Skin image analysis for erythema migrans detection and automated lyme disease referral

Philippe Burlina, N. Joshi, Elise Ng, S. Billings, A. Rebman, John Aucott

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This study develops approaches for the automated referral of individuals with Lyme disease using erythema migrans rash (EM) images with clinical-grade or ‘in the wild’ characteristics. We develop a pre-screener using a Deep Convolutional Neural Network (DCNN) that classifies EM vs. other conditions, including either control/unaffected skin, or skin presenting with other confuser lesions. We test and report performance metrics for the proposed approach on this dataset including Cohen’s Kappa coefficient, area under the receiver operating characteristic (ROC) curve (AUC), accuracy, sensitivity, specificity. The machine classification yields accuracy (and error margin) of 93.04% (1.49), AUC of 0.9504 (0.0156), and Kappa of 0.7549 (0.0586), which is a significant improvement over previously published state-of-the-art methods. Results also suggest substantial agreement between machine and expert clinician annotated gold standard images. The DCNN model developed for this skin classifier is made publicly available and can potentially be used by others for transfer learning to other types of skin lesion classification models including those for skin cancer.

Original languageEnglish (US)
Title of host publicationOR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis - 1st International Workshop, OR 2.0 2018 5th International Workshop, CARE 2018, 7th International Workshop, CLIP 2018, 3rd International Workshop, ISIC 2018 Held in Conjunction with MICCAI 2018
EditorsAnand Malpani, Marco A. Zenati, Cristina Oyarzun Laura, M. Emre Celebi, Duygu Sarikaya, Noel C. Codella, Allan Halpern, Marius Erdt, Lena Maier-Hein, Luo Xiongbiao, Stefan Wesarg, Danail Stoyanov, Zeike Taylor, Klaus Drechsler, Kristin Dana, Anne Martel, Raj Shekhar, Sandrine De Ribaupierre, Tobias Reichl, Jonathan McLeod, Miguel Angel González Ballester, Toby Collins, Marius George Linguraru
PublisherSpringer Verlag
Pages244-251
Number of pages8
ISBN (Print)9783030012007
DOIs
StatePublished - Jan 1 2018
Event1st International Workshop on OR 2.0 Context-Aware Operating Theaters, OR 2.0 2018, 5th International Workshop on Computer Assisted Robotic Endoscopy, CARE 2018, 7th International Workshop on Clinical Image-Based Procedures, CLIP 2018, and 1st International Workshop on Skin Image Analysis, ISIC 2018, held in conjunction with the 21st International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2018 - Granada, Spain
Duration: Sep 16 2018Sep 20 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11041 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other1st International Workshop on OR 2.0 Context-Aware Operating Theaters, OR 2.0 2018, 5th International Workshop on Computer Assisted Robotic Endoscopy, CARE 2018, 7th International Workshop on Clinical Image-Based Procedures, CLIP 2018, and 1st International Workshop on Skin Image Analysis, ISIC 2018, held in conjunction with the 21st International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2018
CountrySpain
CityGranada
Period9/16/189/20/18

Fingerprint

Image Analysis
Image analysis
Skin
Cohen's kappa
Transfer Learning
Neural networks
Receiver Operating Characteristic Curve
Performance Metrics
Neural Network Model
Gold
Margin
Specificity
Cancer
Classifiers
Classify
Classifier
Neural Networks
Coefficient

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Burlina, P., Joshi, N., Ng, E., Billings, S., Rebman, A., & Aucott, J. (2018). Skin image analysis for erythema migrans detection and automated lyme disease referral. In A. Malpani, M. A. Zenati, C. Oyarzun Laura, M. E. Celebi, D. Sarikaya, N. C. Codella, A. Halpern, M. Erdt, L. Maier-Hein, L. Xiongbiao, S. Wesarg, D. Stoyanov, Z. Taylor, K. Drechsler, K. Dana, A. Martel, R. Shekhar, S. De Ribaupierre, T. Reichl, J. McLeod, M. A. González Ballester, T. Collins, ... M. G. Linguraru (Eds.), OR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis - 1st International Workshop, OR 2.0 2018 5th International Workshop, CARE 2018, 7th International Workshop, CLIP 2018, 3rd International Workshop, ISIC 2018 Held in Conjunction with MICCAI 2018 (pp. 244-251). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11041 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-01201-4_26

Skin image analysis for erythema migrans detection and automated lyme disease referral. / Burlina, Philippe; Joshi, N.; Ng, Elise; Billings, S.; Rebman, A.; Aucott, John.

