Objectives: While the first algorithms were based on Machine Learning, today's FotoFinder AI for pre-assessment of skin lesions uses most modern Deep Learning algorithms. The human capability to learn from examples and experiences is transferred to the computer. Artificial neural networks enable complex learning which is similar to biological learning processes.
Introduction: Anyone working with the Moleanalyzer pro AI Assistant will experience a new dimension in skin checks. Thanks to the significant help in the pre-assessment of melanocytic and non-melanocytic skin lesions, the expert system with its AI Scoring generates enthusiasm among physicians and patients. The specificity and sensitivity are impressive, as proven by numerous studies.
Together with the Bodyscan master software feature, the Moleanalyzer pro unleashes the full potential of FotoFinder Artificial Intelligence. The Total Body Dermoscopy workflow reduces the examination time for patients.
Materials / method: A market-approved CNN (Moleanalyzer-Pro, FotoFinder Systems) was used for binary classifications of 100 dermoscopic images of FSL. The same lesions were used in a two-level reader study including 64 dermatologists (level I: dermoscopy only; level II: dermoscopy, clinical close-up images, textual information). Primary endpoints were the CNN's sensitivity and specificity in comparison with the dermatologists' management decisions in level II. Generalizability of the CNN results was tested by using four additional external data sets.
Results: The CNN's sensitivity, specificity and ROC AUC were 96.2% [87.0%-98.9%], 68.8% [54.7%-80.1%] and 0.929 [0.880-0.978], respectively. In level II, the dermatologists' management decisions showed a mean sensitivity of 84.2% [82.2%-86.2%] and specificity of 69.4% [66.0%-72.8%]. When fixing the CNN's specificity at the dermatologists' mean specificity (69.4%), the CNN's sensitivity (96.2% [87.0%-98.9%]) was significantly higher than that of dermatologists (84.2% [82.2%-86.2%]; p < 0.001). Dermatologists of all training levels were outperformed by the CNN (all p < 0.001).
Conclusion: When applied as an assistant system, the CNN's higher sensitivity at an equivalent specificity may result in an improved early detection of face and scalp skin cancers.
Declaraciones
¿Ha recibido algún tipo de financiamiento para realizar su investigación sobre esta temática?
No
¿Ha recibido algún pago, honorario u otra compensación por su trabajo acerca de esta investigación?
No
¿Tiene vínculos financieros con alguna entidad que podría llegar a competir estrechamente con los medicamentos, materiales o instrumentos tratados en su investigación?
No
¿Posee o ha solicitado una patente relacionada a los materiales, productos o instrumentos utilizados en su investigación?
No
Este trabajo no cuenta con el apoyo de ningún financiamiento directo o indirecto. El autor asume plena responsabilidad sobre el mismo.
