Deep learning-based detection of motion artifacts in probe-based confocal laser endomicroscopy images

Aubreville, Marc and Stoeve, Maike and Oetter, Nicolai and Goncalves, Miguel and Knipfer, Christian and Neumann, Helmut and Bohr, Christopher and Stelzle, Florian and Maier, Andreas (2019) Deep learning-based detection of motion artifacts in probe-based confocal laser endomicroscopy images. INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY, 14 (1). pp. 31-42. ISSN 1861-6410, 1861-6429

Full text not available from this repository. (Request a copy)

Abstract

Purpose:Probe-based confocal laser endomicroscopy (pCLE) is a subcellular in vivo imaging technique capable of producing images that enable diagnosis of malign structural modifications in epithelial tissue. Images acquired with pCLE are, however, often tainted by significant artifacts that impair diagnosis. This is especially detrimental for automated image analysis, which is why said images are often excluded from recognition pipelines.MethodsWe present an approach for the automatic detection of motion artifacts in pCLE images and apply this methodology to a data set of 15 thousand images of epithelial tissue acquired in the oral cavity and the vocal folds. The approach is based on transfer learning from intermediate endpoints within a pre-trained Inception v3 network with tailored preprocessing. For detection within the non-rectangular pCLE images, we perform pooling within the activation maps of the network and evaluate this at different network depths.ResultsWe achieved area under the ROC curve values of 0.92 with the proposed method, compared to 0.80 for the best feature-based machine learning approach. Our overall accuracy with the presented approach is 94.8%.ConclusionOver traditional machine learning approaches with state-of-the-art features, we achieved significantly improved overall performance.

Item Type: Article
Uncontrolled Keywords: HEAD; MICROSCOPY; DIAGNOSIS; Deep convolutional neural networks; Confocal laser endomicroscopy; Motion artifact detection
Subjects: 600 Technology > 610 Medical sciences Medicine
Divisions: Medicine > Lehrstuhl für Hals-Nasen-Ohren-Heilkunde
Depositing User: Dr. Gernot Deinzer
Date Deposited: 22 Apr 2020 06:14
Last Modified: 22 Apr 2020 06:14
URI: https://pred.uni-regensburg.de/id/eprint/27876

Actions (login required)

View Item View Item