Images Of Red Blood Cells Research Articles

Smartphone-based sickle cell disease detection and monitoring for point-of-care settings

Sickle cell disease (SCD) is a worldwide hematological disorder causing painful episodes, anemia, organ damage, stroke, and even deaths. It is more common in sub-Saharan Africa and other resource-limited countries. Conventional laboratory-based diagnostic methods for SCD are time-consuming, complex, and cannot be performed at point-of-care (POC) and home settings. Optical microscope-based classification and counting demands a significant amount of time, extensive setup, and cost along with the skilled human labor to distinguish the normal red blood cells (RBCs) from sickled cells. There is an unmet need to develop a POC and home-based test to diagnose and monitor SCD and reduce mortality in resource-limited settings. An early-stage and timely diagnosis of SCD can help in the effective management of the disease. In this article, we utilized a smartphone-based image acquisition method for capturing RBC images from the SCD patients in normoxia and hypoxia conditions. A computer algorithm is developed to differentiate RBCs from the patient's blood before and after cell sickling. Using the developed smartphone-based technique, we obtained similar percentage of sickle cells in blood samples as analyzed by conventional method (standard microscope). The developed method of testing demonstrates the potential utility of the smartphone-based test for reducing the overall cost of screening and management for SCD, thus increasing the practicality of smartphone-based screening technique for SCD in low-resource settings. Our setup does not require any special storage requirements. This is the characteristic advantage of our technique as compared to other hemoglobin-based POC diagnostic techniques.

Quantification of malaria parasitaemia using trainable semantic segmentation and capsnet

Malaria is a life-threatening mosquito (Anopheles)-borne blood disease caused by the plasmodium parasite. Microscopic examination of peripheral blood smears by experts helps to identify parasites precisely. The manual assessment technique is a tedious and time-consuming process. The present study focuses on developing a hybrid screening algorithm for automated identification and classification of malaria parasite-infected red blood cells (RBCs). Initially, a semantic blood cell segmentation method is adopted where a supervised classifier-regulated pixel-based segmentation is adopted to segment individual RBC present in an image. In pixel-based classification, foreground (RBCs) and background regions are considered, a pixel-based large feature dataset is generated, and an artificial neural network (ANN) classifier is trained. The trained model generates a probability map of an image which is later post-processed by Graph-cut and Marker-controlled Watershed method for developing cropped RBC image set. The proposed segmentation method achieves 99.1% accuracy. Finally, a trained modified Capsule Network (CapsNet) model is used for classification of segmented blood cells to identify the species and stages of the parasites. Here, two specific parasite species viz., Plasmodium vivax and Plasmodium falciparum with stages are considered for classification. The performance of the proposed two-steps hybrid malaria screening is promising and the training and testing on local and benchmark dataset with respect to ground truth yield 98.7% accuracy.

Automated Semantic Segmentation of Red Blood Cells for Sickle Cell Disease.

Red blood cell (RBC) segmentation and classification from microscopic images is a crucial step for the diagnosis of sickle cell disease (SCD). In this work, we adopt a deep learning based semantic segmentation framework to solve the RBC classification task. A major challenge for robust segmentation and classification is the large variations on the size, shape and viewpoint of the cells, combining with the low image quality caused by noise and artifacts. To address these challenges, we apply deformable convolution layers to the classic U-Net structure and implement the deformable U-Net (dU-Net). U-Net architecture has been shown to offer accurate localization for image semantic segmentation. Moreover, deformable convolution enables free-form deformation of the feature learning process, thus making the network more robust to various cell morphologies and image settings. dU-Net is tested on microscopic red blood cell images from patients with sickle cell disease. Results show that dU-Net can achieve highest accuracy for both binary segmentation and multi-class semantic segmentation tasks, comparing with both unsupervised and state-of-the-art deep learning based supervised segmentation methods. Through detailed investigation of the segmentation results, we further conclude that the performance improvement is mainly caused by the deformable convolution layer, which has better ability to separate the touching cells, discriminate the background noise and predict correct cell shapes without any shape priors.

A Deep Learning Approach for Segmentation of Red Blood Cell Images and Malaria Detection.

