Stabilized Sparse Ordinal Regression for Medical Risk Stratification

TL;DR

This paper introduces a stabilized sparse ordinal regression framework for medical risk stratification using EMR data, addressing challenges of high dimensionality, noise, and stability, and demonstrating superior performance in suicide risk prediction.

Contribution

It presents a novel ordinal regression approach with domain-specific feature interaction networks to improve stability and interpretability in EMR-based risk prediction.

Findings

Outperforms clinicians in suicide risk prediction accuracy

Identifies risk factors consistent with mental health knowledge

Produces more stable models against data resampling

Abstract

The recent wide adoption of Electronic Medical Records (EMR) presents great opportunities and challenges for data mining. The EMR data is largely temporal, often noisy, irregular and high dimensional. This paper constructs a novel ordinal regression framework for predicting medical risk stratification from EMR. First, a conceptual view of EMR as a temporal image is constructed to extract a diverse set of features. Second, ordinal modeling is applied for predicting cumulative or progressive risk. The challenges are building a transparent predictive model that works with a large number of weakly predictive features, and at the same time, is stable against resampling variations. Our solution employs sparsity methods that are stabilized through domain-specific feature interaction networks. We introduces two indices that measure the model stability against data resampling. Feature networks are used to generate two multivariate Gaussian priors with sparse precision matrices (the Laplacian and Random Walk). We apply the framework on a large short-term suicide risk prediction problem and demonstrate that our methods outperform clinicians to a large-margin, discover suicide risk factors that conform with mental health knowledge, and produce models with enhanced stability.