Summary
Deep learning algorithms are applied on optical coherence tomography (OCT) with cellular resolution for skin cancer detection. Results on animal model show >90% accuracy for the classification of normal, dysplasia, and squamous cell carcinoma. An OCT engineering model has been developed by the Apollo Medical Optics. Clinical collaborations on real-time diagnosis of skin or cornea diseases/cancers are welcome.
 
Applications
Point of care testing on the following clinics:
Dermatology: skin cancer detection, volumetric nuclear cytoplasmic ratio analysis
Ophthalmology: corneal subbasal nerve plexus mapping and feature extraction
 
Advantages
Non-invasiveness, real time, and cellular resolution are achieved for in-vivo biomedical imaging
 
Related Links
National Taiwan University:
http://www.ee.ntu.edu.tw/bio1?teacher_id=941019&p=3
Apollo Medical Optics:
http://www.apollomedicaloptics.com/index.php
 
Keywords
Cellular-resolution optical coherence tomography, deep learning algorithm, skin cancers, cornea nerve imaging