EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Classifying real-world macroscopic images in the primary–secondary care interface using transfer learning: implications for development of artificial intelligence solutions using nondermoscopic images

Jacob Carse, Tamás Süveges, Gillian Chin, Shareen Muthiah, Colin Morton, Charlotte Proby iD, Emanuele Trucco, Colin Fleming, Stephen McKenna iD

Clinical and Experimental Dermatology · 25.06.2024 · 49 (7) · 699-706

DOI: 10.1093/ced/llad400 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Clinical and Experimental Dermatology, online erschienen am 22.11.2023. Verfasst von Jacob Carse und 8 weiteren Autor:innen. Beteiligte Einrichtung: CVIP (Computer Vision and Image Processing), School of Science and Engineering, University of Dundee, Dundee, UK. Hauptthemen laut MEDLINE: Secondary Health Care, Deep Learning, Dermoscopy. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 21 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Institute for Health and Care Research, NHS Transformation Directorate, Department of Health and Social Care.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

The application of deep learning (DL) to diagnostic dermatology has been the subject of numerous studies, with some reporting skin lesion classification performance on curated datasets comparable to that of experienced dermatologists. Most skin disease images encountered in clinical settings are macroscopic, without dermoscopic information, and exhibit considerable variability. Further research is necessary to determine the generalizability of DL algorithms across populations and acquisition settings.

Objectives

To assess the extent to which DL can generalize to nondermoscopic datasets acquired at the primary–secondary care interface in the National Health Service (NHS); to explore how to obtain a clinically satisfactory performance on nonstandardized, real-world local data without the availability of large diagnostically labelled local datasets; and to measure the impact of pretraining DL algorithms on external, public datasets.

Methods

Diagnostic macroscopic image datasets were created from previous referrals from primary to secondary care. These included 2213 images referred from primary care practitioners in NHS Tayside and 1510 images from NHS Forth Valley acquired by medical photographers. Two further datasets with identical diagnostic labels were obtained from sources in the public domain, namely the International Skin Imaging Collaboration (ISIC) dermoscopic dataset and the SD-260 nondermoscopic dataset. DL algorithms, specifically EfficientNets and Self-attention with Window-wise Inner-product based Network (SWIN) transformers, were trained using data from each of these datasets. Algorithms were also fine-tuned on images from the NHS datasets after pretraining on different data combinations, including the larger public-domain datasets. Receiver operating characteristic curves and the area under such curves (AUC) were used to assess performance.

Results

SWIN transformers tested on Forth Valley data had AUCs of 0.85 and 0.89 when trained on SD-260 and Forth Valley data, respectively. Training on SD-260 followed by fine tuning of Forth Valley data gave an AUC of 0.91. Similar effects of pretraining and tuning on local data were observed using Tayside data and EfficientNets. Pretraining on the larger dermoscopic image dataset (ISIC 2019) provided no additional benefit.

Conclusions

Pretraining on public macroscopic images followed by tuning to local data gave promising results. Further improvements are needed to afford deployment in real clinical pathways. Larger datasets local to the target domain might be expected to yield further improved performance.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Secondary Health CareDeep LearningDermoscopyHumansArtificial IntelligenceSkin DiseasesPrimary Health CareAlgorithmsState MedicineDermatology

Autor:innen und Institutionen

  • CVIP (Computer Vision and Image Processing), School of Science and Engineering, University of Dundee, Dundee, UK.
  • Department of Dermatology, Ninewells Hospital and Medical School, Dundee, UK.
  • Department of Dermatology, Forth Valley Dermatology Centre, Stirling, UK.
  • School of Medicine, Ninewells Hospital and Medical School, University of Dundee, Dundee, UK.

ORCID-Profile: Charlotte Proby, Stephen McKenna

Förderung

  • National Institute for Health and Care Research
  • NHS Transformation Directorate
  • Department of Health and Social Care
  • UK Engineering and Physical Sciences Research Council

Zitieren

APA

Carse, J., Süveges, T., Chin, G., Muthiah, S., Morton, C., Proby, C., Trucco, E., Fleming, C., & McKenna, S. (2024). Classifying real-world macroscopic images in the primary–secondary care interface using transfer learning: implications for development of artificial intelligence solutions using nondermoscopic images. Clinical and Experimental Dermatology, 49(7), 699-706. https://doi.org/10.1093/ced/llad400

Vancouver

Carse J, Süveges T, Chin G, Muthiah S, Morton C, Proby C, et al. Classifying real-world macroscopic images in the primary–secondary care interface using transfer learning: implications for development of artificial intelligence solutions using nondermoscopic images. Clinical and Experimental Dermatology. 2024;49(7):699-706. doi:10.1093/ced/llad400

BibTeX
@article{Carse2024,
  author = {Carse, Jacob and Süveges, Tamás and Chin, Gillian and Muthiah, Shareen and Morton, Colin and Proby, Charlotte and Trucco, Emanuele and Fleming, Colin and McKenna, Stephen},
  title = {Classifying real-world macroscopic images in the primary–secondary care interface using transfer learning: implications for development of artificial intelligence solutions using nondermoscopic images},
  journal = {Clinical and Experimental Dermatology},
  year = {2024},
  volume = {49},
  number = {7},
  pages = {699--706},
  doi = {10.1093/ced/llad400}
}
Kostenlos starten →