Abstract
Visual examination of blood smears is an essential tool for analysis, prevention, and remediation of several types of maladies. The interest of computer-aided decision has been acknowledged in many medicinal instances (e.g., automatic ways and means are being explored to spot, classify, and measure visual items in hematological cytology [HC]). This chapter proposes an entirely automated blood smear diagnosis system for hemograms, which can lessen the time spent to scrutinize a slide. The present framework relies on morphological operations (MOs) and soft segmentation by means of the watershed transform (WT). Experiments demonstrate the method efficacy to count white blood cells (WBCs) and red blood cells (RBCs). Some considerations about implementations, design advice and possible variants, as well as improvements are discussed. The future of automated medical analysis is contemplated.
TopIntroduction
Hematologic examinations are relevant for diagnosing disorders of the hematologic system, and they help in the identification of many systemic and organ illnesses. Even though in most cases the disease diagnosis cannot be established solely on a Complete Blood Cell (CBC) count, the hemogram may add valuable evidence in the diagnosis, surveying, and elaboration of a prognosis concerning the future disease progression in an individual. This chapter provides physicians, veterinaries and laboratory technicians with the essentials on the design of an automated digital system for the analysis of Blood Cells (BCs), and diseases related to anomalies in the hematological profile. Since automated cell counters have become increasingly popular, the focus of the present system is on anomalies noticeable by automated cell counters such as abnormalities in the amount, volume, and hemoglobin concentration of BCs rather than on abnormal morphology. Nevertheless, if pathological or irregular outcomes are discovered, or abnormal morphology or BC function damage is suspected, a further microscopic blood smear analysis is strongly advised (Gulati et al., 2002) .
Blood analyses entail a blood smear obtained by sliding a few microliters of blood on a glass slide and subsequent use of dyes, which allow the visualization of cellular structures. Only the final portion of the slide is used to perform the counts, because the anterior portions have clustered and/or overlapping cells, preventing a reliable count. This blood smear is used to analyze the morphology and staining of Red Blood Cells (RBCs) as well as White Blood Cells (WBCs) and it can help to investigate the presence of anisocytosis, such as sickle cell anemia (sickle-shaped RBCs), anemia spherocytosis (round RBCs), hereditary elliptosis (elongated RBCs), among others besides aiding in the identification of blood parasites, such as leishmaniosis, malaria, and so on (Melo, & Silveira, 2013; Ciesla, 2012; Xavier et al., 2016).
Given the above, making the BC methodology more accessible to the less favored populations may help in the detection and consequent early treatment of the pathologies.
It is important to note that there are many tools to solve health problems using a high-level programming language and interactive environment for numerical computation, visualization and programming. This work relies on the MATLAB IDE that is a proprietary software but programs developed in it can be executed in Free Open Source packages like Octave and Scilab (Hansen, 2011; Roux, 2016; Wouwer et al., 2014). It can analyze data, develop algorithms, create models and applications, build mathematical tools capable of exploring multiple approaches and search for a faster solution than traditional programming languages such as C/C++ or Java (Reyes-Aldosoro, 2015).
First, image preprocessing of the digitized smears is applied for suppressing noise, ameliorating the luminance/contrast differences in the blood smear color image and, in some cases, removing unnecessary image items such as platelets. Second, a segmentation process discovers and isolates the important entities in the image. The third stage characterizes the objects previously extracted to be used in the last phase, i.e., the classification stage. Feature selection can lessen the redundant information required to describe important parts of the image. Selected features are fed to the classification process, which makes the class assignment (Diaz, & Manzanera, 2009).
It is important to stress the fact that the framework discussed here can be easily adapted to veterinary use provided the parameters and databases are adjusted to the animal undergoing examinations. The system in (Herman et al., 2018) is a dedicated analyzer for cows. Figure 1 shows a blood smear for a human and a bovine.
Figure 1. Blood smears for a human (a) and a cow (b)
Key Terms in this Chapter
Average White Blood Count (WBC): This important parameter counts the number of leukocytes present in the bloodstream ideally between 3000 and 10000 cells per mm 3 of blood. The WBC values depend on the age and sex of the patient. More than 10,000 leukocytes per mm3 strongly indicates infections and inflammation, whether acute or chronic.
Average Red Blood Count (RBC): This important parameter counts the number of erythrocytes present in the bloodstream ideally between 4.0 to 6.0 x 10 6 mm 3 of blood. These values are variable according to the age and sex of the patient. Changes in the mean value of red blood cells may indicate genetic anemias, deficiency anemia, leukemias, polycythemias, or blood loss.
Flow Cytometry: Is a technique performed by means of an apparatus called a cytometry whose objective is to analyze particles and/or cells suspended in a liquid medium and submitted to a flow, which simulates the passage of blood from a blood vessel. In this way, it is possible to perform the quantification, analysis, and classification of cells and/or particles.
Content-Based Image Retrieval (CBIR): Is a technique that consists of the application of computational vision and pattern recognition methodologies to solve image retrieval problems in large databases. Visual features such as coloring, texture, and so on are used for the creation of a feature descriptor.
Blood Smear: It consists of the deposition of a few microliters of blood in a glass slide. It is from the blood smear that the qualitative and quantitative evaluations of blood cells are performed.
Mean Corpuscular Hemoglobin (MCH): Is a parameter of evaluation of the erythrocyte indices, being responsible for the evaluation of the staining of the erythrocytes. Thus, this analysis is of great importance for the classification of anemia. It is the average amount of hemoglobin present within each erythrocyte, being represented in picograms (pg).
Watershed Transform: Is an image processing technique that had its principle based on observations of nature. It presents great applicability in the detection and labeling of objects present in an image.
Mean Corpuscular Hemoglobin Concentration (MCHC): Is a parameter of evaluation of the erythrocyte indices, being responsible for the evaluation of the staining of the erythrocytes. Thus, this analysis is of great importance for the classification of anemia. Is the mean hemoglobin concentration in 100 mL erythrocytes being plotted in gram per deciliter (g/dL). CHCM differs from HCM by the fact that the concentration takes cell volume into account and HCM only relates total hemoglobin to the number of erythrocytes.
Complete Blood Count: Is the result of the counting of blood cells by means of counting of Neubauer chamber with the posterior submission of the values found in consecrated mathematical formulas in the medical literature.