This study looked at how different MRI techniques find spots in the liver. Researchers used a new method called Deep Learning Reconstruction and compared it to a standard, traditional scan method. They looked at scans from 53 patients who were in the hepatobiliary phase of the imaging process.
The main finding was that the deep learning method found more small lesions between 5 and 10 millimeters in size. The study reported a statistical difference with a p-value of 0.030 for these small spots. However, for larger lesions over 10 millimeters, both methods performed the same. There were no differences found for high-signal lesions either.
The researchers did not report any safety concerns or side effects from using the new scan technology. Because this was a single-center study with only 53 patients, the results come from a limited group. Readers should understand that this is early evidence from one location. More research is needed to see if these findings hold true in other hospitals or with larger groups of people.