Header

UZH-Logo

Maintenance Infos

Suspicious Skin Lesion Detection in Wide-Field Body Images using Deep Learning Outlier Detection


Garcia, Javier Barranco; Tanadini-Lang, Stephanie; Andratschke, Nicolaus; Gassner, Mathia; Braun, Ralph P (2022). Suspicious Skin Lesion Detection in Wide-Field Body Images using Deep Learning Outlier Detection. In: 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), Glasgow, Scotland, United Kingdom, 11 July 2022 - 15 July 2022. Institute of Electrical and Electronics Engineers, 2928-2932.

Abstract

During consultation dermatologists have to address hundreds of lesions in a limited amount of time. They will not only evaluate the single lesion of interest but more importantly the context of it. Visually comparing the similarity of the majority of lesions within the same patient provides a strong indication for lesions with significantly differing aspects. Deep learning algorithms are capable to identify such outliers, i.e. images that differ considerably from the expected appearance on a larger cohort, and highlight the main differences in those cases. In the present study we evaluate the use of autoencoders as unsupervised tools to detect suspicious skin lesions based on evaluation of real world data acquired during consultation at the USZ Dermatology Clinic. Clinical Relevance— Deep learning algorithms are showing many promising results in dermatology lesion classification. However the context of the lesion is normally not considered in the analysis which prevents these tools to transition into routine practice. An outlier detector based on real world data would allow a dermatologist or general practitioner to detect the suspicious lesions for further examination. The algorithm would additionally provide useful insights by highlighting the feature differences between the original outlier (malignant lesion) and the lesion reconstructed by the autoencoder

Abstract

During consultation dermatologists have to address hundreds of lesions in a limited amount of time. They will not only evaluate the single lesion of interest but more importantly the context of it. Visually comparing the similarity of the majority of lesions within the same patient provides a strong indication for lesions with significantly differing aspects. Deep learning algorithms are capable to identify such outliers, i.e. images that differ considerably from the expected appearance on a larger cohort, and highlight the main differences in those cases. In the present study we evaluate the use of autoencoders as unsupervised tools to detect suspicious skin lesions based on evaluation of real world data acquired during consultation at the USZ Dermatology Clinic. Clinical Relevance— Deep learning algorithms are showing many promising results in dermatology lesion classification. However the context of the lesion is normally not considered in the analysis which prevents these tools to transition into routine practice. An outlier detector based on real world data would allow a dermatologist or general practitioner to detect the suspicious lesions for further examination. The algorithm would additionally provide useful insights by highlighting the feature differences between the original outlier (malignant lesion) and the lesion reconstructed by the autoencoder

Statistics

Citations

Dimensions.ai Metrics

Altmetrics

Downloads

3 downloads since deposited on 16 Nov 2022
0 downloads since 12 months
Detailed statistics

Additional indexing

Item Type:Conference or Workshop Item (Paper), refereed, original work
Communities & Collections:04 Faculty of Medicine > University Hospital Zurich > Clinic for Radiation Oncology
04 Faculty of Medicine > University Hospital Zurich > Dermatology Clinic
Dewey Decimal Classification:610 Medicine & health
Scopus Subject Areas:Physical Sciences > Signal Processing
Physical Sciences > Biomedical Engineering
Physical Sciences > Computer Vision and Pattern Recognition
Health Sciences > Health Informatics
Language:English
Event End Date:15 July 2022
Deposited On:16 Nov 2022 12:34
Last Modified:19 Jul 2023 11:22
Publisher:Institute of Electrical and Electronics Engineers
Series Name:IEEE Engineering in Medicine and Biology Society. Annual International Conference Proceedings
ISSN:2375-7477
OA Status:Closed
Publisher DOI:https://doi.org/10.1109/embc48229.2022.9871655
PubMed ID:36085609