Malaria is an endemic life-threating disease caused by the unicellular protozoan parasites of the genus Plasmodium. Confirming the presence of parasites early in all malaria cases ensures species-specific antimalarial treatment, reducing the mortality rate, and points to other illnesses in negative cases. However, the gold standard remains the light microscopy of May-Grünwald–Giemsa (MGG)-stained thin and thick peripheral blood (PB) films. This is a time-consuming procedure, dependent on a pathologist’s skills, meaning that healthcare providers may encounter difficulty in diagnosing malaria in places where it is not endemic. This work presents a novel three-stage pipeline to (1) segment erythrocytes, (2) crop and mask them, and (3) classify them into malaria infected or not. The first and third steps involved the design, training, validation and testing of a Segmentation Neural Network and a Convolutional Neural Network from scratch using a Graphic Processing Unit. Segmentation achieved a global accuracy of 93.72% over the test set and the specificity for malaria detection in red blood cells (RBCs) was 87.04%. This work shows the potential that deep learning has in the digital pathology field and opens the way for future improvements, as well as for broadening the use of the created networks.

Development of MatLab Coding for Early Detection of Leukemia through Automated Analysis of RBCs

Background: Bone marrow biopsy has become an integral part of leukemia diagnosis and its treatment. Several advancements are being made towards the analysis of digital images of biopsy samples. Recently, the FDA approved the procedures of digital health. In tune with that, digital image analysis has become propelled. With the advent of high-throughput technologies, the scientific community focuses on the red blood cells (RBCs) for early detection of cancer, including leukemia. The reasons are due to their abundance in peripheral blood and hence, easily accessible compared to the bone marrow biopsy procedure. High magnification and high-resolution electron microscopy-based ultra-structural analysis of RBCs already proved the utility of the hypothesis about a decade ago. However, in clinical set-up, electron microscopy-based procedures are the major bottleneck in the implementation of early detection of leukemia. Algorithm-based computer vision may be suitable to overcome this limitation. Methods: An intensive search with PubMed and Google for early diagnosis of leukemia through RBC light microscopic images was made. For this search, the image processing algorithm for RBC was also made in PubMed, IEEE Xplorer and Google; and the latest developments are noted. To fill the existing gap, a user-friendly MatLab coding is developed for automated analysis of RBC images. Results: RBC images from both normal and leukemia were analyzed with the developed code. Each RBC cells were analyzed individually for each sample of normal and leukemia. Therefore, in the output cellular characteristics namely, radius, perimeter, area, convexity and solidity were represented in a quantitative manner. Comparison of mean values between normal and leukemia groups for corresponding variables showed statistical significance. In the test run, data of 82 RBC cells from single leukemia sample when compared with the data of pooled data of normal samples also showed significant difference (P<0.05). Conclusion: Thus, the developed code successfully distinguishes between RBC cells of leukemia and normal. We hope that this RBC based developed code would be useful in identifying earlystage of leukemia in an individual patient; however, for this, ~100 RBC cells are needed to be analyzed. As the developed code is based on RBC based diagnosis therefore, may be applied to other cancers and if needed, further modification is possible due to availability of code. Thus, it is expected that the developed MatLab code may play a role in preventive oncology.

A deep learning AlexNet model for classification of red blood cells in sickle cell anemia

Sickle cell anemia (SCA) is a serious hematological disorder, where affected patients are frequently hospitalized throughout a lifetime and even can cause death. The manual method of detecting and classifying abnormal cells of SCA patient blood film through a microscope is time-consuming, tedious, prone to error, and require a trained hematologist. The affected patient has many cell shapes that show important biomechanical characteristics. Hence, having an effective way of classifying the abnormalities present in the SCA disease will give a better insight into managing the concerned patient's life. This work proposed algorithm in two-phase firstly, automation of red blood cells (RBCs) extraction to identify the RBC region of interest (ROI) from the patient’s blood smear image. Secondly, deep learning AlexNet model is employed to classify and predict the abnormalities presence in SCA patients. The study was performed with (over 9,000 single RBC images) taken from 130 SCA patient each class having 750 cells. To develop a shape factor quantification and general multiscale shape analysis. We reveal that the proposed framework can classify 15 types of RBC shapes including normal in an automated manner with a deep AlexNet transfer learning model. The cell's name classification prediction accuracy, sensitivity, specificity, and precision of 95.92%, 77%, 98.82%, and 90% were achieved, respectively.

Malaria Detection Using Image Processing

Malaria is one of a serious infectious disease in the world caused by a peripheral blood parasite of the genus Plasmodium. Traditional microscopy method, for malaria diagnosis is old method and has been occasionally proved inefficient since it is time consuming and results are difficult to reproduce. As it constitutes a serious global health problem, the evaluation process is of high importance. In this work, an accurate, easy, rapid and affordable model of malaria diagnosis using stained thin blood smear images was developed. The method makes use of the intensity features of Plasmodium parasites and also erythrocyte or red blood cell images. Images of infected and non-infected red blood cells were taken, pre-processed, and then relevant features were extracted from them and eventually diagnosis was made based on the features extracted from the images. A set of features based on intensity have been proposed, and the performance of these features on the red blood cell samples from the database have been evaluated using an artificial neural network (ANN) classifier. The results have shown that these features could be extremely successful if used for malaria parasite detection.

