Evaluating glioma growth predictions as a forward ranking problem

TL;DR

This paper introduces a novel framework for evaluating glioma growth predictions by framing it as a ranking problem, emphasizing spatial infiltration patterns and separating model fit from predictive performance.

Contribution

It proposes a ranking-based evaluation method using average precision, providing a new perspective on assessing tumor growth models beyond traditional segmentation accuracy.

Findings

Better model fit does not always lead to improved predictive performance.

Ranking-based evaluation offers a comprehensive assessment of growth predictions.

The framework separates model fitting from future prediction accuracy.

Abstract

The problem of tumor growth prediction is challenging, but promising results have been achieved with both model-driven and statistical methods. In this work, we present a framework for the evaluation of growth predictions that focuses on the spatial infiltration patterns, and specifically evaluating a prediction of future growth. We propose to frame the problem as a ranking problem rather than a segmentation problem. Using the average precision as a metric, we can evaluate the results with segmentations while using the full spatiotemporal prediction. Furthermore, by separating the model goodness-of-fit from future predictive performance, we show that in some cases, a better fit of model parameters does not guarantee a better the predictive power.