Training Strategies for Radiology Deep Learning Models in Data-limited Scenarios.

Data-driven approaches have great potential to shape future practices in radiology. The most straightforward strategy to obtain clinically accurate models is to use large, well-curated and annotated datasets. However, patient privacy constraints, tedious annotation processes, and the limited availability of radiologists pose challenges to building such datasets. This review details model training strategies in scenarios with limited data, insufficiently labeled data, and/or limited expert resources. This review discusses strategies to enlarge the data sample, decrease the time burden of manual supervised labeling, adjust the neural network architecture to improve model performance, apply semisupervised approaches, and leverage efficiencies from pretrained models.

Informations de copyright

2021 by the Radiological Society of North America, Inc.

Déclaration de conflit d'intérêts

Disclosures of Conflicts of Interest: S.C. No relevant relationships. X.V.N. Equity ownership in and dividends from multiple publicly traded companies that may be considered “broadly relevant to artificial intelligence” (NVIDIA, Amazon, Microsoft, AMD, Apple). L.R.F. Patents issued, no royalties: Radiographic marker that displays upright angle on portable radiographs (Patent no. 9,541,822 B2) and multiscale universal CT window (Patent no. 8,406,493 B2); royalties from Springer (combat radiology, unrelated); research agreement with Philips. L.M.P. associate editor of Radiology: Artificial Intelligence.