Classification with CNN features and SVM on Embedded DSP Core for Colorectal Magnified NBI Endoscopic Video Image

Masayuki ODAGAWA; Takumi OKAMOTO; Tetsushi KOIDE; Toru TAMAKI; Shigeto YOSHIDA; Hiroshi MIENO; Shinji TANAKA

doi:10.1587/transfun.2021EAP1036

Classification with CNN features and SVM on Embedded DSP Core for Colorectal Magnified NBI Endoscopic Video Image

Masayuki ODAGAWA, Takumi OKAMOTO, Tetsushi KOIDE, Toru TAMAKI, Shigeto YOSHIDA, Hiroshi MIENO, Shinji TANAKA

Full Text Views

0

Cite this

Summary :

In this paper, we present a classification method for a Computer-Aided Diagnosis (CAD) system in a colorectal magnified Narrow Band Imaging (NBI) endoscopy. In an endoscopic video image, color shift, blurring or reflection of light occurs in a lesion area, which affects the discrimination result by a computer. Therefore, in order to identify lesions with high robustness and stable classification to these images specific to video frame, we implement a CAD system for colorectal endoscopic images with the Convolutional Neural Network (CNN) feature and Support Vector Machine (SVM) classification on the embedded DSP core. To improve the robustness of CAD system, we construct the SVM learned by multiple image sizes data sets so as to adapt to the noise peculiar to the video image. We confirmed that the proposed method achieves higher robustness, stable, and high classification accuracy in the endoscopic video image. The proposed method also can cope with differences in resolution by old and new endoscopes and perform stably with respect to the input endoscopic video image.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E105-A No.1 pp.25-34

Publication Date: 2022/01/01

Publicized: 2021/07/21

Online ISSN: 1745-1337

DOI: 10.1587/transfun.2021EAP1036

Type of Manuscript: PAPER

Category: VLSI Design Technology and CAD

Authors

Masayuki ODAGAWA
  Hiroshima University,Cadence Design Systems Japan
Takumi OKAMOTO
  Cadence Design Systems Japan
Tetsushi KOIDE
  Hiroshima University
Toru TAMAKI
  Nagoya Institute of Technology
Shigeto YOSHIDA
  Medical Corporation JR Hiroshima Hospital
Hiroshi MIENO
  Medical Corporation JR Hiroshima Hospital
Shinji TANAKA
  Hiroshima University

Keyword

medical image/video processing, computer-aided diagnosis system (CAD), convolutional neural network (CNN), support vector machine (SVM)

Cite this

Copy

Masayuki ODAGAWA, Takumi OKAMOTO, Tetsushi KOIDE, Toru TAMAKI, Shigeto YOSHIDA, Hiroshi MIENO, Shinji TANAKA, "Classification with CNN features and SVM on Embedded DSP Core for Colorectal Magnified NBI Endoscopic Video Image" in IEICE TRANSACTIONS on Fundamentals, vol. E105-A, no. 1, pp. 25-34, January 2022, doi: 10.1587/transfun.2021EAP1036.
Abstract: In this paper, we present a classification method for a Computer-Aided Diagnosis (CAD) system in a colorectal magnified Narrow Band Imaging (NBI) endoscopy. In an endoscopic video image, color shift, blurring or reflection of light occurs in a lesion area, which affects the discrimination result by a computer. Therefore, in order to identify lesions with high robustness and stable classification to these images specific to video frame, we implement a CAD system for colorectal endoscopic images with the Convolutional Neural Network (CNN) feature and Support Vector Machine (SVM) classification on the embedded DSP core. To improve the robustness of CAD system, we construct the SVM learned by multiple image sizes data sets so as to adapt to the noise peculiar to the video image. We confirmed that the proposed method achieves higher robustness, stable, and high classification accuracy in the endoscopic video image. The proposed method also can cope with differences in resolution by old and new endoscopes and perform stably with respect to the input endoscopic video image.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.2021EAP1036/_p

Copy

@ARTICLE{e105-a_1_25,
author={Masayuki ODAGAWA, Takumi OKAMOTO, Tetsushi KOIDE, Toru TAMAKI, Shigeto YOSHIDA, Hiroshi MIENO, Shinji TANAKA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Classification with CNN features and SVM on Embedded DSP Core for Colorectal Magnified NBI Endoscopic Video Image},
year={2022},
volume={E105-A},
number={1},
pages={25-34},
abstract={In this paper, we present a classification method for a Computer-Aided Diagnosis (CAD) system in a colorectal magnified Narrow Band Imaging (NBI) endoscopy. In an endoscopic video image, color shift, blurring or reflection of light occurs in a lesion area, which affects the discrimination result by a computer. Therefore, in order to identify lesions with high robustness and stable classification to these images specific to video frame, we implement a CAD system for colorectal endoscopic images with the Convolutional Neural Network (CNN) feature and Support Vector Machine (SVM) classification on the embedded DSP core. To improve the robustness of CAD system, we construct the SVM learned by multiple image sizes data sets so as to adapt to the noise peculiar to the video image. We confirmed that the proposed method achieves higher robustness, stable, and high classification accuracy in the endoscopic video image. The proposed method also can cope with differences in resolution by old and new endoscopes and perform stably with respect to the input endoscopic video image.},
keywords={},
doi={10.1587/transfun.2021EAP1036},
ISSN={1745-1337},
month={January},}

Copy

TY - JOUR
TI - Classification with CNN features and SVM on Embedded DSP Core for Colorectal Magnified NBI Endoscopic Video Image
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 25
EP - 34
AU - Masayuki ODAGAWA
AU - Takumi OKAMOTO
AU - Tetsushi KOIDE
AU - Toru TAMAKI
AU - Shigeto YOSHIDA
AU - Hiroshi MIENO
AU - Shinji TANAKA
PY - 2022
DO - 10.1587/transfun.2021EAP1036
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E105-A
IS - 1
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - January 2022
AB - In this paper, we present a classification method for a Computer-Aided Diagnosis (CAD) system in a colorectal magnified Narrow Band Imaging (NBI) endoscopy. In an endoscopic video image, color shift, blurring or reflection of light occurs in a lesion area, which affects the discrimination result by a computer. Therefore, in order to identify lesions with high robustness and stable classification to these images specific to video frame, we implement a CAD system for colorectal endoscopic images with the Convolutional Neural Network (CNN) feature and Support Vector Machine (SVM) classification on the embedded DSP core. To improve the robustness of CAD system, we construct the SVM learned by multiple image sizes data sets so as to adapt to the noise peculiar to the video image. We confirmed that the proposed method achieves higher robustness, stable, and high classification accuracy in the endoscopic video image. The proposed method also can cope with differences in resolution by old and new endoscopes and perform stably with respect to the input endoscopic video image.
ER -