Meeting Banner
Abstract #3723

Robust Reproducible Semi-Automated Perfusion-Diffusion Mismatch Assessment in Acute Ischemic Stroke Setting

Venkata Veerendranadh Chebrolu1, Dattesh D. Shanbhag2, Arun Govinda Rao3, Patrice Hervo4, Marc-Antoine Labeyrie5, Catherine Oppenheim5, 6, Rakesh Mullick2

1Medical Image Analysis Lab, GE Global Research, Bangalore , Karnataka, India; 2Medical Image Analysis Lab, GE Global Research, Bangalore, Karnataka, India; 3GE Healthcare, Bangalore, Karnataka, India; 4GE Healthcare, Buc, France; 5Departments of Radiology and Neurology, Centre Hospitalier, Sainte-Anne, Paris, France; 6Universit Paris Descartes , Paris, France

In semi-automated stroke lesion segmentation based on user defined seed inputs, the location and shape of the input could vary. We developed robust and reproducible semi-automated DWI and PWI lesion segmentation algorithms and evaluated their performance and reproducibility in assessing perfusion-diffusion mismatch in a cohort of acute ischemic stroke patients. Repeated measures ANOVA did not show any statistically significant differences between ground-truth lesion volumes and those obtained using the semi-automated methods (p < 0.05). Mismatch agreement was achieved in 88% of the cases with a kappa (&#954;) of 0.766.

Abstract PDF

Keywords

according achieved acute agreement analyze another applying approaches assessment audience automated background belonged blood brain bring cause central cerebellum cerebral cerebrum channel classes connected considered criteria death decision defined demonstrating described develop diffusion disability discarded distance done evaluate example excellent expected extremity fast feedback filter final functionality generated generation global good greater ground guidelines head hence highest histogram hours hypo included infarct infarction inputs intensity inter interaction intervention intra kappa kinds known lesion lesions location manual manually marking matching maximizes measured measures medical million minutes mismatch needed neurology observer onset overall overlap partitioned patients people performance perfusion physicists plots positive probable processed proposed proximity psychiatry radiologist radiologists radiology registering removed repeated reported reproducibility reproducible residue risk robust salvageable scanner seed segmentation segmentations segments selected semi senior shape slice started statistical stroke successfully suffering summary support target threshold timely tool transform treatment truth truths user vary ventricle ventricular volume volumes whisker year