Aims and objectives
Hepatosteatosis refers to the abnormal accumulation of triglycerides in more than 5% of hepatocytes (Lee DH,
2017; Torres,
Williams,
& Harrison,
2012).
Significant etiologies of hepatosteatosis include consumption of alcohol and Non-Alcoholic Fatty Liver Disease (NAFLD) (Miyoshi & Kamada,
2015).
NAFLD is a spectrum ranging from benign Isolated Fatty Liver (IFL) to Non-Alcoholic Steatohepatitis (NASH),
which involves hepatocellular inflammation and the potential to develop fibrosis,
cirrhosis and hepatocellular carcinoma (Torres et al,
2012).
NAFLD and NASH have prevalences of 46% and 12.2% respectively among asymptomatic...
Methods and materials
Study Design:
Patients were selected and a database was created of unenhanced and Portal Venous CT images through selection of mild,
moderate and severe cases of hepatic steatosis as defined by the unenhanced phase appearance.
Normal control cases of normal density healthy livers were used for each of the unenhanced CT and portal venous CT groups.
The unenhanced CT phase images were used as an internal reference standard for hepatic steatosis.
1.5 cm2 Regions of Interest (ROI) were taken from hepatic segments V,
VI,
VII,...
Results
Evaluation of 3D-CNN segmentation on the held-out local Test set yielded a Mean Average Error (MAE) of 2.55 HU for the auto-segmented Median PVP density compared to the manual ROI PVP density (Fig. 3).
When evaluating the True versus the Simulated Plain Density on the Training dataset,
the linear regression model demonstrated a R2 of 0.873,
RMSE of 6.979,
and MAE of 5.82HU.
The full end-to-end algorithm was evaluated on the 40 Test cases,
which yielded a Simulated Average Plain Density MAE of 4.78HU.
Conclusion
In our study,
a hybrid 3D-CNN/statistical approach is able to predict Average Plain Density of a liver from contrasted examinations at a level comparable with that of human radiologist’s assessment to within 4-5HU from standardized measurements of Uncontrasted CT.
The use of Plain CT as an internal reference standard provided sufficient data for training both Unenhanced and PV DCNN’s to identify hepatic steatosis to the degree possible from Unenhanced CT.
Of note,
this task is performed in a fully-automated fashion with zero user interaction from...
References
Hamer,
O.
W.,
Aguirre,
D.
A.,
Casola,
G.,
Lavine,
J.
E.,
Woenckhaus,
M.,
& Sirlin,
C.
B.
(2006).
Fatty liver: imaging patterns and pitfalls.
Radiographics,
26(6),
1637-1653.
Lee,
D.
H.
(2017).
Imaging evaluation of non-alcoholic fatty liver disease: focused on quantification.
Clinical and molecular hepatology,
23(4),
290.
Lee,
S.
S.,
& Park,
S.
H.
(2014).
Radiologic evaluation of nonalcoholic fatty liver disease.
World journal of gastroenterology: WJG,
20(23),
7392.
McFadden S,
Roding T,
de Vries G,
Benwell M,
Bijwaard H,
Scheurleer J.
Digital imaging and...