Effects of Different Scan Projections on the Quantitative Ultrasound-Based Evaluation of Hepatic Steatosis.
Laura De RosaAntonio SalvatiFerruccio BoninoMaurizia Rossana BrunettoFrancesco FaitaPublished in: Healthcare (Basel, Switzerland) (2022)
Non-alcoholic fatty liver disease (NAFLD) is becoming a global public health issue and the identification of the steatosis severity is very important for the patients' health. Ultrasound (US) images of 214 patients were acquired in two different scan views (subcostal and intercostal). A classification of the level of steatosis was made by a qualitative evaluation of the liver ultrasound images. Furthermore, an US image processing algorithm provided quantitative parameters (hepatic-renal ratio (HR) and Steato-score) designed to quantifying the fatty liver content. The aim of the study is to evaluate the differences in the assessment of hepatic steatosis acquiring and processing different US scan views. No significant differences were obtained calculating the HR and the Steato-score parameters, not even with the classification of patients on the basis of body mass index (BMI) and of different classes of steatosis severity. Significant differences between the two parameters were found only for patients with absence or mild level of steatosis. These results show that the two different scan projections do not greatly affect HR and the Steato-score assessment. Accordingly, the US-based steatosis assessment is independent from the view of the acquisitions, thus making the subcostal and intercostal scans interchangeable, especially for patients with moderate and severe steatosis.
Keyphrases
- public health
- deep learning
- end stage renal disease
- body mass index
- insulin resistance
- computed tomography
- high fat diet
- ejection fraction
- newly diagnosed
- healthcare
- machine learning
- peritoneal dialysis
- prognostic factors
- high fat diet induced
- high resolution
- type diabetes
- risk assessment
- early onset
- patient reported outcomes
- skeletal muscle
- convolutional neural network
- mental health
- physical activity
- climate change
- mass spectrometry
- global health