Artificial Intelligence in Health Care: Benefits and Challenges of Machine Learning Technologies for Medical Diagnostics

What GAO Found

Several machine studying (ML) applied sciences can be found in the U.S. to help with the diagnostic course of. The ensuing advantages embrace earlier detection of ailments; extra constant evaluation of medical information; and elevated entry to care, significantly for underserved populations. GAO recognized a spread of ML-based applied sciences for 5 chosen ailments — sure cancers, diabetic retinopathy, Alzheimer’s illness, coronary heart illness, and COVID-19 —with most applied sciences counting on information from imaging corresponding to x-rays or magnetic resonance imaging (MRI). However, these ML applied sciences have usually not been broadly adopted.

Academic, authorities, and personal sector researchers are working to increase the capabilities of ML-based medical diagnostic applied sciences. In addition, GAO recognized three broader rising approaches—autonomous, adaptive, and consumer-oriented ML-diagnostics—that may be utilized to diagnose a spread of ailments. These advances may improve medical professionals’ capabilities and enhance affected person therapies but additionally have sure limitations. For instance, adaptive applied sciences might enhance accuracy by incorporating extra information to replace themselves, however computerized incorporation of low-quality information might result in inconsistent or poorer algorithmic efficiency.

Spectrum of adaptive algorithms

We recognized a number of challenges affecting the event and adoption of ML in medical diagnostics:

Demonstrating real-world efficiency throughout numerous scientific settings and in rigorous research.
Meeting scientific wants, corresponding to creating applied sciences that combine into scientific workflows.
Addressing regulatory gaps, corresponding to offering clear steerage for the event of adaptive algorithms.
These challenges have an effect on varied stakeholders together with expertise builders, medical suppliers, and sufferers, and might gradual the event and adoption of these applied sciences.

GAO developed three coverage choices that might assist handle these challenges or improve the advantages of ML diagnostic applied sciences. These coverage choices establish potential actions by policymakers, which embrace Congress, federal companies, state and native governments, tutorial and analysis establishments, and business. See beneath for a abstract of the coverage choices and related alternatives and issues.

Policy Options to Help Address Challenges or Enhance Benefits of ML Diagnostic Technologies

 
Opportunities
Considerations

Evaluation (reportpage 28)

Policymakers may create incentives, steerage, or insurance policies to encourage or require the analysis of ML diagnostic applied sciences throughout a spread of deployment situations and demographics consultant of the supposed use.

This coverage possibility may assist handle the problem of demonstrating actual world efficiency.

Stakeholders may higher perceive the efficiency of these applied sciences throughout numerous situations and assist to establish biases, limitations, and alternatives for enchancment.
Could inform suppliers’ adoption selections, doubtlessly resulting in elevated adoption by enhancing belief.
Information from evaluations will help inform the choices of policymakers, corresponding to selections about regulatory necessities.

May be time-intensive, which may delay the motion of these applied sciences into {the marketplace}, doubtlessly affecting sufferers and professionals who may benefit from these applied sciences.
More rigorous analysis will probably result in further prices, corresponding to direct prices for funding the research. Developers might not be incentivized to conduct these evaluations if it may present their merchandise in a unfavourable mild, so policymakers may take into account whether or not evaluations needs to be carried out or reviewed by impartial events, in line with business officers.

Data Access (reportpage 29)

Policymakers may develop or increase entry to high-quality medical information to develop and take a look at ML medical diagnostic applied sciences. Examples embrace requirements for accumulating and sharing information, creating information commons, or utilizing incentives to encourage information sharing.

This coverage possibility may assist handle the problem of demonstrating actual world efficiency.

Developing or increasing entry to high-quality datasets may assist facilitate coaching and testing ML applied sciences throughout numerous and consultant situations. This may enhance the applied sciences’ efficiency and generalizability, assist builders perceive their efficiency and areas for enchancment, and assist to construct belief and adoption in these applied sciences.
Expanding entry may allow builders to save lots of time in the event course of, which may shorten the time it takes for these applied sciences to be accessible for adoption.

Entities that personal information could also be reluctant to share them for a quantity of causes. For instance, these entities might take into account their information worthwhile or proprietary. Some entities may additionally be involved in regards to the privateness of their sufferers and the supposed use and safety of their information.
Data sharing mechanisms could also be of restricted use to researchers and builders relying on the standard and interoperability of these information, and curating and storing information might be costly and might require public and personal sources.

Collaboration (reportpage 30)

Policymakers may promote collaboration amongst builders, suppliers, and regulators in the event and adoption of ML diagnostic applied sciences. For instance, policymakers may convene multidisciplinary consultants collectively in the design and improvement of these applied sciences via workshops and conferences.

This coverage possibility may assist handle the challenges of assembly medical wants and addressing regulatory gaps.

Collaboration between ML builders and suppliers may assist be sure that the applied sciences handle scientific wants. For instance, collaboration between builders and medical professionals may assist builders create ML applied sciences that combine into medical professionals’ workflows, and reduce time, effort, and disruption.
Collaboration amongst builders and medical suppliers may assist in the creation and entry of ML prepared information, in line with NIH officers.

As beforehand reported, suppliers might not have time to each collaborate with builders and deal with sufferers; nonetheless, organizations can present protected time for workers to have interaction in innovation actions corresponding to collaboration. 
If builders solely collaborate with suppliers in particular settings, their applied sciences might not be usable throughout a spread of situations and settings, corresponding to throughout completely different affected person sorts or expertise programs.

Source: GAO. | GAO-22-104629

Why GAO Did This Study

Diagnostic errors have an effect on greater than 12 million Americans annually, with combination prices probably in extra of $100 billion, in line with a report by the Society to Improve Diagnosis in Medicine. ML, a subfield of synthetic intelligence, has emerged as a strong instrument for fixing complicated issues in numerous domains, together with medical diagnostics. However, challenges to the event and use of machine studying applied sciences in medical diagnostics increase technological, financial, and regulatory questions.

GAO was requested to conduct a expertise evaluation on the present and rising makes use of of machine studying in medical diagnostics, in addition to the challenges and coverage implications of these applied sciences. This report discusses (1) presently accessible ML medical diagnostic applied sciences for 5 chosen ailments, (2) rising ML medical diagnostic applied sciences, (3) challenges affecting the event and adoption of ML applied sciences for medical prognosis, and (4) coverage choices to assist handle these challenges.

GAO assessed accessible and rising ML applied sciences; interviewed stakeholders from authorities, business, and academia; convened a gathering of consultants in collaboration with the National Academy of Medicine; and reviewed studies and scientific literature. GAO is figuring out coverage choices in this report.

For extra data, contact Karen L. Howard at (202) 512-6888 or [email protected].

https://www.gao.gov/products/gao-22-104629

Recommended For You