I nstitutskolloquium - Uniklinik RWTH Aachen
Transcription
I nstitutskolloquium - Uniklinik RWTH Aachen
Institutskolloquium Thema: Analysis of Breast Cancer Histopathological Images for Diagnosis using Deep Learning Referent: Ph.D G. Nagarajan, IIT Madras, Dept. of Applied Mechanics, Chennai, India Ort: Hörsaal 3 Ebene E, Uniklinik RWTH Aachen Pauwelsstr. 30, 52074 Aachen Zeit: Freitag, den 15.07.2016, von 14.00 – 15.30 Uhr Sprache: englisch Abstract: Cancer is a massive public health problem in the world. Breast cancer is the leading type of malignant tumor observed in women and the effective treatment depends on its early diagnosis. Histopathological image analysis is the gold standard in diagnosing all types of cancer and widely used for Breast cancer diagnosis. The breast cancer diagnosis, including grading and staging, is done by pathologists applying visual inspection of histological samples under microscope. Histopathological analysis is a highly timeconsuming specialized task. Existing techniques involve errors due to manual interpretation and variances due to subjective evaluation. Classification and segmentation of histopathological images into distinct histopathology pattern is the prior goal in computer aided image analysis for cancer diagnosis application. Histopathological images are prone to noise, uneven staining and variations in shape that makes automated analysis a challenging task. Deep learning can extract and organize the discriminative information from the images. In this work, an attempt is being made to comprehensively analyse and classify the breast cancer histopathological images using deep learning. In this seminar, the technological challenges, methods of addressing them and their clinical relevance would be discussed. The Institute of Electrical and Electronics Engineers Deutsche Sektion Joint Chapter EMB http://ewh.ieee.org/r8/germany/e mb/