OR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis - 1st International Workshop, OR 2.0 2018 5th International Workshop, CARE 2018, 7th International Workshop, CLIP 2018, 3rd International Workshop, ISIC 2018 Held in Conjunction with MICCAI 2018. ed. / Anand Malpani; Marco A. Zenati; Cristina Oyarzun Laura; M. Emre Celebi; Duygu Sarikaya; Noel C. Codella; Allan Halpern; Marius Erdt; Lena Maier-Hein; Luo Xiongbiao; Stefan Wesarg; Danail Stoyanov; Zeike Taylor; Klaus Drechsler; Kristin Dana; Anne Martel; Raj Shekhar; Sandrine De Ribaupierre; Tobias Reichl; Jonathan McLeod; Miguel Angel González Ballester; Toby Collins; Marius George Linguraru. Springer Verlag, 2018. p. 244-251 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11041 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Burlina, P, Joshi, N, Ng, E, Billings, S, Rebman, A & Aucott, J 2018, Skin image analysis for erythema migrans detection and automated lyme disease referral. in A Malpani, MA Zenati, C Oyarzun Laura, ME Celebi, D Sarikaya, NC Codella, A Halpern, M Erdt, L Maier-Hein, L Xiongbiao, S Wesarg, D Stoyanov, Z Taylor, K Drechsler, K Dana, A Martel, R Shekhar, S De Ribaupierre, T Reichl, J McLeod, MA González Ballester, T Collins & MG Linguraru (eds), OR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis - 1st International Workshop, OR 2.0 2018 5th International Workshop, CARE 2018, 7th International Workshop, CLIP 2018, 3rd International Workshop, ISIC 2018 Held in Conjunction with MICCAI 2018. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11041 LNCS, Springer Verlag, pp. 244-251, 1st International Workshop on OR 2.0 Context-Aware Operating Theaters, OR 2.0 2018, 5th International Workshop on Computer Assisted Robotic Endoscopy, CARE 2018, 7th International Workshop on Clinical Image-Based Procedures, CLIP 2018, and 1st International Workshop on Skin Image Analysis, ISIC 2018, held in conjunction with the 21st International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2018, Granada, Spain, 9/16/18. https://doi.org/10.1007/978-3-030-01201-4_26
Burlina P, Joshi N, Ng E, Billings S, Rebman A, Aucott J. Skin image analysis for erythema migrans detection and automated lyme disease referral. In Malpani A, Zenati MA, Oyarzun Laura C, Celebi ME, Sarikaya D, Codella NC, Halpern A, Erdt M, Maier-Hein L, Xiongbiao L, Wesarg S, Stoyanov D, Taylor Z, Drechsler K, Dana K, Martel A, Shekhar R, De Ribaupierre S, Reichl T, McLeod J, González Ballester MA, Collins T, Linguraru MG, editors, OR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis - 1st International Workshop, OR 2.0 2018 5th International Workshop, CARE 2018, 7th International Workshop, CLIP 2018, 3rd International Workshop, ISIC 2018 Held in Conjunction with MICCAI 2018. Springer Verlag. 2018. p. 244-251. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-01201-4_26
Burlina, Philippe ; Joshi, N. ; Ng, Elise ; Billings, S. ; Rebman, A. ; Aucott, John. / Skin image analysis for erythema migrans detection and automated lyme disease referral. OR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis - 1st International Workshop, OR 2.0 2018 5th International Workshop, CARE 2018, 7th International Workshop, CLIP 2018, 3rd International Workshop, ISIC 2018 Held in Conjunction with MICCAI 2018. editor / Anand Malpani ; Marco A. Zenati ; Cristina Oyarzun Laura ; M. Emre Celebi ; Duygu Sarikaya ; Noel C. Codella ; Allan Halpern ; Marius Erdt ; Lena Maier-Hein ; Luo Xiongbiao ; Stefan Wesarg ; Danail Stoyanov ; Zeike Taylor ; Klaus Drechsler ; Kristin Dana ; Anne Martel ; Raj Shekhar ; Sandrine De Ribaupierre ; Tobias Reichl ; Jonathan McLeod ; Miguel Angel González Ballester ; Toby Collins ; Marius George Linguraru. Springer Verlag, 2018. pp. 244-251 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{c3b5b485f503423d9f41e85efb066ad5,
title = "Skin image analysis for erythema migrans detection and automated lyme disease referral",
abstract = "This study develops approaches for the automated referral of individuals with Lyme disease using erythema migrans rash (EM) images with clinical-grade or ‘in the wild’ characteristics. We develop a pre-screener using a Deep Convolutional Neural Network (DCNN) that classifies EM vs. other conditions, including either control/unaffected skin, or skin presenting with other confuser lesions. We test and report performance metrics for the proposed approach on this dataset including Cohen’s Kappa coefficient, area under the receiver operating characteristic (ROC) curve (AUC), accuracy, sensitivity, specificity. The machine classification yields accuracy (and error margin) of 93.04{\%} (1.49), AUC of 0.9504 (0.0156), and Kappa of 0.7549 (0.0586), which is a significant improvement over previously published state-of-the-art methods. Results also suggest substantial agreement between machine and expert clinician annotated gold standard images. The DCNN model developed for this skin classifier is made publicly available and can potentially be used by others for transfer learning to other types of skin lesion classification models including those for skin cancer.",
author = "Philippe Burlina and N. Joshi and Elise Ng and S. Billings and A. Rebman and John Aucott",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/978-3-030-01201-4_26",
language = "English (US)",
isbn = "9783030012007",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "244--251",
editor = "Anand Malpani and Zenati, {Marco A.} and {Oyarzun Laura}, Cristina and Celebi, {M. Emre} and Duygu Sarikaya and Codella, {Noel C.} and Allan Halpern and Marius Erdt and Lena Maier-Hein and Luo Xiongbiao and Stefan Wesarg and Danail Stoyanov and Zeike Taylor and Klaus Drechsler and Kristin Dana and Anne Martel and Raj Shekhar and {De Ribaupierre}, Sandrine and Tobias Reichl and Jonathan McLeod and {Gonz{\'a}lez Ballester}, {Miguel Angel} and Toby Collins and Linguraru, {Marius George}",
booktitle = "OR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis - 1st International Workshop, OR 2.0 2018 5th International Workshop, CARE 2018, 7th International Workshop, CLIP 2018, 3rd International Workshop, ISIC 2018 Held in Conjunction with MICCAI 2018",

