Algorithmic framework for HEp-2 fluorescence pattern classification to aid auto-immune diseases diagnosis

Abstract

Fluorescence microscopy allows the acquisition of the spectroscopic properties of fluorescent reporter molecules at levels of resolution too small to be seen with the naked eye. The Indirect Immune Fluorescence Test is the method used to identify antinuclear antibodies. The main principle of this method is to identify the auto-antibodies in a patient's blood serum by staining affected cell structures. The resulting auto-antibody specific fluorescence patterns can be visualized by a fluorescence microscope and examined by a physician to determine a diagnosis. More than 30 different nuclear and cytoplasmic fluorescence patterns are known, which are characterized by a set of a 100 different auto-antibodies. The quality of a suspicion diagnosis strongly depends on the experience of the physicians and, as such, can be very subjective. This paper focuses on the development and evaluation of image processing and classification algorithms for HEp-2 Cell segmentation and cell type classification in order to better detect a suspicion diagnosis for auto-immune diseases.

HEp-2 cell pattern segmentation for the support of autoimmune disease diagnosis

“…The cause of all auto-immune diseases is an unwanted immune response against human cell structures being wrongly interpreted and fought as foreign bodies. Often the course of the disease is chronic and leads to seriously damaging tissue and organs [4]. Examples of such autoimmune diseases are Rheumatoid Arthritis, Multiple Sclerosis and Diabetes mellitus type 1.…”

Design of a Computer-Assisted System to Automatically Detect Cell Types Using ANA IIF Images for the Diagnosis of Autoimmune Diseases

“…It was shown that the segmentation performance of the proposed method achieved 89 % when evaluated based on percent volume overlap (PVO) rate [13,14]. The classification performance using a support vector machine (SVM) [15,16] classifier designed based on the features calculated from the segmented cells achieved an average accuracy of 96.90 %, outperforming the methods proposed in other previous studies [1,[5][6][7][8][9][10].…”

“…Generally, these systems include a module for image segmentation designed based on Otsu thresholding [1][2][3] and watershed segmentation [4], and another module for image recognition built based on various pattern recognition methods [1,[5][6][7][8][9][10].…”

Automatic Classification of Human Epithelial Type 2 Cell Indirect Immunofluorescence Images using Cell Pyramid Matching

“…Perner et al [23] Textural Decision Tree Hiemann et al [15] Structural; textural LogisticModel Tree Elbischger et al [8] Image statistics; cell shape; textural Nearest Neighbour (NN) Hsieh et al [16] Image statistics; textural Learning Vector Quantisation (LVQ) Soda et al [34] Specific set of features (e.g. textural) for each class Multi Expert System Cordelli et al [7] Image statistics; textural; morphological AdaBoost Strandmark et al [35] Morphological; image statistics; textural Random Forest Ali et al [2] Biological-Inspired Descriptor Boosted k-NN Classifier Theodorakopoulos et al [36] Morphological and texture features Kernel SVM (KSVM) Thibault et al [37] Morphological and texture features Linear Regression, Random Forest Ghosh et al [13] Histograms of Oriented Gradients, SVM image statistics and textural Li et al [19] Textural and image statistics SVM Di Cataldo et al [5] GLCM and DCT features SVM Snell et al [33] Texture and shape Multistage classifier Ersoy et al [9] Local shape measures, gradient and textural ShareBoost Wiliem et al [41] Bag of visual words with dual-region structure Nearest Convex Hull Classifier (NCH) and apply an automated feature selection process [15].…”