MRI-based prostate cancer detection with high-level representation and hierarchical classification

Yulian Zhu; Li Wang; Mingxia Liu; Chunjun Qian; Ambereen Yousuf; Aytekin Oto; Dinggang Shen

doi:10.1002/mp.12116

MRI-based prostate cancer detection with high-level representation and hierarchical classification

Yulian Zhu, Li Wang, Mingxia Liu, Chunjun Qian, Ambereen Yousuf, Aytekin Oto, Dinggang Shen

Department of Brain and Cognitive Engineering

Research output: Contribution to journal › Article › peer-review

31 Citations (Scopus)

Abstract

PURPOSE: Extracting the high-level feature representation by using deep neural networks for detection of prostate cancer, and then based on high-level feature representation constructing hierarchical classification to refine the detection results.

METHODS: High-level feature representation is first learned by a deep learning network, where multiparametric MR images are used as the input data. Then, based on the learned high-level features, a hierarchical classification method is developed, where multiple random forest classifiers are iteratively constructed to refine the detection results of prostate cancer.

RESULTS: The experiments were carried on 21 real patient subjects, and the proposed method achieves an averaged section-based evaluation (SBE) of 89.90%, an averaged sensitivity of 91.51%, and an averaged specificity of 88.47%.

CONCLUSIONS: The high-level features learned from our proposed method can achieve better performance than the conventional handcrafted features (e.g., LBP and Haar-like features) in detecting prostate cancer regions, also the context features obtained from the proposed hierarchical classification approach are effective in refining cancer detection result.

Original language	English
Pages (from-to)	1028-1039
Number of pages	12
Journal	Medical physics
Volume	44
Issue number	3
DOIs	https://doi.org/10.1002/mp.12116
Publication status	Published - 2017 Mar 1

Keywords

deep learning
hierarchical classification
magnetic resonance imaging (MRI)
prostate cancer detection
random forest

ASJC Scopus subject areas

Biophysics
Radiology Nuclear Medicine and imaging

UN SDGs

This output contributes to the following Sustainable Development Goal(s)

Access to Document

10.1002/mp.12116

Fingerprint Dive into the research topics of 'MRI-based prostate cancer detection with high-level representation and hierarchical classification'. Together they form a unique fingerprint.

Cite this

@article{abf6cba0d0fa47b5a1c7d92b318d423e,

title = "MRI-based prostate cancer detection with high-level representation and hierarchical classification",

abstract = "PURPOSE: Extracting the high-level feature representation by using deep neural networks for detection of prostate cancer, and then based on high-level feature representation constructing hierarchical classification to refine the detection results.METHODS: High-level feature representation is first learned by a deep learning network, where multiparametric MR images are used as the input data. Then, based on the learned high-level features, a hierarchical classification method is developed, where multiple random forest classifiers are iteratively constructed to refine the detection results of prostate cancer.RESULTS: The experiments were carried on 21 real patient subjects, and the proposed method achieves an averaged section-based evaluation (SBE) of 89.90%, an averaged sensitivity of 91.51%, and an averaged specificity of 88.47%.CONCLUSIONS: The high-level features learned from our proposed method can achieve better performance than the conventional handcrafted features (e.g., LBP and Haar-like features) in detecting prostate cancer regions, also the context features obtained from the proposed hierarchical classification approach are effective in refining cancer detection result.",

keywords = "deep learning, hierarchical classification, magnetic resonance imaging (MRI), prostate cancer detection, random forest",

author = "Yulian Zhu and Li Wang and Mingxia Liu and Chunjun Qian and Ambereen Yousuf and Aytekin Oto and Dinggang Shen",

note = "Publisher Copyright: {\textcopyright} 2017 American Association of Physicists in Medicine. Copyright: This record is sourced from MEDLINE{\textregistered}/PubMed{\textregistered}, a database of the U.S. National Library of Medicine",

year = "2017",

month = mar,

day = "1",

doi = "10.1002/mp.12116",

language = "English",

volume = "44",

pages = "1028--1039",

journal = "Medical Physics",

issn = "0094-2405",

publisher = "AAPM - American Association of Physicists in Medicine",

number = "3",

}

TY - JOUR

T1 - MRI-based prostate cancer detection with high-level representation and hierarchical classification

AU - Zhu, Yulian

AU - Wang, Li

AU - Liu, Mingxia

AU - Qian, Chunjun

AU - Yousuf, Ambereen

AU - Oto, Aytekin

AU - Shen, Dinggang

PY - 2017/3/1

Y1 - 2017/3/1

N2 - PURPOSE: Extracting the high-level feature representation by using deep neural networks for detection of prostate cancer, and then based on high-level feature representation constructing hierarchical classification to refine the detection results.METHODS: High-level feature representation is first learned by a deep learning network, where multiparametric MR images are used as the input data. Then, based on the learned high-level features, a hierarchical classification method is developed, where multiple random forest classifiers are iteratively constructed to refine the detection results of prostate cancer.RESULTS: The experiments were carried on 21 real patient subjects, and the proposed method achieves an averaged section-based evaluation (SBE) of 89.90%, an averaged sensitivity of 91.51%, and an averaged specificity of 88.47%.CONCLUSIONS: The high-level features learned from our proposed method can achieve better performance than the conventional handcrafted features (e.g., LBP and Haar-like features) in detecting prostate cancer regions, also the context features obtained from the proposed hierarchical classification approach are effective in refining cancer detection result.

AB - PURPOSE: Extracting the high-level feature representation by using deep neural networks for detection of prostate cancer, and then based on high-level feature representation constructing hierarchical classification to refine the detection results.METHODS: High-level feature representation is first learned by a deep learning network, where multiparametric MR images are used as the input data. Then, based on the learned high-level features, a hierarchical classification method is developed, where multiple random forest classifiers are iteratively constructed to refine the detection results of prostate cancer.RESULTS: The experiments were carried on 21 real patient subjects, and the proposed method achieves an averaged section-based evaluation (SBE) of 89.90%, an averaged sensitivity of 91.51%, and an averaged specificity of 88.47%.CONCLUSIONS: The high-level features learned from our proposed method can achieve better performance than the conventional handcrafted features (e.g., LBP and Haar-like features) in detecting prostate cancer regions, also the context features obtained from the proposed hierarchical classification approach are effective in refining cancer detection result.

KW - deep learning

KW - hierarchical classification

KW - magnetic resonance imaging (MRI)

KW - prostate cancer detection

KW - random forest

UR - http://www.scopus.com/inward/record.url?scp=85016235192&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85016235192&partnerID=8YFLogxK

U2 - 10.1002/mp.12116

DO - 10.1002/mp.12116

M3 - Article

C2 - 28107548

AN - SCOPUS:85016235192

VL - 44

SP - 1028

EP - 1039

JO - Medical Physics

JF - Medical Physics

SN - 0094-2405

IS - 3

ER -

MRI-based prostate cancer detection with high-level representation and hierarchical classification

Abstract

Keywords

ASJC Scopus subject areas

UN SDGs

Access to Document

Other files and links

Cite this