Détection du carcinome basocellulaire dans des images OCT plein champ utilisant un réseau de neurones convolutif
- Resource Type
- Authors
- John R. Durkin; Vannary Meas-Yedid; Jean-Christophe Olivo-Marin; C. Boceara; Diana Mandache; E. Dalimier
- Source
- IEEE 15th International Symposium on Biomedical Imaging
IEEE 15th International Symposium on Biomedical Imaging, Apr 2018, Washington, United States. ⟨10.1109/ISBI.2018.8363689⟩
ISBI
- Subject
- Computer science
medicine.medical_treatment
02 engineering and technology
Convolutional neural network
[INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI]
03 medical and health sciences
0302 clinical medicine
Optical coherence tomography
convolutional neural networks
0202 electrical engineering, electronic engineering, information engineering
medicine
Mohs surgery
[INFO.INFO-IM]Computer Science [cs]/Medical Imaging
Basal cell carcinoma
030212 general & internal medicine
Generalized normal distribution
medicine.diagnostic_test
business.industry
Deep learning
Digital pathology
020207 software engineering
Pattern recognition
Optical Biopsy
medicine.disease
Cancer cell
nonmelanoma skin cancer
Artificial intelligence
business
digital pathology
Classifier (UML)
full field optical coherence tomography
- Language
- English
International audience; In this paper we introduce a new application that exploits the emerging imaging modality of full field optical coherence tomography (FFOCT) as a means of optical biopsy. The objective is to build a computer-aided diagnosis (CAD) tool that can speed up the detection of tumoral areas in skin excisions resulting from Mohs surgery. Since there is little prior knowledge about the appearance of cancer cell morphology in this type of imagery, deep learning techniques are applied. Using convolutional neural networks (CNN), we train a feature extractor able to find representative characteristics for FFOCT data and a classifier that learns a generalized distribution of the data. With a dataset of 40 high-resolution images, we obtained a classification accuracy of 95.93%.