Congress:
ECR24
Poster Number:
C-16188
Type:
EPOS Radiologist (scientific)
Authorblock:
J. F. Heidenreich1, J. Gödiker2, B. Schlevogt3, P. Schindler2; 1Würzburg/DE, 2Muenster/DE, 3Osnabrück/DE
Disclosures:
Julius Frederik Heidenreich:
Nothing to disclose
Juliane Gödiker:
Nothing to disclose
Bernhard Schlevogt:
Nothing to disclose
Philipp Schindler:
Nothing to disclose
Keywords:
Abdomen, Artificial Intelligence, Liver, MR, MR-Cholangiography, Segmentation, Genetic defects
Multiple biliary hamartomas (MBH), also known as "von Meyenburg complexes", are benign malformations characterized by dilated bile ducts embedded in fibrous stroma, usually less than 10 mm in diameter. MBH are part of a wider spectrum of ductal plate malformations, which are generally presumed to have a genetic basis. However, the specific genetic cause of MBH is unknown. MBH are usually incidental findings in liver imaging and accurate identification is critical as they can be mistaken for malignant growths. This study aims to determine whether magnetic resonance imaging (MRI) combined with machine learning-supported post-processing can improve the detection of MBHs and also uncover their genetic basis.