Abstract #2279

Fast and Robust Detection of Fetal Brain in MRI using Transfer Learning based FCN

DEFENG WANG^1,2, Jinpeng Li², YISHAN LUO^2,3, DANTONG MIAO⁴, XIN ZHANG⁴, Queenie Chan⁵, Winnie CW CHU², LIN SHI^6,7,8, and BING ZHANG⁴

¹Shenzhen Research Institute, The Chinese University of Hong Kong, Shenzhen, People's Republic of China, ²Department of Imaging and Interventional Radiology, The Chinese University of Hong Kong, Hong Kong, Hong Kong, ³Shenzhen Research Institute,The Chinese University of Hong Kong, Shenzhen, People's Republic of China, ⁴Department of Radiology, Nanjing Drum Tower Hospital,The Affiliated Hospital of Nanjing University Medical School, Nanjing, People's Republic of China, ⁵Philips Healthcare, Hong Kong, ⁶Chow Yuk Ho Technology Center for Innovative Medicine, Hong Kong, ⁷Therese Pei Fong Chow Research Centre for Prevention of Dementia, ⁸Department of Medicine and Therapeutics, The Chinese University of Hong Kong, Hong Kong

In this work, we proposed a transfer learning based FCN method which can automatically detect the fetal brain in MRI. We used the off-the-shelf model weights trained on nature images to initialize a fully connected network (FCN), and then fine-tuned the model on the fetal MRIs. We tested our method on two datasets with different MRI sequences, and the results demonstrated that the proposed method is automatic, fast and robust for detection of fetal brain in MRI.

This abstract and the presentation materials are available to members only; a login is required.

Join Here