Imaging human blood cells in vivo with oblique back-illumination capillaroscopy.

We present a non-invasive, label-free method of imaging blood cells flowing through human capillaries in vivo using oblique back-illumination capillaroscopy (OBC). Green light illumination allows simultaneous phase and absorption contrast, enhancing the ability to distinguish red and white blood cells. Single-sided illumination through the objective lens enables 200 Hz imaging with close illumination-detection separation and a simplified setup. Phase contrast is optimized when the illumination axis is offset from the detection axis by approximately 225 µm when imaging ∼80 µm deep in phantoms and human ventral tongue. We demonstrate high-speed imaging of individual red blood cells, white blood cells with sub-cellular detail, and platelets flowing through capillaries and vessels in human tongue. A custom pneumatic cap placed over the objective lens stabilizes the field of view, enabling longitudinal imaging of a single capillary for up to seven minutes. We present high-quality images of blood cells in individuals with Fitzpatrick skin phototypes II, IV, and VI, showing that the technique is robust to high peripheral melanin concentration. The signal quality, speed, simplicity, and robustness of this approach underscores its potential for non-invasive blood cell counting.

Implementation of Convolutional Neural Network in the classification of red blood cells have affected of malaria

Malaria is a disease caused by plasmodium which attacks red blood cells. Diagnosis of malaria can be made by examining the patient's red blood cells using a microscope. Convolutional Neural Network (CNN) is a deep learning method that is growing rapidly. CNN is often used in image classification. The CNN process usually requires considerable resources. This is one of the weaknesses of CNN. In this study, the CNN architecture used in the classification of red blood cell images is LeNet-5 and DRNet. The data used is a segmented image of red blood cells and is secondary data. Before conducting the data training, data pre-processing and data augmentation from the dataset was carried out. The number of layers of the LeNet-5 and DRNet models were 4 and 7. The test accuracy of the LeNet-5 and DrNet models was 95% and 97.3%, respectively. From the test results, it was found that the LeNet-5 model was more suitable in terms of red blood cell classification. By using the LeNet-5 architecture, the resources used to perform classification can be reduced compared to previous studies where the accuracy obtained is also the same because the number of layers is less, which is only 4 layers

Sub-nanometer height sensitivity by phase shifting interference microscopy under environmental fluctuations.

Phase shifting interferometric (PSI) techniques are among the most sensitive phase measurement methods. Owing to its high sensitivity, any minute phase change caused due to environmental instability results into, inaccurate phase measurement. Consequently, a well calibrated piezo electric transducer (PZT) and highly-stable environment is mandatory for measuring accurate phase map using PSI implementation. Here, we present an inverse approach, which can retrieve phase maps of the samples with negligible errors under environmental fluctuations. The method is implemented by recording a video of continuous temporally phase shifted interferograms and phase shifts were calculated between all the data frames using Fourier transform algorithm with a high accuracy ≤ 5.5 × 10-4 π rad. To demonstrate the robustness of the proposed method, a manual translation of the stage was employed to introduce continuous temporal phase shift between data frames. The developed algorithm is first verified by performing quantitative phase imaging of optical waveguide and red blood cells using uncalibrated PZT under the influence of vibrations/air turbulence and compared with the well calibrated PZT results. Furthermore, we demonstrated the potential of the proposed approach by acquiring the quantitative phase imaging of an optical waveguide with a rib height of only 2 nm and liver sinusoidal endothelial cells (LSECs). By using 12-bit CMOS camera the height of shallow rib waveguide is measured with a height sensitivity of 4 Å without using PZT and in presence of environmental fluctuations.vn.

Semi-supervised learning with deep convolutional generative adversarial networks for canine red blood cells morphology classification

Information of Red Blood Cell (RBC) morphology, obtained by analysing RBC images, is regularly requested by veterinarians to diagnose anaemic dogs. Machine learning techniques have been exploited to speed up the image classification. Recently, many researchers used deep learning techniques for classification; however, a large quantity of labelled data is necessary to extract performance with them. A lack of annotated data, due to time and costs for pathologist and their limited numbers, has become a difficulty. This limits the amount of annotated data and leads to a large number of unannotated data, preventing traditional deep learning algorithms from being effective. We show that a semi-supervised learning method, using the Generative Adversarial Networks (GANs) for canine RBC morphology classification, can solve the lack of labelled data, when we want to train a deep learning classifier. Our semi-supervised GAN can use both labelled and unlabelled data and showed that they can achieve the same level of performance as a traditional convolutional neural network, with a smaller number of labelled images. Furthermore, we showed that augmenting the limited numbers of a labelled images enhanced the overall performance. A key benefit of our method is reduced pathologist cost and time to annotate cell images for developing a deep learning classifier.

