Performance evaluation of radiologists with artificial neural network for differential diagnosis of intra-axial cerebral tumors on MR images

K. Yamashita, Takashi Yoshiura, H. Arimura, F. Mihara, T. Noguchi, A. Hiwatashi, O. Togao, Y. Yamashita, T. Shono, S. Kumazawa, Y. Higashida, H. Honda

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

BACKGROUND AND PURPOSE: Previous studies have suggested that use of an artificial neural network (ANN) system is beneficial for radiological diagnosis. Our purposes in this study were to construct an ANN for the differential diagnosis of intra-axial cerebral tumors on MR images and to evaluate the effect of ANN outputs on radiologists' diagnostic performance. MATERIALS AND METHODS: We collected MR images of 126 patients with intra-axial cerebral tumors (58 high-grade gliomas, 37 low-grade gliomas, 19 metastatic tumors, and 12 malignant lymphomas). We constructed a single 3-layer feed-forward ANN with a Levenberg-Marquardt algorithm. The ANN was designed to differentiate among 4 categories of tumors (high-grade gliomas, low-grade gliomas, metastases, and malignant lymphomas) with use of 2 clinical parameters and 13 radiologic findings in MR images. Subjective ratings for the 13 radiologic findings were provided independently by 2 attending radiologists. All 126 cases were used for training and testing of the ANN based on a leave-one-out-by-case method. In the observer test, MR images were viewed by 9 radiologists, first without and then with ANN outputs. Each radiologist's performance was evaluated through a receiver operating characteristic (ROC) analysis on a continuous rating scale. RESULTS: The averaged area under the ROC curve for ANN alone was 0.949. The diagnostic performance of the 9 radiologists increased from 0.899 to 0.946 (P < .001) when they used ANN outputs. CONCLUSIONS: The ANN can provide useful output as a second opinion to improve radiologists' diagnostic performance in the differential diagnosis of intra-axial cerebral tumors seen on MR imaging.

Original languageEnglish
Pages (from-to)1153-1158
Number of pages6
JournalAmerican Journal of Neuroradiology
Volume29
Issue number6
DOIs
Publication statusPublished - Jun 1 2008

Fingerprint

Differential Diagnosis
Glioma
Neoplasms
ROC Curve
Lymphoma
Radiologists
Referral and Consultation
Neoplasm Metastasis

All Science Journal Classification (ASJC) codes

  • Radiology Nuclear Medicine and imaging
  • Clinical Neurology

Cite this

Performance evaluation of radiologists with artificial neural network for differential diagnosis of intra-axial cerebral tumors on MR images. / Yamashita, K.; Yoshiura, Takashi; Arimura, H.; Mihara, F.; Noguchi, T.; Hiwatashi, A.; Togao, O.; Yamashita, Y.; Shono, T.; Kumazawa, S.; Higashida, Y.; Honda, H.

In: American Journal of Neuroradiology, Vol. 29, No. 6, 01.06.2008, p. 1153-1158.

Research output: Contribution to journalArticle

Yamashita, K. ; Yoshiura, Takashi ; Arimura, H. ; Mihara, F. ; Noguchi, T. ; Hiwatashi, A. ; Togao, O. ; Yamashita, Y. ; Shono, T. ; Kumazawa, S. ; Higashida, Y. ; Honda, H. / Performance evaluation of radiologists with artificial neural network for differential diagnosis of intra-axial cerebral tumors on MR images. In: American Journal of Neuroradiology. 2008 ; Vol. 29, No. 6. pp. 1153-1158.
@article{ae440ac391fb49119c09e28c0242586a,
title = "Performance evaluation of radiologists with artificial neural network for differential diagnosis of intra-axial cerebral tumors on MR images",
abstract = "BACKGROUND AND PURPOSE: Previous studies have suggested that use of an artificial neural network (ANN) system is beneficial for radiological diagnosis. Our purposes in this study were to construct an ANN for the differential diagnosis of intra-axial cerebral tumors on MR images and to evaluate the effect of ANN outputs on radiologists' diagnostic performance. MATERIALS AND METHODS: We collected MR images of 126 patients with intra-axial cerebral tumors (58 high-grade gliomas, 37 low-grade gliomas, 19 metastatic tumors, and 12 malignant lymphomas). We constructed a single 3-layer feed-forward ANN with a Levenberg-Marquardt algorithm. The ANN was designed to differentiate among 4 categories of tumors (high-grade gliomas, low-grade gliomas, metastases, and malignant lymphomas) with use of 2 clinical parameters and 13 radiologic findings in MR images. Subjective ratings for the 13 radiologic findings were provided independently by 2 attending radiologists. All 126 cases were used for training and testing of the ANN based on a leave-one-out-by-case method. In the observer test, MR images were viewed by 9 radiologists, first without and then with ANN outputs. Each radiologist's performance was evaluated through a receiver operating characteristic (ROC) analysis on a continuous rating scale. RESULTS: The averaged area under the ROC curve for ANN alone was 0.949. The diagnostic performance of the 9 radiologists increased from 0.899 to 0.946 (P < .001) when they used ANN outputs. CONCLUSIONS: The ANN can provide useful output as a second opinion to improve radiologists' diagnostic performance in the differential diagnosis of intra-axial cerebral tumors seen on MR imaging.",
author = "K. Yamashita and Takashi Yoshiura and H. Arimura and F. Mihara and T. Noguchi and A. Hiwatashi and O. Togao and Y. Yamashita and T. Shono and S. Kumazawa and Y. Higashida and H. Honda",
year = "2008",
month = "6",
day = "1",
doi = "10.3174/ajnr.A1037",
language = "English",
volume = "29",
pages = "1153--1158",
journal = "American Journal of Neuroradiology",
issn = "0195-6108",
publisher = "American Society of Neuroradiology",
number = "6",