}

TY - GEN

T1 - Skin image analysis for erythema migrans detection and automated lyme disease referral

AU - Burlina, Philippe

AU - Joshi, N.

AU - Ng, Elise

AU - Billings, S.

AU - Rebman, A.

AU - Aucott, John

PY - 2018/1/1

Y1 - 2018/1/1

N2 - This study develops approaches for the automated referral of individuals with Lyme disease using erythema migrans rash (EM) images with clinical-grade or ‘in the wild’ characteristics. We develop a pre-screener using a Deep Convolutional Neural Network (DCNN) that classifies EM vs. other conditions, including either control/unaffected skin, or skin presenting with other confuser lesions. We test and report performance metrics for the proposed approach on this dataset including Cohen’s Kappa coefficient, area under the receiver operating characteristic (ROC) curve (AUC), accuracy, sensitivity, specificity. The machine classification yields accuracy (and error margin) of 93.04% (1.49), AUC of 0.9504 (0.0156), and Kappa of 0.7549 (0.0586), which is a significant improvement over previously published state-of-the-art methods. Results also suggest substantial agreement between machine and expert clinician annotated gold standard images. The DCNN model developed for this skin classifier is made publicly available and can potentially be used by others for transfer learning to other types of skin lesion classification models including those for skin cancer.

AB - This study develops approaches for the automated referral of individuals with Lyme disease using erythema migrans rash (EM) images with clinical-grade or ‘in the wild’ characteristics. We develop a pre-screener using a Deep Convolutional Neural Network (DCNN) that classifies EM vs. other conditions, including either control/unaffected skin, or skin presenting with other confuser lesions. We test and report performance metrics for the proposed approach on this dataset including Cohen’s Kappa coefficient, area under the receiver operating characteristic (ROC) curve (AUC), accuracy, sensitivity, specificity. The machine classification yields accuracy (and error margin) of 93.04% (1.49), AUC of 0.9504 (0.0156), and Kappa of 0.7549 (0.0586), which is a significant improvement over previously published state-of-the-art methods. Results also suggest substantial agreement between machine and expert clinician annotated gold standard images. The DCNN model developed for this skin classifier is made publicly available and can potentially be used by others for transfer learning to other types of skin lesion classification models including those for skin cancer.

UR - http://www.scopus.com/inward/record.url?scp=85054840112&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85054840112&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-01201-4_26

DO - 10.1007/978-3-030-01201-4_26

M3 - Conference contribution

AN - SCOPUS:85054840112

SN - 9783030012007

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 244

EP - 251

BT - OR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis - 1st International Workshop, OR 2.0 2018 5th International Workshop, CARE 2018, 7th International Workshop, CLIP 2018, 3rd International Workshop, ISIC 2018 Held in Conjunction with MICCAI 2018

A2 - Malpani, Anand

A2 - Zenati, Marco A.

A2 - Oyarzun Laura, Cristina

A2 - Celebi, M. Emre

A2 - Sarikaya, Duygu

A2 - Codella, Noel C.

A2 - Halpern, Allan

A2 - Erdt, Marius

A2 - Maier-Hein, Lena

A2 - Xiongbiao, Luo

A2 - Wesarg, Stefan

A2 - Stoyanov, Danail

A2 - Taylor, Zeike

A2 - Drechsler, Klaus

A2 - Dana, Kristin

A2 - Martel, Anne

A2 - Shekhar, Raj

A2 - De Ribaupierre, Sandrine

A2 - Reichl, Tobias

A2 - McLeod, Jonathan

A2 - González Ballester, Miguel Angel

A2 - Collins, Toby

A2 - Linguraru, Marius George

PB - Springer Verlag

ER -