Noninvasive quantitative ultrasound fatty liver evaluation of hepato-renal index in pediatric patients using 3D-slicer

Abstract

Abstract Aim Hepato-renal index (HRI) calculated by ultrasound images has been shown to be an effective, noninvasive tool to screen patients with steatosis. The aim of this study was to non-invasively explore a new method for the calculation, directly on DICOM images, of HRI in pediatric patients using 3D Slicer, a free and open-source software for medical image analysis, especially used for artificial intelligence data annotation. Previous studies in literature were based on non-medical image format (such as png and tiff) and analyzed by using ImageJ, a popular image analysis software. Materials and methods DICOM images were collected retrospectively between November 2022 and December 2023 at a tertiary institution on any patient under 18 years referred to US assessment of suspected steatosis. The HRI was measured on sagittal images with a clear visualization of both the liver and the kidney by 3D Slicer. The HRI was calculated as the ratio of average pixel intensity values between the two ROIs. Also, we correlated the index with the qualitative operator assessment grade of steatosis made by two expert radiologists with more than 15 years of experience. Results 49 patients (59% male, age 11.6 ± 2.6 year) were recruited. Of these, 29 (59%) had qualitative ultrasound confirmation of the diagnosis. The Bland–Altman plots showed a good agreement between the HRI indexes calculated with the standard tool ImageJ and 3D Slicer. Furthermore, the quantitative HRI calculated by the two software showed a correlation (Spearman’s coefficient = 0.52, P < 0.0001) with the qualitative operator measurement. Conclusions For the first time, an estimation of HRI directly on DICOM images was executed by using 3D Slicer, obtained quantitative information directly from clinically approved image format.