Review of Artifact Detection Methods for Automated Analysis and Diagnosis in Digital Pathology
Hossain Shakhawat, Sakir Hossain, Alamgir Kabir, Sakib Mahmud, M. M. Manjurul Islam, Faisal Tariq
Abstract
This chapter introduces the whole slide imaging system and the major artifacts found in whole slide image (WSI), addresses the challenges of automated analysis caused by the artifacts, and reviews the existing artifact detection methods. The WSI scanner is considered the key to digital pathology. In digital pathology, automated image analysis tools are utilized which automated analysis which can be broadly categorized into two types: Morphological image analysis and molecular image analysis. Artifacts that are produced at different stages of glass slide preparation are considered tissue artifacts and the artifacts produced while scanning the specimen are scanning artifacts. Scanning artifacts are generated mainly due to the hardware failure of the WSI scanner. The chapter reviews the artifact-detection methods by categorizing them into three groups based on the used technology as image enhancement-based approach, machine learning approach, and others. Markiefka et al. performed the diagnosis of deep learning-based classification method.