| 1. |
- Haenssle, H A, et al.
(författare)
-
Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists.
- 2018
-
Ingår i: Annals of Oncology. - : Elsevier BV. - 1569-8041 .- 0923-7534. ; 29:8, s. 1836-1842
-
Tidskriftsartikel (refereegranskat)abstract
- Deep learning convolutional neural networks (CNN) may facilitate melanoma detection, but data comparing a CNN's diagnostic performance to larger groups of dermatologists are lacking.Google's Inception v4 CNN architecture was trained and validated using dermoscopic images and corresponding diagnoses. In a comparative cross-sectional reader study a 100-image test-set was used (level-I: dermoscopy only; level-II: dermoscopy plus clinical information and images). Main outcome measures were sensitivity, specificity and area under the curve (AUC) of receiver operating characteristics (ROC) for diagnostic classification (dichotomous) of lesions by the CNN versus an international group of 58 dermatologists during level-I or -II of the reader study. Secondary end points included the dermatologists' diagnostic performance in their management decisions and differences in the diagnostic performance of dermatologists during level-I and -II of the reader study. Additionally, the CNN's performance was compared with the top-five algorithms of the 2016 International Symposium on Biomedical Imaging (ISBI) challenge.In level-I dermatologists achieved a mean (±standard deviation) sensitivity and specificity for lesion classification of 86.6% (±9.3%) and 71.3% (±11.2%), respectively. More clinical information (level-II) improved the sensitivity to 88.9% (±9.6%, P=0.19) and specificity to 75.7% (±11.7%, P<0.05). The CNN ROC curve revealed a higher specificity of 82.5% when compared with dermatologists in level-I (71.3%, P<0.01) and level-II (75.7%, P<0.01) at their sensitivities of 86.6% and 88.9%, respectively. The CNN ROC AUC was greater than the mean ROC area of dermatologists (0.86 versus 0.79, P<0.01). The CNN scored results close to the top three algorithms of the ISBI 2016 challenge.For the first time we compared a CNN's diagnostic performance with a large international group of 58 dermatologists, including 30 experts. Most dermatologists were outperformed by the CNN. Irrespective of any physicians' experience, they may benefit from assistance by a CNN's image classification.This study was registered at the German Clinical Trial Register (DRKS-Study-ID: DRKS00013570; https://www.drks.de/drks_web/).
|
|
| 3. |
- Menzies, Scott W, et al.
(författare)
-
Dermoscopic Evaluation of Nodular Melanoma.
- 2013
-
Ingår i: JAMA dermatology (Chicago, Ill.). - : American Medical Association (AMA). - 2168-6084 .- 2168-6068. ; 149:6, s. 699-709
-
Tidskriftsartikel (refereegranskat)abstract
- IMPORTANCE Nodular melanoma (NM) is a rapidly progressing potentially lethal skin tumor for which early diagnosis is critical. OBJECTIVE To determine the dermoscopy features of NM. DESIGN Eighty-three cases of NM, 134 of invasive non-NM, 115 of nodular benign melanocytic tumors, and 135 of nodular nonmelanocytic tumors were scored for dermoscopy features using modified and previously described methods. Lesions were separated into amelanotic/hypomelanotic or pigmented to assess outcomes. SETTING Predominantly hospital-based clinics from 5 continents. MAIN OUTCOME MEASURES Sensitivity, specificity, and odds ratios for features/models for the diagnosis of melanoma. RESULTS Nodular melanoma occurred more frequently as amelanotic/hypomelanotic (37.3%) than did invasive non-NM (7.5%). Pigmented NM had a more frequent (compared with invasive non-NM; in descending order of odds ratio) symmetrical pigmentation pattern (5.8% vs 0.8%), large-diameter vessels, areas of homogeneous blue pigmentation, symmetrical shape, predominant peripheral vessels, blue-white veil, pink color, black color, and milky red/pink areas. Pigmented NM less frequently displayed an atypical broadened network, pigment network or pseudonetwork, multiple blue-gray dots, scarlike depigmentation, irregularly distributed and sized brown dots and globules, tan color, irregularly shaped depigmentation, and irregularly distributed and sized dots and globules of any color. The most important positive correlating features of pigmented NM vs nodular nonmelanoma were peripheral black dots/globules, multiple brown dots, irregular black dots/globules, blue-white veil, homogeneous blue pigmentation, 5 to 6 colors, and black color. A model to classify a lesion as melanocytic gave a high sensitivity (>98.0%) for both nodular pigmented and nonnodular pigmented melanoma but a lower sensitivity for amelanotic/hypomelanotic NM (84%). A method for diagnosing amelanotic/hypomelanotic malignant lesions (including basal cell carcinoma) gave a 93% sensitivity and 70% specificity for NM. CONCLUSIONS AND RELEVANCE When a progressively growing, symmetrically patterned melanocytic nodule is identified, NM needs to be excluded.
|
|
