Press & Events
Project results short videos series: Optical Coherence Tomography (OCT) Classifier
A deep learning model for the automatic classification of murine OCT images as benign (healthy & hyperplasic) versus malignant (adenomatous & adenocarcinoma) has been developed at the PICCOLO project. The algorithm has been trained, validated and tested with a preliminary version of the Murine model ex-vivo OCT database also produced at the project. The evaluation over individual B-scan images reports 96.95±1.4% sensitivity, 80.94±15% specificity and over C-scan volumes 98.21±1.90% sensitivity and 78,65±20% specificity.
These results confirm that the combination of this novel imaging technology together with deep learning algorithms leads the foundations for the development of real-time optical biopsy applications. Besides, it can be combined with other complementary novel imaging approaches as the MPM Classifier for improved diagnosis.
The details of this output will be openly accessible through publications in scientific journals at the scientific community´s disposal.
Tecnalia Research & Innovation (TECNALIA)