Meeting Banner
Abstract #3133

Multi-stage ensemble machine learning for predicting the pathology of thyroid micronodules on small-datasets high b-value thyroid DWI

ChengLong Deng1,2, BingChao Wu1,2, QingJun Wang3, QingLei Shi4, Bei Guan1,2, Dacheng Qu5, ChenXi Li1,2, DaoGuang Zan1,2, XiaoLin Chen1,2, and YongJi Wang1,2
1Collaborative Innovation Center, Institute of Software Chinese Academy of Sciences, Beijing, China, 2University of Chinese Academy of Sciences, Beijing, China, 3Department of Radiology, The 6th Medical Center of Chinese PLA General Hospital, Beijing, China, 4MR Scientific Marketing, Siemens Healthcare, Beijing, China, 5School of Computer Science and Technology, Beijing Institute of Technology, Beijing, China


In this paper, a multi-stage ensemble learning based on the majority voting mechanism was designed to leverage the contradiction between an insufficient number of thyroid MRI and well-trained deep learning models that accurately predicted the pathology of thyroid micronodules. And its clinical applicability value was also assessed in terms of micronodule risk stratification and optimal regimen selection on high b-value (2000 s/mm2) diffusion-weighted images. Experimental results proved that our model had the capability of effectively distinguishing benign and malignant micronodules on small-dataset thyroid DWI images.

This abstract and the presentation materials are available to members only; a login is required.

Join Here