Decoding Pain: Uncovering the Factors that Affect Performance of Neuroimaging-Based Pain Models

Abstract

AO_SCPLOWBSTRACTC_SCPLOWNeuroimaging-based pain biomarkers, when combined with machine learning techniques, have demonstrated potential in decoding pain intensity and diagnosing clinical pain conditions. However, a systematic evaluation of how different modeling options affect model performance remains unexplored. This study presents results from a comprehensive literature survey and benchmarking analysis. We conducted a survey of 57 previously published articles that included neuroimaging-based predictive modeling of pain, comparing classification and prediction performance based on the following modeling variables--the levels of data, spatial scales, idiographic vs. population models, and sample sizes. The findings revealed a preference for population-level modeling with brain-wide features, aligning with the goal of clinical translation of neuroimaging biomarkers. However, a systematic evaluation of the influence of different modeling options was hindered by a limited number of independent test results. This prompted us to conduct benchmarking analyses using a locally collected functional Magnetic Resonance Imaging (fMRI) dataset (N = 124) involving an experimental thermal pain task. The results clearly demonstrated that data levels, spatial scales, and sample sizes significantly impact model performance. Specifically, incorporating more pain-related brain regions, increasing sample sizes, and using less data averaging in training while increasing it in testing enhanced performance. These findings provide a useful reference for decision-making in the development of neuroimaging-based biomarkers, highlighting the importance of the careful selection of modeling strategies and options to build better-performing neuroimaging pain biomarkers. These findings offer useful guidance for developing neuroimaging-based biomarkers, underscoring the importance of strategic selection of modeling approaches to build better-performing neuroimaging pain biomarkers.