Texture Analysis to Differentiate Malignant Renal Tumors in Children Using Gray-Scale Ultrasonography Images

Hyun Joo Shin, Jin Young Kwak, Eunjung Lee, Mi Jung Lee, Haesung Yoon, Kyunghwa Han, Myung Joon Kim

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)

Abstract

We assessed the feasibility of texture analysis to differentiate Wilms tumor, clear cell sarcoma and rhabdoid tumor of the kidney in children using gray-scale ultrasonography images. Children who had pre-operative renal ultrasonography images of the three tumors from January 2002 to February 2017 were retrospectively included as the test set, and children with the same criteria from March 2017 to December 2018 were included as the validation set. From histogram and second-order statistics, features were compared between the tumors, and diagnostic performances were assessed. Among a total of 32 children (24 children with Wilms tumors, five children with clear cell sarcomas and three children with rhabdoid tumors) from the test set, features from the second-order statistics showed an area under the curve greater than 0.89 for differentiating Wilms tumor from the others. These features aided in the differentiation of tumor type in the two children with Wilms tumors in the validation set. Therefore, texture analysis from gray-scale ultrasonography images can be used to differentiate Wilms tumors from clear cell sarcomas and rhabdoid tumors in children.

Original languageEnglish
Pages (from-to)2205-2212
Number of pages8
JournalUltrasound in Medicine and Biology
Volume45
Issue number8
DOIs
Publication statusPublished - 2019 Aug

Bibliographical note

Publisher Copyright:
© 2019

All Science Journal Classification (ASJC) codes

  • Radiological and Ultrasound Technology
  • Biophysics
  • Acoustics and Ultrasonics

Fingerprint

Dive into the research topics of 'Texture Analysis to Differentiate Malignant Renal Tumors in Children Using Gray-Scale Ultrasonography Images'. Together they form a unique fingerprint.

Cite this