Borrowing From the Future: Enhancing Early Risk Assessment through Contrastive Learning

TL;DR

This paper introduces Borrowing From the Future (BFF), a contrastive learning framework that leverages later-stage data to improve early-stage risk predictions in pediatric health assessments.

Contribution

The study presents a novel contrastive multi-modal approach that uses future information to enhance early risk assessment accuracy.

Findings

BFF improves early risk prediction performance across multiple pediatric tasks.

The framework effectively borrows signals from future stages to inform earlier assessments.

Results show consistent accuracy gains in real-world datasets.

Abstract

Risk assessments for a pediatric population are often conducted across multiple stages. For example, clinicians may evaluate risks prenatally, at birth, and during Well-Child visits. Although predictions made at later stages typically achieve higher precision, it is clinically desirable to make reliable risk assessments as early as possible. Therefore, this study focuses on improving prediction performance in early-stage risk assessments. Our solution, \textbf{Borrowing From the Future (BFF)}, is a contrastive multi-modal framework that treats each time window as a distinct modality. In BFF, a model is trained on all available data throughout the time while performing a risk assessment using up-to-date information. This contrastive framework allows the model to ``borrow'' informative signals from later stages (e.g., Well-Child visits) to implicitly supervise the learning at earlier stages (e.g., prenatal/birth stages). We validate BFF on two real-world pediatric outcome prediction tasks, demonstrating consistent improvements in early risk assessments. The code is available at https://github.com/scotsun/bff.