Objective: Atherosclerotic plaques have a complex composition, consisting of inflammation, fibrosis, cholesterol crystals, hemorrhage, and/or calcification. The segmentation and quantification of plaque features in histopathology images form the foundation for studies evaluating plaque instability and the mechanisms that underlie the atherosclerotic process. Manual segmentation of plaque features from histology images is a tedious, time-consuming, and subjective visual recognition task. Herein, we present a fully automatic approach using state-of-the-art deep learning techniques to identify three major features of the atherosclerotic plaque: calcification, lipid core, and fibrosis. Methods: Plaques (n=70) were collected from patients who underwent a carotid endarterectomy at McGill University-affiliated hospitals. Hematoxylin and Eosin-stained sections were obtained from the region with the largest plaque burden. The “ground truth annotations” for lipid core, calcification, and fibrosis were performed manually by three blinded cardiovascular pathologists, using Sedeen Viewer. A total of 23,000 patches with 512x512 pixel size were extracted from our image dataset, and divided into train, validate, and test sets. Using Transfer Learning, multi-class U-Net models for semantic segmentation were trained on the patches to extract fibrosis, lipid, and calcification plaque features. Evaluation of model performance was based on the mean value of the Intersection over Union (Mean-IOU) between the prediction results and the “ground truth annotations”. Results: Our models resulted in an overall performance of 77% for test images, and a per-class performance for the three plaque features: fibrosis = 0.77±0.2, lipid core = 0.80±0.3, calcification = 0.75±0.25. However, a qualitative evaluation by the pathologists confirmed that the prediction results in fact outperformed the “ground truth annotations”, and detected non-annotated regions. Conclusion: To our knowledge, this is a first attempt at developing a fully automatic approach for atherosclerotic plaque feature segmentation from histology images. Our models can accelerate atherosclerosis research, by improving the speed, quality, and reproducibility of plaque analysis.

Harrison's principles of internal medicine [2] Automated cell counter device (principles, calibration, quality control and error) [3] Abnormal red blood cells detection using adaptive neuro-fuzzy system [4] Digital image processing [5] A tlreshold selection method from gray-level histograms [6] Morphological image analysis: Principles and applications [7] Computer and robot vision [8] Torus [9] ANFIS: Adaptive-network-based fuzzy inference system [10] An introduction to ROC analysis [11] Simultaneous determination of size and refractive index of red blood cells by light scattering measurements [12]Estimation of human red blood cells size using light scattering images Aims Size, shape, and volume of Red Blood Cells are important factors in diagnosing bloodassociated disorders such as iron deficiency and anemia. Every day, thousands of blood samples are tested by microscopes and automated cell counter devices in pathology laboratories around the world, which may be expensive and time-consuming. The objective of this study was to measure mean corpuscular volume of abnormal red blood cells using the adaptive neuro-fuzzy system with image processing.Materials & Methods This study was conducted on 60 blood samples from the archive of pathology laboratory of Sarem hospital including 40 normal samples and 20 abnormal samples. Adaptive local thresholding and bounding box methods were used to extract the inner and outer diameters of red cells to calculate MCV. An adaptive-network-based fuzzy inference system was used to classify blood samples to normal and abnormal groups. In this method, normal and abnormal blood samples were examined using image processing techniques and MATLAB software. Findings The Accuracy of the proposed method and area under the curve were found as 96.6% and 0.995%, respectively. Conclusion The proposed method provides diagnostic capability using a drop of the blood sample and showed suitable performance on pathological images. The designed automatic system can be a convenient and cost effective alternative for common laboratory procedures. In addition, the method provides a basis for calculating other parameters of blood test or CBC such as mean cell hemoglobin, mean cell hemoglobin concentration, RDW, hematocrit, and red blood cell count.

Volume of red blood cell is an important factor in distinguishing its abnormalities. Mean corpuscular volume (MCV) of red blood cells contributes much to differentiation of several blood diseases like iron deficiency and other types of anemia. This paper proposes an automated system to classify blood samples using cell microscopic images instead of pathology test results. Adaptive local thresholding is first used to segment cell images. The volumes of red cells are then estimated by assuming torus geometry for cells. Finally, an adaptive network-based fuzzy inference system (ANFIS) is used to classify blood samples to normal and abnormal. Accuracy of the proposed system and area under Receiver Operating Characteristics (ROC) curve are 100% and 1 respectively.

No abstract