Keywords:
Tissue characterisation, Technology assessment, Segmentation, Computer Applications-Detection, diagnosis, Image manipulation / Reconstruction, CT, Liver, Computer applications
Authors:
A. P. Gonchar, A. Elizarov, N. S. Kulberg, V. Gombolevskiy, I. Blokhin, T. I. Alekseeva, A. Krysanova, M. Suleymanova, D. A. Chernyshev; Moscow/RU
DOI:
10.26044/ecr2019/C-3001
Aims and objectives
The frequency of moderate-to-severe hepatic steatosis in an asymptomatic population increases annually [1].
Today non-contrast computed tomography (CT) provides fast,
reproducible and noninvasive hepatic steatosis quantification [2,3].
Liver segmentation can be done by semi-automatic and automatic methods.
Semi-automatic methods can require a large amount of radiologist’s time for interactive segmentation [4,
5].
Attenuation-based algorithms,
various clustering methods,
pattern recognition,
and neural networks are used for automated liver segmentation [6-10].
Despite the significant advances in this area,
available algorithms have some weaknesses,
mainly the lack of accuracy and reliability and vendor dependency.
Thus,
at the moment no vendor-agnostic software allows for fast automated liver segmentation in a large number of cases in routine practice.
The purpose of this study is to develop a program for automatic liver density measurement in CT and low dose computed tomography (LDCT).
We utilized data from LDCT lung cancer screening trial in Moscow,
Russia.
The endpoint was accurate mean liver density measurement in Hounsfield units (HU).