Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Performance improvement of automated melanoma diagnosis system by data augmentation

Kato, Kana ; Nemoto, Mitsutaka ; Kimura, Yuichi ; Kiyohara, Yoshio ; Koga, Hiroshi ; Yamazaki, Naoya ; Christensen, Gustav LU ; Ingvar, Christian LU ; Nielsen, Kari LU orcid and Nakamura, Atsushi , et al. (2020) In Advanced Biomedical Engineering 9. p.62-70
Abstract

Color information is an important tool for diagnosing melanoma. In this study, we used a hyper-spectral imager (HSI), which can measure color information in detail, to develop an automated melanoma diagnosis system. In recent years, the effectiveness of deep learning has become more widely accepted in the field of image recognition. We therefore integrated the deep convolutional neural network with transfer learning into our system. We tried data augmentation to demonstrate how our system improves diagnostic performance. 283 melanoma lesions and 336 non-melanoma lesions were used for the analysis. The data measured by HSI, called the hyperspectral data (HSD), were converted to a single-wavelength image averaged over plus or minus 3 nm.... (More)

Color information is an important tool for diagnosing melanoma. In this study, we used a hyper-spectral imager (HSI), which can measure color information in detail, to develop an automated melanoma diagnosis system. In recent years, the effectiveness of deep learning has become more widely accepted in the field of image recognition. We therefore integrated the deep convolutional neural network with transfer learning into our system. We tried data augmentation to demonstrate how our system improves diagnostic performance. 283 melanoma lesions and 336 non-melanoma lesions were used for the analysis. The data measured by HSI, called the hyperspectral data (HSD), were converted to a single-wavelength image averaged over plus or minus 3 nm. We used GoogLeNet which was pre-trained by ImageNet and then was transferred to analyze the HSD. In the transfer learning, we used not only the original HSD but also artificial augmentation dataset to improve the melanoma classification performance of GoogLeNet. Since GoogLeNet requires three-channel images as input, three wavelengths were selected from those single-wavelength images and assigned to three channels in wavelength order from short to long. The sensitivity and specificity of our system were estimated by 5-fold cross-val-idation. The results of a combination of 530, 560, and 590 nm (combination A) and 500, 620, and 740 nm (com-bination B) were compared. We also compared the diagnostic performance with and without the data augmentation. All images were augmented by inverting the image vertically and/or horizontally. Without data augmentation, the respective sensitivity and specificity of our system were 77.4% and 75.6% for combination A and 73.1% and 80.6% for combination B. With data augmentation, these numbers improved to 79.9% and 82.4% for combination A and 76.7% and 82.2% for combination B. From these results, we conclude that the diagnostic performance of our system has been improved by data augmentation. Furthermore, our system suc-ceeds to differentiate melanoma with a sensitivity of almost 80%.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; ; and , et al. (More)
; ; ; ; ; ; ; ; ; ; and (Less)
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Data augmentation, Deep learning, Hyperspectral imager, Melanoma, Transfer learning
in
Advanced Biomedical Engineering
volume
9
pages
9 pages
publisher
Japanese Society for Medical and Biological Engineering
external identifiers
  • scopus:85082580920
ISSN
2187-5219
DOI
10.14326/abe.9.62
language
English
LU publication?
yes
id
c33b4900-5ec3-496c-898a-c93572d5e3e8
date added to LUP
2020-04-28 15:01:10
date last changed
2022-04-18 22:06:49
@article{c33b4900-5ec3-496c-898a-c93572d5e3e8,
  abstract     = {{<p>Color information is an important tool for diagnosing melanoma. In this study, we used a hyper-spectral imager (HSI), which can measure color information in detail, to develop an automated melanoma diagnosis system. In recent years, the effectiveness of deep learning has become more widely accepted in the field of image recognition. We therefore integrated the deep convolutional neural network with transfer learning into our system. We tried data augmentation to demonstrate how our system improves diagnostic performance. 283 melanoma lesions and 336 non-melanoma lesions were used for the analysis. The data measured by HSI, called the hyperspectral data (HSD), were converted to a single-wavelength image averaged over plus or minus 3 nm. We used GoogLeNet which was pre-trained by ImageNet and then was transferred to analyze the HSD. In the transfer learning, we used not only the original HSD but also artificial augmentation dataset to improve the melanoma classification performance of GoogLeNet. Since GoogLeNet requires three-channel images as input, three wavelengths were selected from those single-wavelength images and assigned to three channels in wavelength order from short to long. The sensitivity and specificity of our system were estimated by 5-fold cross-val-idation. The results of a combination of 530, 560, and 590 nm (combination A) and 500, 620, and 740 nm (com-bination B) were compared. We also compared the diagnostic performance with and without the data augmentation. All images were augmented by inverting the image vertically and/or horizontally. Without data augmentation, the respective sensitivity and specificity of our system were 77.4% and 75.6% for combination A and 73.1% and 80.6% for combination B. With data augmentation, these numbers improved to 79.9% and 82.4% for combination A and 76.7% and 82.2% for combination B. From these results, we conclude that the diagnostic performance of our system has been improved by data augmentation. Furthermore, our system suc-ceeds to differentiate melanoma with a sensitivity of almost 80%.</p>}},
  author       = {{Kato, Kana and Nemoto, Mitsutaka and Kimura, Yuichi and Kiyohara, Yoshio and Koga, Hiroshi and Yamazaki, Naoya and Christensen, Gustav and Ingvar, Christian and Nielsen, Kari and Nakamura, Atsushi and Sota, Takayuki and Nagaoka, Takashi}},
  issn         = {{2187-5219}},
  keywords     = {{Data augmentation; Deep learning; Hyperspectral imager; Melanoma; Transfer learning}},
  language     = {{eng}},
  month        = {{03}},
  pages        = {{62--70}},
  publisher    = {{Japanese Society for Medical and Biological Engineering}},
  series       = {{Advanced Biomedical Engineering}},
  title        = {{Performance improvement of automated melanoma diagnosis system by data augmentation}},
  url          = {{http://dx.doi.org/10.14326/abe.9.62}},
  doi          = {{10.14326/abe.9.62}},
  volume       = {{9}},
  year         = {{2020}},
}