Quantitative absorption imaging of red blood cells to determine physical and mechanical properties.

Red blood cells or erythrocytes, constituting 40 to 45 percent of the total volume of human blood are vesicles filled with hemoglobin with a fluid-like lipid bilayer membrane connected to a 2D spectrin network. The shape, volume, hemoglobin mass, and membrane stiffness of RBCs are important characteristics that influence their ability to circulate through the body and transport oxygen to tissues. In this study, we show that a simple two-LED set up in conjunction with standard microscope imaging can accurately determine the physical and mechanical properties of single RBCs. The Beer–Lambert law and undulatory motion dynamics of the membrane have been used to measure the total volume, hemoglobin mass, membrane tension coefficient, and bending modulus of RBCs. We also show that this method is sensitive enough to distinguish between the mechanical properties of RBCs during morphological changes from a typical discocyte to echinocytes and spherocytes. Measured values of the tension coefficient and bending modulus are 1.27 × 10−6 J m−2 and 7.09 × 10−20 J for discocytes, 4.80 × 10−6 J m−2 and 7.70 × 10−20 J for echinocytes, and 9.85 × 10−6 J m−2 and 9.69 × 10−20 J for spherocytes, respectively. This quantitative light absorption imaging reduces the complexity related to the quantitative imaging of the biophysical and mechanical properties of a single RBC that may lead to enhanced yet simplified point of care devices for analyzing blood cells.

Red blood cell classification on thin blood smear images for malaria diagnosis

Parasite detection is important for the diagnosis of many blood-borne diseases including malaria. As part of a program to develop a fast, accurate, and affordable automatic device for diagnosing malaria, a critical step is to automatically classify individual red blood cells in thin blood smear images. To automatically recognize malaria parasites in an image, this paper presents a red blood cell classification study for malaria diagnosis. To diagnose malaria, the threshold-based segmentation is implemented using the Otsu’s method succeeded by the distance transform and statistical classifier. The methods are applied to red blood cell images obtained from Kaggle. These experimental results show that the classification recognizes malaria parasite with 94.60% accuracy, 96.20% specificity, and 93% sensitivity.

Defects of Red Blood Cell Membranes in Patients with Brain Dysfunction (Pilot Study)

The aimof the paper: to identify promising diagnostic and prognostic biomarkers of pathological processes development based on the red blood cell membrane morphology and nanostructure in patients with brain disorders in the Intensive Care Unit.Materials and methods. The study included 24 patients from the anesthesiology and resuscitation ward of the Federal Research and Clinical Center of Intensive Care Medicine and Rehabilitology. Blood was acquired from the patients for standard tests, and all further tests were performed in vitro. The images of red blood cells were obtained using the atomic force microscope «NTEGRA Рrima» (NT-MDT, Russia) in semi-contact mode.Results. Patients from the anesthesiology and intensive care ward with traumatic brain injury, ischemic and hemorrhagic stroke, cerebral edema, and post-hypoxic encephalopathy had different blood cell shapes and localized defects of different topology on the surface of erythrocyte membranes including defects of pallor, torus, and nanostructure.Conclusion. In this pilot study we have shown that several defects represent the trigger mechanisms for the development of a total membrane damage. Local topographic defects of nanostructures and abnormalities of erythrocyte morphology are irreversible. The number and quality of these abnormalities may eventually be used as a diagnostic and prognostic biomarker of pathological processes.

Machine learning-based screening of red blood cells using quantitative phase imaging with micro-spectrocolorimetry

We report simultaneous micro-spectrocolorimetry and quantitative phase imaging (QPI) of human red blood cells (RBCs) using white light interference microscopy. To understand the mechanism of disease at the cellular level and to increase the diagnostic performance of QPI, we integrated it with micro-spectrocolorimetry. Color-coordinates of various spatial locations of RBCs along with 3D phase-maps help to quantify different biophysical parameters. Features extracted from this multimodal technique combined with support vector machine achieved average specificity, sensitivity, and accuracy of 95.52%, 95.58%, and 95.55%, respectively with testing data in classification of healthy and unhealthy RBCs. A better result is obtained by using synergies amongst data as compared to QPI based features only.

