Lodewijk Brand’s new journal publication, titled “Joint Multi-Modal Longitudinal Regression and Classification for Alzheimer’s Disease Prediction”, has recently been accepted into the IEEE Transactions on Medical Imaging. This work is an extension of the conference version presented at the Medical Image Computing and Computer Assisted Intervention (MICCAI 2018).

In this work the authors expand upon the Joint Multi-Modal Longitudinal Regression and Classification method that is able to identify brain and genetic biomarkers that are predictive of an AD diagnosis. The proposed method performs the classification and regression tasks at the same time and uses three distinct regularization terms to link the two tasks together and capture the temporal and structural relationships within the input data.

The code for the method is provided at: https://github.com/minds-mines/jmmlrc