Improving Clinical Decision Support through Interpretable Machine Learning and Error Handling in Electronic Health Records

TL;DR

Trust-MAPS embeds clinical domain knowledge into EMR data processing by projecting data onto physiologically feasible constraints and measuring deviations from healthy homeostasis as trust-scores. This two-step projection (physical then normal) both corrects data errors and provides interpretable, cluster-consistent features that feed into a cluster-then-predict ML pipeline (SMOTE-assisted XGBoost). On the PhysioNet 2019 sepsis dataset, Trust-MAPS yields a substantial AUROC improvement (~0.91 six hours prior) and enhanced precision, with trust-scores consistently ranking among top predictors and offering clinically meaningful explanations. The approach reduces bias from erroneous data, improves model calibration, and acts as a modular preprocessing tool that can generalize to other health-care decision-support tasks and beyond.

Abstract

The objective of this work is to develop an Electronic Medical Record (EMR) data processing tool that confers clinical context to Machine Learning (ML) algorithms for error handling, bias mitigation and interpretability. We present Trust-MAPS, an algorithm that translates clinical domain knowledge into high-dimensional, mixed-integer programming models that capture physiological and biological constraints on clinical measurements. EMR data is projected onto this constrained space, effectively bringing outliers to fall within a physiologically feasible range. We then compute the distance of each data point from the constrained space modeling healthy physiology to quantify deviation from the norm. These distances, termed "trust-scores," are integrated into the feature space for downstream ML applications. We demonstrate the utility of Trust-MAPS by training a binary classifier for early sepsis prediction on data from the 2019 PhysioNet Computing in Cardiology Challenge, using the XGBoost algorithm and applying SMOTE for overcoming class-imbalance. The Trust-MAPS framework shows desirable behavior in handling potential errors and boosting predictive performance. We achieve an AUROC of 0.91 (0.89, 0.92 : 95% CI) for predicting sepsis 6 hours before onset - a marked 15% improvement over a baseline model trained without Trust-MAPS. Trust-scores emerge as clinically meaningful features that not only boost predictive performance for clinical decision support tasks, but also lend interpretability to ML models. This work is the first to translate clinical domain knowledge into mathematical constraints, model cross-vital dependencies, and identify aberrations in high-dimensional medical data. Our method allows for error handling in EMR, and confers interpretability and superior predictive power to models trained for clinical decision support.

Figures (23)

• Figure 1: Trust-MAPS augment the "corrected data" computed using projections onto physical constraints $P$, with "trust-scores" which capture the distance of the corrected data from homeostasis constraints. This augmented data is then used in downstream ML for predictions, and significantly improves the performance. There are a number of steps needed to normalize the length of the stay of patients, impute missing values in the data to get the new patient data in standardized form, and then data augmentation with trust-scores for each patient vital (See Figure \ref{['fig:pipeline']} in the Supplemental \ref{['sec:methods']}), before using projections. The pipeline is very general, and can be used for more advanced predictive methods (e.g., those that process clinical notes).
• Figure 2: Example of Trust-MAPS on Respiratory Rate
• Figure 3: Example of Trust-MAPS on FiO2
• Figure 5: An illustrative example of data correction by projection on the Physical Constraints Set $P$ demonstrated through (a) the constraint on the relationship between MAP, DBP and SBP, where the imputed values of DBP default to the average value but the projected values satisfy the interactions between these vitals given by constraint #4 in Supplemental Section \ref{['sec:phys proj']}, (b) A comparison of imputation of Base Excess by methods described in the Supplemental Section \ref{['app:alt_imputation']}. The Trust-MAPS approach is labeled "Linear Imputation and Projections." (c) A comparison of imputation of temperature.
• Figure 6: Left: A plot comparing Receiver-Operating Characteristic Curves for sepsis prediction machine learning model on the dataset processing with each step of the Trust-MAPS process, and baseline models trained without Trust-MAPS. Right: A plot comparing Precision-Recall Curves for sepsis prediction machine learning model on the dataset processing with each step of the Trust-MAPS process, and baseline models trained without Trust-MAPS.