"The greater the degree of hepatic fibrosis the more likely the patient will develop liver cancer," said Dr. Hiroki Kato, lead author of the study.
A radiologist duo reviewed the MR images of 52 patients, and compared their diagnoses to the data that was provided by the CAD system. The computer algorithm, when compared to the review of the images by two radiologists, proved to be more accurate in diagnosing liver fibrosis.
"If such algorithms are efficiently incorporated into routine clinical practice, they will increase radiologists' confidence in the diagnosis of hepatic fibrosis. Successful clinical introduction of a CAD system will obviate the need for more invasive examinations such as biopsy in the diagnosis of hepatic fibrosis," said Dr. Kato.
"It was surprising for us to find that the CAD system was superior to the radiologists in the detection and grading of hepatic fibrosis. The radiologists were experienced in diagnosing hepatic fibrosis so we assumed that the diagnostic performance by radiologists would be comparable or superior to the CAD system," he added.
The new findings have been reported in the journal American Journal of Roentgenology, published by the American Roentgen Ray Society.