Spatially Aware Deep Learning Cuts IVIM MRI Parameter Errors by 81% in Esophageal Cancer

A spatially aware neural network (Spatial NCDE) reduced IVIM parameter estimation error by 81% (D) vs. least squares and 55% vs. UNet in simulations. In esophageal cancer patients, improvement in predicting chemoradiotherapy response was not statistically significant.

• At SNR 20, MSE reductions for D, D*, and f were 81%, 85%, 83% vs. least squares and 55%, 67%, 44% vs. UNet.
• The method is acquisition-independent, accommodating protocol variations.
• Patient-based predictive ability for neoadjuvant chemoradiotherapy response improved but did not reach statistical significance.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.