News • Technology assessment

AI in medical research: Promise and challenges

In an editorial published in PNAS Nexus, Monica M. Bertagnolli assesses the promise of artificial intelligence and machine learning (AI/ML) to study and improve health.

portrait of Monica Bertagnolli
Monica M. Bertagnolli, MD

Image source: NIH

The editorial, which was published in PNAS Nexus, was written by Dr. Bertagnolli in her capacity as director of the National Cancer Institute. 

AI/ML offers powerful new tools to analyze highly complex datasets, and researchers across biomedicine are taking advantage. However, Dr. Bertagnolli argues that human judgment is still required. Humans must select and develop the right computational models and ensure that the data used to train machine learning models are relevant, complete, high quality, and sufficiently copious. Many machine learning insights emerge from a "black box" without transparency into the logic underlying the predictions, which can impede acceptance for AI/ML-informed methods in clinical practice. 

"Explainable AI" can crack open the box to allow researchers more access to the causal links the methods are capturing. AI/ML-informed methods must also meet patient needs in the real world, so interdisciplinary collaborations should include those engaged in clinical care.

Researchers must also watch for bias; unrecognized confounders such as race and socioeconomic status can produce results that discriminate against some patient groups. AI/ML is an exciting new tool that also demands increased responsibility. Ultimately, AI is only as smart and as responsible as the humans who wield it. 

In the same issue, Victor J. Dzau, President of the National Academy of Medicine shares his perspective on the same topic


Source: PNAS Nexus

21.12.2023

More on the subject:

Related articles

Photo

News • Color adjustment technique

More consistent histopathology slides for machine learning use

In a recent study, researchers proposed a novel technique to help make stained histopathological image datasets more useful for many emerging machine-learning-based classification systems.

Photo

News • Towards personalised risk assessment

Training an AI to early detect dementia with big data

With enough medical training data, AI can predict health conditions with astounding accuracy. Now, researchers want to use brain scans of the entire Scottish population to have an algorithm early…

Photo

News • Machine learning forecast of MS

Multiple sclerosis: predicting progression with AI

Can an AI predict the progression of multiple sclerosis in a patient over the subsequent months and years? Researchers in Belgium used machine learning to do just that – with promising results.

Related products

Subscribe to Newsletter