Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images
MetadataShow full item record
The automatic analysis of digital pathology images is becoming of increasing interest for the development of novel therapeutic drugs and of the associated companion diagnostic tests in oncology. A precise quantification of the tumor microenvironment and therefore an accurate segmentation of the tumor extend are critical in this context. In this paper, we present a new approach based on visual context Random Forest to generate high resolution segmentation maps from Deep Learning coarse segmentation maps. Through an example inimmunofluorescence, we show that this method enables an accurate and fast detection of the tumor structures in terms of qualitative and quantitative evaluation against three baseline approaches. For the method to be resilient to the high variability of staining intensity, a novel locally adaptive normalization algorithm is moreover introduced.
Brieu , N , Gavriel , C , Harrison , D J , Caie , P D & Schmidt , G 2018 , Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images . in J E Tomaszewski & M N Gurcan (eds) , Medical Imaging 2018 : Digital Pathology . , 105810P , Proceedings of SPIE , vol. 10581 , SPIE , Symposium: SPIE Medical Imaging , Houston , United States , 10/02/18 . DOI: 10.1117/12.2292794conference
Medical Imaging 2018
© 2018, SPIE. This work has been made available online in accordance with the publisher’s policies. This is the final published version of the work, which was originally published at https://doi.org/10.1117/12.2292794
Items in the St Andrews Research Repository are protected by copyright, with all rights reserved, unless otherwise indicated.