TY - GEN
T1 - Skin image analysis for erythema migrans detection and automated lyme disease referral
AU - Burlina, Philippe
AU - Joshi, N.
AU - Ng, E.
AU - Billings, S.
AU - Rebman, A.
AU - Aucott, J.
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
T2 - 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
Y2 - 16 September 2018 through 20 September 2018
ER -