Meeting Banner
Abstract #3535

Planar Hip Cartilage Quality Maps a Novel Approach to 3D Cartilage Assessment by Combining DGEMRIC with Automated Segmentation

Carl Siversson1, 2, Alireza Akhondi-Asl1, Young-Jo Kim3, Simon K. Warfield1

1Computational Radiology Laboratory, Children's Hospital Boston, Harvard Medical School, Boston, MA, United States; 2Dept. of Medical Radiation Physics, Lund University, Malmo, Sweden; 3Department of Orthopaedics, Children's Hospital Boston, Harvard Medical School, Boston, MA, United States

Delayed Gadolinium enhanced MRI of cartilage (dGEMRIC) is a technique for assessing cartilage quality using T1 quantification. In this work it is shown that by combining an isotropic 3D T1 map with an isotropic 3D TrueFISP image it is both feasible and robust to apply fully automated cartilage segmentation to hip-dGEMRIC data. It is also shown that it is possible to apply cartographic algorithms to the segmented cartilage, generating planar T1 maps covering the entire cartilage surface. Such planar maps are very suitable for studying patterns of cartilage degeneration.

Keywords

accurate achieve additional administration analyzed applications applied around assessing assessment automated automatic automatically available azimuthal care cartilage cartographic children clinical coefficient combination combining compensate comprising computational consuming cover degeneration delayed dept developed development dice disease done dose double engineering enhanced entire epidemiological equal evaluation excellent excluded extracting feasible femur finally fovea full fusion gadolinium generated generating hand hospital immediately impingement important include instead intravenous intuitive isotropic label laboratory leave local manual manually maps matrix measured medical minutes modern molecular movement nine novel outside part patient patients patterns performance physics planar possibilities practical predictive previous projection quality quantification quantitative radially radiation radiology receiving recent registered registration related representative retrospectively rigid robust role scanner school segmentation segmentations segmented selection several sphere staple still structure studies subjects surgery systems table target tissue training transformed unfold utilize utilized verified volume volumes years young