Machine learning models are becoming commonplace in the domain of medical imaging, and with these methods comes an ever-increasing need for more data. However, to preserve patient anonymity it is frequently impractical or prohibited to transfer protected health information (PHI) between institutions. Additionally, due to the nature of some studies, there may not be a large public dataset available on which to train models. To address this conundrum, we analyze the efficacy of transferring the model itself in lieu of data between different sites. By doing so we accomplish two goals: 1) the model gains access to training on a larger dataset that it could not normally obtain and 2) the model better generalizes, having trained on data from separate locations. In this paper, we implement multi-site learning with disparate datasets from the National Institutes of Health (NIH) and Vanderbilt University Medical Center (VUMC) without compromising PHI. Three neural networks are trained to convergence on a computed tomography (CT) brain hematoma segmentation task: one only with NIH data, one only with VUMC data, and one multi-site model alternating between NIH and VUMC data. Resultant lesion masks with the multi-site model attain an average Dice similarity coefficient of 0.64 and the automatically segmented hematoma volumes correlate to those done manually with a Pearson correlation coefficient of 0.87, corresponding to an 8% and 5% improvement, respectively, over the single-site model counterparts.

Date Published: 15 March 2019
PDF: 8 pages
Proc. SPIE 10949, Medical Imaging 2019: Image Processing, 109490A (15 March 2019); doi: 10.1117/12.2511997

Published in SPIE Proceedings Vol. 10949:
Medical Imaging 2019: Image Processing
Elsa D. Angelini; Bennett A. Landman, Editor(s)