Litcius/Paper detail

Ultrasound Radiomics Effective for Preoperative Identification of True and Pseudo Gallbladder Polyps Based on Spatial and Morphological Features

Haixia Yuan, Qihui Yu, Yanqun Zhang, Qing Yu, Qi Zhang, Wen‐Ping Wang

2020Frontiers in Oncology21 citationsDOIOpen Access PDF

Abstract

contributed equally to this article.Purpose To explore the value of ultrasound radiomics in the preoperative identification of gallbladder true-and pseudo-polyps and evaluate the diagnostic accuracy.In total 99 pathological proven gallbladder polyps in 96 patients were enrolled in this study, including 58 cholesterol polyps in 55 patients and 41 gallbladder tubular adenomas in 41 patients. Preoperative ultrasound images were acquired including spatial and morphological features were extracted from each lesion, then two stages of feature selection methods were consecutively adopted by using Fisher inter-intra class variance ratios and Z-scores of these high-dimensional features, so as to select a few intrinsic and important features for differential diagnosis with the support vector machine.Results A total of 80 radiomics features including spatial and morphological features were extracted from each polyp. Eight intrinsic features were selected after the twostage feature selection. Cont14 and Entr6 in the cholesterol polyps group were significantly higher than those in gallbladder adenomas group (4.063±1.682 vs.Ultrasound radiomics analysis based on the spatial and morphological features extracted from ultrasound images could effectively contribute to preoperative diagnosis of gallbladder ture-and pseudo-polyps, which is valuable for clinical management of gallbladder polyps.

Topics & Concepts

RadiomicsMedicineUltrasoundGallbladderRadiologyIdentification (biology)UltrasonographyGeneral surgerySurgeryBiologyBotanyRadiomics and Machine Learning in Medical ImagingAdvanced X-ray and CT ImagingMRI in cancer diagnosis
Ultrasound Radiomics Effective for Preoperative Identification of True and Pseudo Gallbladder Polyps Based on Spatial and Morphological Features | Litcius