Academic Bibliography

 Search 200 years of publications by Ghent University researchers.
Advanced search
1 file | 1.69 MB
Add to list
  1. Search results
  2. Dermatologist versus artificial intel...

Dermatologist versus artificial intelligence confidence in dermoscopy diagnosis : complementary information that may affect decision-making

Pieter Van Molle (UGent) , Sofie Mylle (UGent) , Tim Verbelen (UGent) , Cedric De Boom, Bert Vankeirsbilck (UGent) , Evelien Verhaeghe (UGent) , Bart Dhoedt (UGent) and Lieve Brochez (UGent)
(2023) EXPERIMENTAL DERMATOLOGY. 32(10). p.1744-1751
Author
Organization
Abstract
In dermatology, deep learning may be applied for skin lesion classification. However, for a given input image, a neural network only outputs a label, obtained using the class probabilities, which do not model uncertainty. Our group developed a novel method to quantify uncertainty in stochastic neural networks. In this study, we aimed to train such network for skin lesion classification and evaluate its diagnostic performance and uncertainty, and compare the results to the assessments by a group of dermatologists. By passing duplicates of an image through such a stochastic neural network, we obtained distributions per class, rather than a single probability value. We interpreted the overlap between these distributions as the output uncertainty, where a high overlap indicated a high uncertainty, and vice versa. We had 29 dermatologists diagnose a series of skin lesions and rate their confidence. We compared these results to those of the network. The network achieved a sensitivity and specificity of 50% and 88%, comparable to the average dermatologist (respectively 68% and 73%). Higher confidence/less uncertainty was associated with better diagnostic performance both in the neural network and in dermatologists. We found no correlation between the uncertainty of the neural network and the confidence of dermatologists (R = -0.06, p = 0.77). Dermatologists should not blindly trust the output of a neural network, especially when its uncertainty is high. The addition of an uncertainty score may stimulate the human-computer interaction.
Keywords
PERFORMANCE, computer vision, deep learning, neural networks, skin lesion classification, uncertainty

Downloads

  • Download
    (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.69 MB

Citation

Please use this url to cite or link to this publication:

MLA
Van Molle, Pieter, et al. “Dermatologist versus Artificial Intelligence Confidence in Dermoscopy Diagnosis : Complementary Information That May Affect Decision-Making.” EXPERIMENTAL DERMATOLOGY, vol. 32, no. 10, 2023, pp. 1744–51, doi:10.1111/exd.14892.
APA
Van Molle, P., Mylle, S., Verbelen, T., De Boom, C., Vankeirsbilck, B., Verhaeghe, E., … Brochez, L. (2023). Dermatologist versus artificial intelligence confidence in dermoscopy diagnosis : complementary information that may affect decision-making. EXPERIMENTAL DERMATOLOGY, 32(10), 1744–1751. https://doi.org/10.1111/exd.14892
Chicago author-date
Van Molle, Pieter, Sofie Mylle, Tim Verbelen, Cedric De Boom, Bert Vankeirsbilck, Evelien Verhaeghe, Bart Dhoedt, and Lieve Brochez. 2023. “Dermatologist versus Artificial Intelligence Confidence in Dermoscopy Diagnosis : Complementary Information That May Affect Decision-Making.” EXPERIMENTAL DERMATOLOGY 32 (10): 1744–51. https://doi.org/10.1111/exd.14892.
Chicago author-date (all authors)
Van Molle, Pieter, Sofie Mylle, Tim Verbelen, Cedric De Boom, Bert Vankeirsbilck, Evelien Verhaeghe, Bart Dhoedt, and Lieve Brochez. 2023. “Dermatologist versus Artificial Intelligence Confidence in Dermoscopy Diagnosis : Complementary Information That May Affect Decision-Making.” EXPERIMENTAL DERMATOLOGY 32 (10): 1744–1751. doi:10.1111/exd.14892.
Vancouver
1.
Van Molle P, Mylle S, Verbelen T, De Boom C, Vankeirsbilck B, Verhaeghe E, et al. Dermatologist versus artificial intelligence confidence in dermoscopy diagnosis : complementary information that may affect decision-making. EXPERIMENTAL DERMATOLOGY. 2023;32(10):1744–51.
IEEE
[1]
P. Van Molle et al., “Dermatologist versus artificial intelligence confidence in dermoscopy diagnosis : complementary information that may affect decision-making,” EXPERIMENTAL DERMATOLOGY, vol. 32, no. 10, pp. 1744–1751, 2023.
@article{01H9JAARDQ7N7KAJ1M2053CFPE,
  abstract     = {{In dermatology, deep learning may be applied for skin lesion classification. However, for a given input image, a neural network only outputs a label, obtained using the class probabilities, which do not model uncertainty. Our group developed a novel method to quantify uncertainty in stochastic neural networks. In this study, we aimed to train such network for skin lesion classification and evaluate its diagnostic performance and uncertainty, and compare the results to the assessments by a group of dermatologists. By passing duplicates of an image through such a stochastic neural network, we obtained distributions per class, rather than a single probability value. We interpreted the overlap between these distributions as the output uncertainty, where a high overlap indicated a high uncertainty, and vice versa. We had 29 dermatologists diagnose a series of skin lesions and rate their confidence. We compared these results to those of the network. The network achieved a sensitivity and specificity of 50% and 88%, comparable to the average dermatologist (respectively 68% and 73%). Higher confidence/less uncertainty was associated with better diagnostic performance both in the neural network and in dermatologists. We found no correlation between the uncertainty of the neural network and the confidence of dermatologists (R = -0.06, p = 0.77). Dermatologists should not blindly trust the output of a neural network, especially when its uncertainty is high. The addition of an uncertainty score may stimulate the human-computer interaction.}},
  author       = {{Van Molle, Pieter and Mylle, Sofie and Verbelen, Tim and De Boom, Cedric and Vankeirsbilck, Bert and Verhaeghe, Evelien and Dhoedt, Bart and Brochez, Lieve}},
  issn         = {{0906-6705}},
  journal      = {{EXPERIMENTAL DERMATOLOGY}},
  keywords     = {{PERFORMANCE,computer vision,deep learning,neural networks,skin lesion classification,uncertainty}},
  language     = {{eng}},
  number       = {{10}},
  pages        = {{1744--1751}},
  title        = {{Dermatologist versus artificial intelligence confidence in dermoscopy diagnosis : complementary information that may affect decision-making}},
  url          = {{http://doi.org/10.1111/exd.14892}},
  volume       = {{32}},
  year         = {{2023}},
}
Altmetric
View in Altmetric
Web of Science
Times cited: