Year: 2017
Type: preprint
Abstract: Deep learning has become a promising approach for automated medical diagnoses. When medical data samples are limited, collaboration among multiple institutions is necessary to achieve high algorithm performance. However, sharing patient data often has limitations due to technical, legal, or ethical concerns. In such cases, sharing a deep learning model is a more attractive alternative. The best method of performing such a task is unclear, however. In this study, we simulate the dissemination of learning deep learning network models across four institutions using various heuristics and compare the results with a deep learning model trained on centrally hosted patient data. The heuristics investigated include ensembling single institution models, single weight transfer, and cyclical weight transfer. We evaluated these approaches for image classification in three independent image collections (retinal fundus photos, mammography, and ImageNet). We find that cyclical weight transfer resulted in a performance (testing accuracy = 77.3%) that was closest to that of centrally hosted patient data (testing accuracy = 78.7%). We also found that there is an improvement in the performance of cyclical weight transfer heuristic with high frequency of weight transfer.
Source: arXiv (Cornell University)
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Funders: Massachusetts General Hospital, Centre d'Imagerie BioMédicale, National Institutes of Health, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, National Institute on Drug Abuse, National Cancer Institute, National Institute of Biomedical Imaging and Bioengineering
Awards: G1360098495, Informatics Tools for Optimized Imaging Biomarkers for Cancer Research&Discovery, Quantitative Image Informatics for Cancer Research (QIICR), G3171611343, Advanced Multimodal Neuroimaging Training Grant, Quantitative MRI of Glioblastoma Response, Advanced Multimodal Neuroimaging Training Grant, G6205396609, Center for Functional Imaging Technologies, Qualification and Deployment of Imaging Biomarkers of Cancer Treatment Response, Computing, Optimizing, and Evaluating Quantitative Cancer Imaging Biomarkers
OpenAlex ID: W2754259187