Hieu Trung Huynh PhD, Presenter: Nothing to Disclose

Kenji Suzuki PhD, Abstract Co-Author: Nothing to Disclose

Ibrahim Karademir MD, Abstract Co-Author: Nothing to Disclose

Rony Kampalath MD, Abstract Co-Author: Nothing to Disclose

Aytekin Oto MD, Abstract Co-Author: Honorarium, Koninklijke Philips Electronics NV Research Grant, Koninklijke Philips Electronics NV Research Grant, Bayer AG Research Grant, Visualase Inc Research Grant, General Electric Company

Liver segmentation is challenging due to the intensity similarity between liver and other organs. Our purpose was to develop an accurate automated 3D liver segmentation scheme for measuring liver volumes in MRI.

Our scheme for MRI liver volumetry consisted of five steps. First, an anisotropic diffusion smoothing filter was applied to T1-weighted MR images of the liver in the portal-venous-phase to reduce noise while preserving the liver structure. An edge enhancer and a nonlinear gray-scale converter were applied to enhance the liver boundary. This boundary-enhanced image was used as a speed function for a 3D fast-marching algorithm to generate an initial surface that roughly approximated the liver shape. A 3D geodesic-active-contour segmentation algorithm refined the initial surface so as to more precisely fit the liver boundary. The liver volume was calculated based on the refined liver surface. Our database consisted of 11 cases of hepatic MRI scans with slice thickness and spacing of 5 and 2.5 mm, respectively. The automated MRI liver volumetry was compared with those manually traced by a radiologist, used as “gold standard.”

The mean liver volume obtained by our automated scheme was 1707 ± 369 cc, whereas the mean manual volume was 1731 ± 360 cc. The two volumetrics reached an excellent agreement (intra-class correlation coefficient was 0.98) without statistical significance (p=0.18). The average accuracy and Dice overlap coefficient were 99.4 ± 0.15% and 0.94 ± 0.02, respectively. The mean processing time for our automated scheme was 1.03 ± 0.09 min (CPU: Intel, Xeon, 2.66 GHz), whereas that for manual volumetry was 27.7 ± 3.5 min (p<0.001).

The MR liver volumetrics based on our automated scheme agreed excellently with “gold-standard” manual volumetrics, and required substantially less completion time.

Our automated scheme provides an efficient and accurate method of measuring liver volumes on MR images; thus, it would be useful for radiologists in their liver volume measurement.

Huynh, H, Suzuki, K, Karademir, I, Kampalath, R, Oto, A, MRI Liver Volumetry Using 3D Geodesic Active Contour Segmentation Coupled with a Level-set Algorithm.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12043896.html