Second opinion needed: communicating uncertainty in medical machine learning

By Benjamin Kompa, Jasper Snoek, Andrew Beam. npj Digital Medicine, 2021

463 citations recorded by the source corpus.

There is great excitement that medical artificial intelligence (AI) based on machine learning (ML) can be used to improve decision making at the patient level in a variety of healthcare settings. However, the quantification and communication of uncertainty for individual predictions is often neglected even though uncertainty estimates could lead to more principled decision-making and enable machine learning models to automatically or semi-automatically abstain on samples for which there is high uncertainty. In this article, we provide an overview of different approaches to uncertainty quantification and abstention for machine learning and highlight how these techniques could improve the safety and reliability of current ML systems being used in healthcare settings. Effective quantification and communication of uncertainty could help to engender trust with healthcare workers, while providing safeguards against known failure modes of current machine learning approaches. As machine learning becomes further integrated into healthcare environments, the ability to say “I’m not sure” or “I don’t know” when uncertain is a necessary capability to enable safe clinical deployment.

Subfields: Family Practice, General Health Professions, Health Informatics

View Paper: https://www.semanticscholar.org/paper/e32a2519b59d62cff6cb8136ee242dc3754ed57b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

FUTURE-AI: international consensus guideline for trustworthy and deployable artificial intelligence in healthcare (2023, unknown): “…health organisations, and regulators about the AI model’s properties (eg, hyperparameters), training and testing data, evaluation criteria and results, biases and other limitations, and periodic audits and updates 49-51 ; (c) a publication based on existing AI reporting the bmj | BMJ…”
Deep learning in cancer diagnosis, prognosis and treatment selection (2021, background): “from making predictions and ask for medical experts’ opinion [160].”
Coarse race data conceals disparities in clinical risk score performance (2023, methodology): “…for quantifying uncertainty (Efron and Tibshirani, 1994) which is widely used in medical applications (Mihaylova et al., 2011; Myers et al., 2020; Kompa et al., 2021): for each iteration, we sample datapoints with replacement from the original dataset to produce a “bootstrapped” dataset of the…”
Revisiting the Calibration of Modern Neural Networks (2021, background): “For example, in a medical application, we would like to defer images for which the model makes low-confidence predictions to a physician for review (Kompa et al., 2021).”