Study Data from Center for Devices and Radiological Health Update Understanding of Machine Learning (Evaluating Machine Learning-Based MRI Reconstruction Using Digital Image Quality Phantoms).

A recent study conducted by the Center for Devices and Radiological Health explores the evaluation of machine learning-based magnetic resonance imaging (MRI) reconstruction methods. The researchers propose the use of digital image quality phantoms and automated evaluation methods to assess the performance of these methods. The study demonstrates the utility of this evaluation pipeline by assessing an example machine learning-based reconstruction model. The results indicate that training data acquired with a lower undersampling factor and larger anatomical coverage yield better performance. This comprehensive and standardized pipeline can contribute to a better understanding of the performance and guide future development of machine learning-based reconstruction algorithms. [Extracted from the article]