JMIR Publications

ABSTRACT

Background:

Despite the promise of machine learning (ML) to inform individualized medical care, the clinical utility of ML in medicine has been limited by the minimal interpretability and “black box” nature of these algorithms.

Objective:

To demonstrate a general framework for generating clinically-relevant and interpretable visualizations of “black box” predictions to aid in the clinical translation of ML.

Methods:

To obtain improved transparency of ML, simplified models and visual displays can be generated based upon the visualization approaches commonly used in clinical practice, such as decision trees and effect plots. We illustrate methods for interpretable ML predictions with a random forest approach applied to data from the Left Ventricular (LV) Structural Predictors of Sudden Cardiac Death (SCD) Registry for individualized risk prediction of SCD, a leading cause of death.

Results:

With the LV Structural Predictors of SCD Registry data, SCD risk predictions can be obtained through ML methods such as random forest. The “black box” predictions can be translated into clinically-relevant and interpretable visualizations through a global summary decision tree and an effect estimate visual display. Several risk factors (namely heart failure hospitalization, cardiac magnetic resonance imaging indices, and serum concentration of systemic inflammation) can be clearly visualized as branch points of a decision tree to discriminate between low, intermediate, and high risk patients.

Conclusions:

We provide approaches to increase the clinical translation of ML approaches through clinician-tailored visual displays of “black box” algorithms. We illustrate this general model-agnostic framework by applying it to SCD risk prediction. Although we illustrate the methods using SCD prediction with RF as an illustrative example, the methods presented are applicable more broadly to improving clinical translation of ML, regardless of the specific ML algorithm or clinical application. Since any trained predictive model can be summarized in a simplified manner with this approach, we encourage the use of simplified visual displays as an adjunct to the complex predictive model. Overall, this framework can allow clinicians to peek inside the “black box” and develop a deeper understanding of the model in order to gain trust in the predictions and confidence in applying them to clinical care. Clinical Trial: Left Ventricular (LV) Structural Predictors of Sudden Cardiac Death (SCD) Registry: ClinicalTrials.gov NCT01076660