Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

3
92
2
29

Year Published

2020
2020
2024
2024

Publication Types

Select...
7
1

Relationship

1
7

Authors

Journals

citations
Cited by 146 publications
(126 citation statements)
references
References 19 publications
3
92
2
29
Order By: Relevance
“…Convolutional neural network (CNN) is a type of deep-learning algorithm that resembles the organization of the visual cortex. CNN models have advanced dramatically in recent years, ultimately being able to demonstrate physician-level diagnostic accuracy in a variety of medical fields such as dermatology, particularly skin cancers [1][2][3][4][5][6][7][8][9]. However, most studies have had a retrospective design and whether these data can be reproduced in a real clinical setting has not been assessed in prospective studies.…”
Section: Introductionmentioning
confidence: 99%
“…Convolutional neural network (CNN) is a type of deep-learning algorithm that resembles the organization of the visual cortex. CNN models have advanced dramatically in recent years, ultimately being able to demonstrate physician-level diagnostic accuracy in a variety of medical fields such as dermatology, particularly skin cancers [1][2][3][4][5][6][7][8][9]. However, most studies have had a retrospective design and whether these data can be reproduced in a real clinical setting has not been assessed in prospective studies.…”
Section: Introductionmentioning
confidence: 99%
“…CNNs have classified dermoscopic images of melanoma versus nevi with performances similar to or exceeding dermatologists (Brinker et al, 2019a(Brinker et al, , 2019bCodella et al, 2016;Haenssle et al, 2020Haenssle et al, , 2018Marchetti et al, 2018Marchetti et al, , 2020Phillips et al, 2019;Tschandl et al, 2019b). CNNs have also achieved expert-level diagnosis of nonpigmented skin cancer (Tschandl et al, 2019c) and outperformed dermatologists across five disease classes (Maron et al, 2019).…”
Section: Dermoscopic Imagesmentioning
confidence: 98%
“…It is unknown how CNNs perform compared with dermatologists making face-to-face assessments because studies report dermatologist-level diagnostic accuracy based on clinician evaluations of images in an artificial setting, using curated images, and without providing the full complement of meta-data normally available in clinic and teledermatology settings. Dermatologists improved their diagnostic accuracy when given access to close-up images and limited clinical information such as age, sex, and body site (Haenssle et al, 2020).…”
Section: Emerging Applicationsmentioning
confidence: 99%
See 1 more Smart Citation
“…The latter diagnostic method has a statistically lower error rate and, therefore, most existing work in digital skin diagnosis attaches great importance to a biopsy-verified test set. However, due to the low availability and the high costs of acquiring biopsy-verified images a large amount of non-biopsy-verified images are often included in the training set whose diagnosis are based solely on a consensus decision of several dermatologists or on the lack of temporal lesion changes over several skin examinations (3,8,(11)(12)(13). In doing so, high label noise is introduced in the modeling process.…”
Section: Introductionmentioning
confidence: 99%