Extensive T1-weighted MRI Preprocessing Improves Generalizability of Deep Brain Age Prediction Models

Abstract

Brain age is an estimate of chronological age obtained from T1-weighted magnetic resonance images (T1w MRI) and represents a simple diagnostic biomarker of brain ageing and associated diseases. While the current best accuracy of brain age predictions on T1w MRIs of healthy subjects ranges from two to three years, comparing results from different studies is challenging due to differences in the datasets, T1w preprocessing pipelines, and performance metrics used. This paper investigates the impact of T1w image preprocessing on the performance of four deep learning brain age models presented in recent literature. Four preprocessing pipelines were evaluated, differing in terms of registration, grayscale correction, and software implementation. The results showed that the choice of software or preprocessing steps can significantly affect the prediction error, with a maximum increase of 0.7 years in mean absolute error (MAE) for the same model and dataset. While grayscale correction had no significant impact on MAE, the affine registration, compared to the rigid registration of T1w images to brain atlas was shown to statistically significantly improve MAE. Models trained on 3D images with isotropic 1 mm3 resolution exhibited less sensitivity to the T1w preprocessing variations compared to 2D models or those trained on downsampled 3D images. Some proved invariant to the preprocessing pipeline, however only after offset correction. Our findings generally indicate that extensive T1w preprocessing enhances the MAE, especially when applied to a new dataset. This runs counter to prevailing research literature which suggests that models trained on minimally preprocessed T1w scans are better poised for age predictions on MRIs from unseen scanners. Regardless of model or T1w preprocessing used, we show that to enable generalization of models performance on a new dataset with either the same or different T1w preprocessing than the one applied in model training, some form of offset correction should be applied. HighlightsO_LIIntensive preprocessing improves performance of computationally less demanding models. C_LIO_LIModels trained on 1mm3 MRIs are most insensitive to the choice of preprocessing. C_LIO_LIChange in preprocessing increased MAE up to 0.7 years for the same model and dataset. C_LIO_LIPreprocessing software used may impact model performance. C_LIO_LIPrediction bias is systematic across the whole age span and inherent to each model. C_LI