Related Experiment Video
Updated: Oct 28, 2025
Combining Reflectance Confocal Microscopy with Optical Coherence Tomography for Noninvasive Diagnosis of Skin Cancers via Image Acquisition
Published on: August 18, 2022
Deep Learning for Basal Cell Carcinoma Detection for Reflectance Confocal Microscopy.
Gabriele Campanella1, Cristian Navarrete-Dechent2, Konstantinos Liopyris3
1Department of Pathology, Memorial Sloan Kettering Cancer Center, New York, New York, USA; Graduate School of Medical Sciences, Weill Cornell Medicine, Cornell University, New York, New York, USA.
A new artificial intelligence model can automatically detect basal cell carcinoma (BCC) using reflectance confocal microscopy images. This AI shows expert-level accuracy, improving BCC diagnosis and potentially reducing unnecessary biopsies.
Area of Science:
- Dermatology
- Medical Imaging
- Artificial Intelligence
Background:
- Basal cell carcinoma (BCC) is the most common skin cancer, often diagnosed via visual inspection and biopsy.
- Current diagnostic methods for BCC lack specificity, leading to unnecessary procedures.
- Reflectance confocal microscopy (RCM) offers noninvasive cellular-level imaging, improving BCC diagnostic specificity.
Purpose of the Study:
- To develop and evaluate a deep learning-based artificial intelligence (AI) model for automated BCC detection in RCM images.
- To assess the diagnostic performance and generalizability of the AI model.
Main Methods:
- Development of a deep learning model for BCC detection using RCM images.
- Evaluation of the model's performance using receiver operating characteristic (ROC) curves at both stack and lesion levels.
- Validation of the model on an independent international test set.
Main Results:
- The AI model achieved high diagnostic accuracy, with an area under the curve (AUC) of 89.7% (stack level) and 88.3% (lesion level).
- The model demonstrated strong generalizability, achieving an AUC of 86.1% on an international test set.
- Performance was comparable to that of expert RCM diagnosticians.
Conclusions:
- AI-powered decision support systems can effectively detect BCC in RCM images.
- Automated BCC detection using AI has the potential to optimize skin cancer diagnosis and patient management.
- The developed AI model shows promise for clinical deployment in improving BCC evaluation.