}

TY - JOUR

T1 - Performance evaluation of radiologists with artificial neural network for differential diagnosis of intra-axial cerebral tumors on MR images

AU - Yamashita, K.

AU - Yoshiura, Takashi

AU - Arimura, H.

AU - Mihara, F.

AU - Noguchi, T.

AU - Hiwatashi, A.

AU - Togao, O.

AU - Yamashita, Y.

AU - Shono, T.

AU - Kumazawa, S.

AU - Higashida, Y.

AU - Honda, H.

PY - 2008/6/1

Y1 - 2008/6/1

N2 - BACKGROUND AND PURPOSE: Previous studies have suggested that use of an artificial neural network (ANN) system is beneficial for radiological diagnosis. Our purposes in this study were to construct an ANN for the differential diagnosis of intra-axial cerebral tumors on MR images and to evaluate the effect of ANN outputs on radiologists' diagnostic performance. MATERIALS AND METHODS: We collected MR images of 126 patients with intra-axial cerebral tumors (58 high-grade gliomas, 37 low-grade gliomas, 19 metastatic tumors, and 12 malignant lymphomas). We constructed a single 3-layer feed-forward ANN with a Levenberg-Marquardt algorithm. The ANN was designed to differentiate among 4 categories of tumors (high-grade gliomas, low-grade gliomas, metastases, and malignant lymphomas) with use of 2 clinical parameters and 13 radiologic findings in MR images. Subjective ratings for the 13 radiologic findings were provided independently by 2 attending radiologists. All 126 cases were used for training and testing of the ANN based on a leave-one-out-by-case method. In the observer test, MR images were viewed by 9 radiologists, first without and then with ANN outputs. Each radiologist's performance was evaluated through a receiver operating characteristic (ROC) analysis on a continuous rating scale. RESULTS: The averaged area under the ROC curve for ANN alone was 0.949. The diagnostic performance of the 9 radiologists increased from 0.899 to 0.946 (P < .001) when they used ANN outputs. CONCLUSIONS: The ANN can provide useful output as a second opinion to improve radiologists' diagnostic performance in the differential diagnosis of intra-axial cerebral tumors seen on MR imaging.

AB - BACKGROUND AND PURPOSE: Previous studies have suggested that use of an artificial neural network (ANN) system is beneficial for radiological diagnosis. Our purposes in this study were to construct an ANN for the differential diagnosis of intra-axial cerebral tumors on MR images and to evaluate the effect of ANN outputs on radiologists' diagnostic performance. MATERIALS AND METHODS: We collected MR images of 126 patients with intra-axial cerebral tumors (58 high-grade gliomas, 37 low-grade gliomas, 19 metastatic tumors, and 12 malignant lymphomas). We constructed a single 3-layer feed-forward ANN with a Levenberg-Marquardt algorithm. The ANN was designed to differentiate among 4 categories of tumors (high-grade gliomas, low-grade gliomas, metastases, and malignant lymphomas) with use of 2 clinical parameters and 13 radiologic findings in MR images. Subjective ratings for the 13 radiologic findings were provided independently by 2 attending radiologists. All 126 cases were used for training and testing of the ANN based on a leave-one-out-by-case method. In the observer test, MR images were viewed by 9 radiologists, first without and then with ANN outputs. Each radiologist's performance was evaluated through a receiver operating characteristic (ROC) analysis on a continuous rating scale. RESULTS: The averaged area under the ROC curve for ANN alone was 0.949. The diagnostic performance of the 9 radiologists increased from 0.899 to 0.946 (P < .001) when they used ANN outputs. CONCLUSIONS: The ANN can provide useful output as a second opinion to improve radiologists' diagnostic performance in the differential diagnosis of intra-axial cerebral tumors seen on MR imaging.

UR - http://www.scopus.com/inward/record.url?scp=46749125077&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=46749125077&partnerID=8YFLogxK

U2 - 10.3174/ajnr.A1037

DO - 10.3174/ajnr.A1037

M3 - Article

C2 - 18388216

AN - SCOPUS:46749125077

VL - 29

SP - 1153

EP - 1158

JO - American Journal of Neuroradiology

JF - American Journal of Neuroradiology

SN - 0195-6108

IS - 6

ER -