This international, multi-centric randomized study shows how AI-assisted LC-OCTimproves the diagnostic accuracy of basal cell carcinoma (BCC) compared to traditional clinical and dermoscopic methods, particularly for clinicians withlimited LC-OCT experience. DOI: 10.1111/jdv.70099
Validated by Experts. Built on Real Large-Scale Data.
43 dermatologists across Europe
with varying levels of LC-OCT expertise (<6 months and >2 years).
200 equivocal skin lesions
traditional clinical evaluation or dermoscopy (C&D) only vs. C&D plus AI assisted LC-OCT
AI trained on > 600k images
hand labeled images coming from more than 1.000 patients across 5 different medical centers.
2-Year Gap Bridged by AI. CE-Marked.
95% diagnostic accuracy
for BCC diagnosis by experts using AI-assisted LC-OCT, demonstrating reliability in clinical practice.
+20 points in diagnostic accuracy compared to traditional methods.
Novices: 70% (C&D) to 91% (incl LC-OCT AI)
Experts: 75% (C&D) to 95% (incl LC-OCT AI)
CE-marked technology
Class IIa medical device according to European Regulation 2017/745.
Explore two real cases from the AI BCC study and see how accuracy improves across Clinical & Dermoscopy, LC-OCT, and LC-OCT + AI.
At every step, find out what percentage of dermatologists got it right!
1st Case - Is this a BCC, an AK/SCC, a dermal nevus, or healthy skin?
With the Clinical & Dermoscopy,79% of respondents identified the correct diagnostic. Click on the arrow to see the LC-OCT image.
With the LC-OCT image, 79% of respondents identified the correct diagnostic. Click on the arrow to see the LC-OCT + AI image.
LC-OCT+AI - 97,7% of respondents identified the correct diagnostic. The correct answer is: Actinic Keratosis (BCC score is 0%).
2nd Case - Is this a BCC, an AK/SCC, a dermal nevus, or healthy skin?
With the Clinical & Dermoscopy,53% of respondents identified the correct diagnostic. Click on the arrow to see the LC-OCT image.
With the LC-OCT, 60,5% of respondents identified the correct diagnostic. Click on the arrow to see the LC-OCT + AI image.
LC-OCT+AI - 95,3% of respondents identified the correct diagnostic. The correct answer is: BCC (the score reaches 100%).
The First Real-Time AI in Dermatologic Imaging.
The CE-marked deepLiveTM generates an attention heatmap overlaid on the LC-OCT image, color-coded from blue (0%)to yellow (100%), offering an intuitive visualisation of BCC prediction in real time.
Your AI-Powered Differential Diagnosis.
The AI doesn’t stop at identifying BCC - it also evaluates 6 potential imitators, assigning a probability score to each of them: Dermal nevus, Melanocytic lesions, Actinic Keratosis and Squamous Cell Carcinoma, Sebaceous Hyperplasia, healthy skin, and an "Other" category for various benign pathologies.
Know Your Margins In Vivo. Even Before Cutting.
Aim for reduced reoperation rates, fewer Mohs stages, Optimized lesion selection, Improved aesthetic outcomes. Discover how deepLive AITM helps delineate tumor margins in vivo, providing greater precision and confidence before surgery.
Step 1 - Margin assessment (left image) with LC-OCT combined with AI shows that margins at positions 3 to 6 should be extended(yellow marking), as residual BCC remains present (right). (Image Courtesy of Dr. François Habib)
Step 2 - Following the margin extension (left image), LC-OCT combined with AI confirmed that the margins are tumor-free(blue marking). The doctor can confidently proceed with surgery, ensuring the BCC tumor is completely removed. (Image Courtesy of Dr. François Habib)
WHITE PAPER In-Vivo Evaluation of BCC Margins
Discover how the CE-marked deepLiveTM AI, enhances diagnostic precision and surgical margin evaluation with the AI-assisted LC-OCT.
deepLive™ is a CE-marked Class IIa medical device according to European Regulation 2017/745. Manufactured in France and distributed by Damae Medical. Examinations performed with deepLive™ must be carried out exclusively by a healthcare professional experienced in dealing with the diagnosis and management of dermatologic pathologies, in conjunction with the totality of relevant clinical information about the patient. It is not for use as a standalone diagnostic. Users are strongly advised to refer to the user manual and the device labeling to ensure proper and safe use.
