A Deep Learning Approach for Targeted Contrast-Enhanced Ultrasound Based Prostate Cancer Detection

Yujie Feng, Fan Yang, Xichuan Zhou, Yanli Guo, Fang Tang, Fengbo Ren, Jishun Guo, Shuiwang Ji

Research output: Contribution to journalArticle

Abstract

The important role of angiogenesis in cancer development has driven many researchers to investigate the prospects of noninvasive cancer diagnosis based on the technology of contrast-enhanced ultrasound (CEUS) imaging. This paper presents a deep learning framework to detect prostate cancer in the sequential CEUS images. The proposed method uniformly extracts features from both the spatial and the temporal dimensions by performing three-dimensional convolution operations, which captures the dynamic information of the perfusion process encoded in multiple adjacent frames for prostate cancer detection. The deep learning models were trained and validated against expert delineations over the CEUS images recorded using two types of contrast agents, i.e. the anti-PSMA based agent targeted to prostate cancer cells and the non-targeted blank agent. Experiments showed that the deep learning method achieved over 91% specificity and 90% average accuracy over the targeted CEUS images for prostate cancer detection, which was superior (<formula><tex>$p&lt;0.05$</tex></formula>) than previously reported approaches and implementations.

Original languageEnglish (US)
JournalIEEE/ACM Transactions on Computational Biology and Bioinformatics
DOIs
StateAccepted/In press - May 11 2018

Fingerprint

Prostate Cancer
Ultrasound
Ultrasound Image
Prostatic Neoplasms
Ultrasonics
Learning
Cancer
Angiogenesis
Convolution
Contrast Media
Specificity
Ultrasonography
Neoplasms
Adjacent
Perfusion
Cells
Research Personnel
Imaging
Technology
Imaging techniques

Keywords

  • Blood
  • Cancer detection
  • Contrast-enhanced ultrasound
  • convolutional neural network
  • Feature extraction
  • Machine learning
  • Prostate cancer
  • prostate cancer detection
  • targeted agent
  • Ultrasonic imaging

ASJC Scopus subject areas

  • Biotechnology
  • Genetics
  • Applied Mathematics

Cite this

A Deep Learning Approach for Targeted Contrast-Enhanced Ultrasound Based Prostate Cancer Detection. / Feng, Yujie; Yang, Fan; Zhou, Xichuan; Guo, Yanli; Tang, Fang; Ren, Fengbo; Guo, Jishun; Ji, Shuiwang.

In: IEEE/ACM Transactions on Computational Biology and Bioinformatics, 11.05.2018.

Research output: Contribution to journalArticle

@article{9b467e4a67744e88ae39257bf30264c6,
title = "A Deep Learning Approach for Targeted Contrast-Enhanced Ultrasound Based Prostate Cancer Detection",
abstract = "The important role of angiogenesis in cancer development has driven many researchers to investigate the prospects of noninvasive cancer diagnosis based on the technology of contrast-enhanced ultrasound (CEUS) imaging. This paper presents a deep learning framework to detect prostate cancer in the sequential CEUS images. The proposed method uniformly extracts features from both the spatial and the temporal dimensions by performing three-dimensional convolution operations, which captures the dynamic information of the perfusion process encoded in multiple adjacent frames for prostate cancer detection. The deep learning models were trained and validated against expert delineations over the CEUS images recorded using two types of contrast agents, i.e. the anti-PSMA based agent targeted to prostate cancer cells and the non-targeted blank agent. Experiments showed that the deep learning method achieved over 91{\%} specificity and 90{\%} average accuracy over the targeted CEUS images for prostate cancer detection, which was superior ($p<0.05$) than previously reported approaches and implementations.",
keywords = "Blood, Cancer detection, Contrast-enhanced ultrasound, convolutional neural network, Feature extraction, Machine learning, Prostate cancer, prostate cancer detection, targeted agent, Ultrasonic imaging",
author = "Yujie Feng and Fan Yang and Xichuan Zhou and Yanli Guo and Fang Tang and Fengbo Ren and Jishun Guo and Shuiwang Ji",
year = "2018",
month = "5",
day = "11",
doi = "10.1109/TCBB.2018.2835444",
language = "English (US)",
journal = "IEEE/ACM Transactions on Computational Biology and Bioinformatics",
issn = "1545-5963",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - A Deep Learning Approach for Targeted Contrast-Enhanced Ultrasound Based Prostate Cancer Detection

AU - Feng, Yujie

AU - Yang, Fan

AU - Zhou, Xichuan

AU - Guo, Yanli

AU - Tang, Fang

AU - Ren, Fengbo

AU - Guo, Jishun

AU - Ji, Shuiwang

PY - 2018/5/11

Y1 - 2018/5/11

N2 - The important role of angiogenesis in cancer development has driven many researchers to investigate the prospects of noninvasive cancer diagnosis based on the technology of contrast-enhanced ultrasound (CEUS) imaging. This paper presents a deep learning framework to detect prostate cancer in the sequential CEUS images. The proposed method uniformly extracts features from both the spatial and the temporal dimensions by performing three-dimensional convolution operations, which captures the dynamic information of the perfusion process encoded in multiple adjacent frames for prostate cancer detection. The deep learning models were trained and validated against expert delineations over the CEUS images recorded using two types of contrast agents, i.e. the anti-PSMA based agent targeted to prostate cancer cells and the non-targeted blank agent. Experiments showed that the deep learning method achieved over 91% specificity and 90% average accuracy over the targeted CEUS images for prostate cancer detection, which was superior ($p<0.05$) than previously reported approaches and implementations.

AB - The important role of angiogenesis in cancer development has driven many researchers to investigate the prospects of noninvasive cancer diagnosis based on the technology of contrast-enhanced ultrasound (CEUS) imaging. This paper presents a deep learning framework to detect prostate cancer in the sequential CEUS images. The proposed method uniformly extracts features from both the spatial and the temporal dimensions by performing three-dimensional convolution operations, which captures the dynamic information of the perfusion process encoded in multiple adjacent frames for prostate cancer detection. The deep learning models were trained and validated against expert delineations over the CEUS images recorded using two types of contrast agents, i.e. the anti-PSMA based agent targeted to prostate cancer cells and the non-targeted blank agent. Experiments showed that the deep learning method achieved over 91% specificity and 90% average accuracy over the targeted CEUS images for prostate cancer detection, which was superior ($p<0.05$) than previously reported approaches and implementations.

KW - Blood

KW - Cancer detection

KW - Contrast-enhanced ultrasound

KW - convolutional neural network

KW - Feature extraction

KW - Machine learning

KW - Prostate cancer

KW - prostate cancer detection

KW - targeted agent

KW - Ultrasonic imaging

UR - http://www.scopus.com/inward/record.url?scp=85046756724&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046756724&partnerID=8YFLogxK

U2 - 10.1109/TCBB.2018.2835444

DO - 10.1109/TCBB.2018.2835444

M3 - Article

JO - IEEE/ACM Transactions on Computational Biology and Bioinformatics

JF - IEEE/ACM Transactions on Computational Biology and Bioinformatics

SN - 1545-5963

ER -