Answers are grounded in this paper and cite their sources.
A deep learning based DCE-MRI analysis method was proposed with a dedicated neural network architecture and data generation framework. The proposed method does not need DCE-MRI data acquisition or annotation for training. Compared to conventional non-linear least square (NLLS) fitting methods, the proposed method significantly reduced the average processing time from hours to few minutes while preserved the estimation quality.
No takes yet. Share an insight, caveat, or question.
Zhao et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: