RECPAD é a Conferência Nacional Portuguesa de Reconhecimento de Padrões, patrocinada pela APRP (Associação Portuguesa de Reconhecimento de Padrões). É uma conferência de um dia para promover a colaboração entre a comunidade científica portuguesa nas áreas de Reconhecimento de Padrões, Análise e Processamento de Imagens, Soft Computing e áreas afins. Este ano, o RECPAD 2019 será realizado, pela primeira vez, na Faculdade de Ciências da Universidade do Porto (FCUP) no próximo dia 31 de outubro de 2019.
A conferência contará com uma sessão aberta, gratuita, das 10h45 às 11h30, no Auditório Ferreira da Silva no edifício do DCC (FC6 0.30), na qual a ilustre oradora convidada do IEEE EMBS Carmen Poon fará uma apresentação intitulada: "AI-doscopist: A Deep Learning-based Model for Real-time Cancer Diagnosis during Therapeutic Endoscopy".
Short Bio:
Carmen Poon is assistant professor at the department of surgery of the Chinese University of Hong Kong and her expertise topics are mainly endoscopic informatics, image analysis, deep learning, wearable sensing and surgical robotics
Abstract:
"Therapeutic endoscopy is a cornerstone to screen and prevent colorectal cancers. Nevertheless, polyps can be missed or misdiagnosed during colonoscopy. By using large non-medical and endoscopic databases, Carmen have designed a deep-learning-based model, named “Artificial Endoscopist (a.k.a. AI-doscopist)”, to recognise colonic neoplasia during colonoscopy. AI-doscopist was trained and tested with over millions of images, labelled frame-by-frame with the presence, location, and type of colorectal polyps. Her initial study suggested that AI-doscpist can achieve similar detection rates as the endoscopists. Moreover, if AI-doscopist were to be used in real-time, it can potentially assist endoscopists in detecting one more patient with polyp in every 20-33 colonoscopies."