Image Enhancement Techniques for Automated Histopathological Analysis

Computer-assisted analysis of the tissue biopsy samples is used in order to reduce the time consumed by pathologists while determining the results manually. This has reduced the chances of inaccurate results due to human error thus improving accuracy achieved in detection and subsequent classification of a given sample. However, color consistency in light microscopy-based histology is an increasingly important problem with the advent of Gigapixel digital slide scanners and automatic image analysis. A variety of external problems such as weather, sunlight, dust, improper handling of the physical slides or noise introduced in the image can cause distortion of vital information. As a result, numerous image enhancement techniques have been developed to remove noise and enhance the images in order to recover the lost information. This paper provides a quantitative and qualitative analysis of a few image enhancement techniques applied on IHC stained TILs reinfused breast tissue images that use filtering and background correction using morphological structuring elements to rectify the distortion and provide enhanced images as an output which subsequently will give better results in detection and classification algorithms.