorganized but often also lack information on the type of data collected and the collection condition. Current data is just not sufficient and as a result, it cannot be analyzed or interpreted.

Moreover, current AI also does not take into account social and ethical contexts when data is collected which may create bias and inequities, therefore special attention is required in these areas. It is because of these problems the use of AI in biomedical and behavioral research is still restricted. For researchers to be able to efficiently use AI for research purposes well annotated data sets and the use of best practices for collection is a necessity.  

Consequently, the National Health Organisation launched its Bridge to Artificial Intelligence program (Bridge2AI) to assist researchers in creating guides for ethically sourced data that will help understand how factors such as genes, the environment, and behavior impact an individual’s health.

Under the project, NIH plans to spend $130M over the course of 4 years. However this will depend on the availability of funds in the agency. As of yet, the NIH has already distributed $7.7M to the University of California (UC) Los Angeles, UC San Diego, and the University of Colorado Denver. In addition to this, another $22.4 million have been allocated to the program’s first four data generation projects. The agency has also issued 3 awards to create a Bridge Centre for evaluation, integration, and dissemination. 

The Bridge2AI program aims to create data set types including voice recording and genomic analysis to identify abnormalities in the body. The data collected will be used to improve decision-making in crucial areas such as illness recovery. It will also be used to draw connections between genetic abnormalities and certain health conditions. Data sets created under the program will be used for training projects and the AI tools that will be developed will be used to predict health outcomes and find biomarkers in voice recordings.

The program will incorporate talent from diverse backgrounds to create tools that are responsive to AI approaches. The program aims to ensure that the tools generated do not contain any ethical problems such as problems with data trustworthiness, biases, and privacy. NIH has placed a special emphasis on diversity in this program as it is a crucial part of AI research and technology.

According to the acting director of the NIH, Dr. Lawrence A. Tabak, “It is essential to generate high quality, ethically sourced data sets if we want to transform how we do research, connecting researchers to the right AI technology today will not only improve human health but also tackle the most important research questions we have. The answer for all our questions is just within reach now.”