Advanced search
Publication Cover
Journal of Multidisciplinary Healthcare Volume 14, 2021 - Issue
Submit an article Journal homepage
593
Views
15
CrossRef citations to date
0
Altmetric
Original Research

Deep Learning Classifier with Patient’s Metadata of Dermoscopic Images in Malignant Melanoma Detection

Dina Nur Anggraini Ningrum1 Graduate Institute of Biomedical Informatics, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan;2 Public Health Department, Universitas Negeri Semarang, Semarang City, IndonesiaORCID Iconhttps://orcid.org/0000-0002-8557-2862
ORCID Icon,
Sheng-Po Yuan1 Graduate Institute of Biomedical Informatics, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan;3 Department of Otorhinolaryngology, Shuang-Ho Hospital, Taipei Medical University, New Taipei City, TaiwanORCID Iconhttps://orcid.org/0000-0003-3515-3480
ORCID Icon,
Woon-Man Kung4 Department of Exercise and Health Promotion, College of Kinesiology and Health, Chinese Culture University, Taipei, TaiwanORCID Iconhttps://orcid.org/0000-0001-8311-2902
ORCID Icon,
Chieh-Chen Wu4 Department of Exercise and Health Promotion, College of Kinesiology and Health, Chinese Culture University, Taipei, TaiwanORCID Iconhttps://orcid.org/0000-0003-2205-5849
ORCID Icon,
I-Shiang Tzeng4 Department of Exercise and Health Promotion, College of Kinesiology and Health, Chinese Culture University, Taipei, Taiwan;5 Department of Research, Taipei Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, New Taipei City, Taiwan;6 Department of Statistics, National Taipei University, Taipei, TaiwanORCID Iconhttps://orcid.org/0000-0002-9047-8141
ORCID Icon,
Chu-Ya Huang7 Taiwan College of Healthcare Executives, Taipei, Taiwan
,
Jack Yu-Chuan Li1 Graduate Institute of Biomedical Informatics, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan;8 Department Dermatology, Wan Fang Hospital, Taipei, Taiwan;9 Taipei Medical University Research Center of Cancer Translational Medicine, Taipei, TaiwanCorrespondence[email protected]
ORCID Iconhttps://orcid.org/0000-0001-6497-4232
ORCID Icon &
Yao-Chin Wang10 Graduate Institute of Injury Prevention and Control, College of Public Health, Taipei Medical University, Taipei, Taiwan;11 Department of Emergency Medicine, Min-Sheng General Hospital, Taoyuan, TaiwanCorrespondence[email protected]
ORCID Iconhttps://orcid.org/0000-0002-0094-8016
ORCID Icon show all
Pages 877-885 | Published online: 21 Apr 2021

Figures & data

Figure 1 Dermoscopic image data preprocessing. Original dermoscopic image (left side) obtained from dermatology repository International Skin Imaging Collaboration (ISIC) archive (www.isic-archive.com) was cropped and removed the ruler, sticker, or hair (right side) for further analysis.

Figure 1 Dermoscopic image data preprocessing. Original dermoscopic image (left side) obtained from dermatology repository International Skin Imaging Collaboration (ISIC) archive (www.isic-archive.com) was cropped and removed the ruler, sticker, or hair (right side) for further analysis.

Figure 2 Schematic representation of experimental methodology. Initially input data is preprocessed to generate the dataset to be employed to further establish a CNN model. In parallel, patient metadata variable is also preprocessed to generate dataset for establishing the ANN model. These independent models were further combined to assess the impact on accuracy of the prediction between image only data (CNN model) and image plus patient’s metadata (CNN+ANN model).

Figure 2 Schematic representation of experimental methodology. Initially input data is preprocessed to generate the dataset to be employed to further establish a CNN model. In parallel, patient metadata variable is also preprocessed to generate dataset for establishing the ANN model. These independent models were further combined to assess the impact on accuracy of the prediction between image only data (CNN model) and image plus patient’s metadata (CNN+ANN model).

Figure 3 Model architecture. CNN model architecture (A) and CNN+ANN model architecture (B) showed the flow inside the model (printed from TensorFlow 1.14).

Figure 3 Model architecture. CNN model architecture (A) and CNN+ANN model architecture (B) showed the flow inside the model (printed from TensorFlow 1.14).

Table 1 Distribution of Patient’s Metadata

Figure 4 Mode learning curve. Based on loss and accuracy curve of training and validation vs epoch in each model, the learning optimization and performance of the CNN+ANN model (B) has improved more than the CNN model (A). The loss and accuracy of the CNN+ANN model (B) shows that the model is more representative and less overfitting than the CNN model (A).

Figure 4 Mode learning curve. Based on loss and accuracy curve of training and validation vs epoch in each model, the learning optimization and performance of the CNN+ANN model (B) has improved more than the CNN model (A). The loss and accuracy of the CNN+ANN model (B) shows that the model is more representative and less overfitting than the CNN model (A).

Figure 5 Area under the receiver operating characteristic curve (AUROC). The CNN+ANN model (red line) outperforms the CNN model (blue line).

Figure 5 Area under the receiver operating characteristic curve (AUROC). The CNN+ANN model (red line) outperforms the CNN model (blue line).

Figure 6 T-distributed stochastic neighbor embedding (t-SNE) plot to describe label distribution of malignant melanoma in the CNN model (A) and the CNN+ ANN model (B). The malignant melanoma tends to be classified better in the CNN+ANN model (B) than in the CNN model (A). One datum with malignant melanoma represented by one orange dot, one datum with nonmalignant melanoma represented by one blue dot.

Figure 6 T-distributed stochastic neighbor embedding (t-SNE) plot to describe label distribution of malignant melanoma in the CNN model (A) and the CNN+ ANN model (B). The malignant melanoma tends to be classified better in the CNN+ANN model (B) than in the CNN model (A). One datum with malignant melanoma represented by one orange dot, one datum with nonmalignant melanoma represented by one blue dot.

Table 2 Confusion Matrix CNN

Table 3 Confusion Matrix CNN+ANN

Table 4 Malignant Melanomas Detection Result Using CNN Model and CNN+ANN Model

Related research

People also read lists articles that other readers of this article have read.

Recommended articles lists articles that we recommend and is powered by our AI driven recommendation engine.

Cited by lists all citing articles based on Crossref citations.
Articles with the Crossref icon will open in a new tab.