Endovascular Treatment of Klippel-Trenaunay Syndrome

Endovascular Treatment of Klippel-Trenaunay Syndrome

Assessing Malaria using Neutral Zone Classifiers with Mixture Discriminant Analysis on 2D Images of Red Blood Cells

Background and Aim: We aim to build a classifier to distinguish between malaria-infected red blood cells (RBCs) and healthy cells using the two-dimensional (2D) microscopic images of RBCs. We demonstrate the process of cell segmentation and feature extraction from the 2D images. Methods and Materials: We describe an approach to address the problem using mixture discriminant analysis (MDA) on the 2D image profiles of the RBCs. The extracted features are used with Gaussian MDA to distinguish between healthy and malaria infected cells. We also use the neutral zone classifiers where ambiguous cases are identified separately by the classifier. Results: We compare the classification results from the regular classifiers such as linear discriminant analysis (LDA) or MDA and the methods where neutral zone classifiers are used. We see that including the neutral zone improves the classification results by controlling the false positive and false negatives. The number of misclassifications are seen to be lower than the case without neutral zone classifiers. Conclusion: This paper presents an alternative approach for classification by incorporating neutral zone classifier approach, where a prediction is not made for the ambiguous cases. From the data analysis we see that this approach based on neutral zone classifiers presents a useful alternative in classification problems for various applications.

Impact of Misclassification Rates on Compression Efficiency of Red Blood Cell Images of Malaria Infection Using Deep Learning

Malaria is a severe public health problem worldwide, with some developing countries being most affected. Reliable remote diagnosis of malaria infection will benefit from efficient compression of high-resolution microscopic images. This paper addresses a lossless compression of malaria-infected red blood cell images using deep learning. Specifically, we investigate a practical approach where images are first classified before being compressed using stacked autoencoders. We provide probabilistic analysis on the impact of misclassification rates on compression performance in terms of the information-theoretic measure of entropy. We then use malaria infection image datasets to evaluate the relations between misclassification rates and actually obtainable compressed bit rates using Golomb–Rice codes. Simulation results show that the joint pattern classification/compression method provides more efficient compression than several mainstream lossless compression techniques, such as JPEG2000, JPEG-LS, CALIC, and WebP, by exploiting common features extracted by deep learning on large datasets. This study provides new insight into the interplay between classification accuracy and compression bitrates. The proposed compression method can find useful telemedicine applications where efficient storage and rapid transfer of large image datasets is desirable.

Aplikasi self-organizing mapping sebagai alat deteksi anemia pada citra sel darah merah

Application of self-organizing mapping as anemia detection using an image of red blood cellsBackground: Anemia is a nutritional problem characterized by changes in blood cell size, especially in microcytic or macrocytic anemia. Iron deficiency anemia is included in hypochromic microcytic anemia because it has a smaller than normal size red blood cell and has a lower than normal hemoglobin (Hb) arising from reduced supply of iron for erythropoiesis (cell maturation process red blood). Analysis based on red blood cell image is a tool to detect anemia using technology applications. Self-organizing mapping (SOM) is one of the artificial neural networks by dividing the input pattern into several groups, so the network output is in the form of groups that are most similar to the input.Objective: To measure the accuracy of SOM for detecting the size of red blood cells in anemia condition.Methods: The type of research was an observational laboratory. The study was conducted at the Electrobiomedical Laboratory of Universitas Respati Yogyakarta from January to August 2018. The sample consisted of anemia and non-anemia red blood cells which had been tested in a laboratory of 92 blood preparations. Stage of measuring red blood cells consisted of pre-processing (cropping, gray scaling, contrast enhancement, and screening), segmentation, feature extraction, and image identification with SOM. The image identification results were concluded by calculating the accuracy of the anemia detection system based on laboratory examination results.Results: The characteristic that distinguishes anemia and non-anemia was in the size of red blood cells. Anemic red blood cells had different pixel intensities than non-anemic red blood cells. The image of non-anemia red blood cells had a full round or oval image. From as many as 92 detections of blood images, five blood images were not by the target results of laboratory tests. The accuracy achieved by the system was 94.57%.Conclusions: The accuracy value of anemia detection using SOM can be used to identify the type of anemia based on red blood cell size.

Improved autofocus method for human red blood cell images.

This paper presents an improved autofocus method for human red blood cell images in a microscope. The products of the sum modulus difference and the real-valued fast Fourier transform function are multiplied to obtain an improved sharpness evaluation using the properties of a Gaussian function. It is superior to traditional evaluations in terms of unimodality, steepness, and sensitivity. A new quantitative criterion is proposed to represent the ability of sharpness evaluation against noise. An adaptive focus window with great robustness is proposed that can reduce the computation cost and adverse effects of the background. The better performances of the proposed algorithms are all proved by experiment results, and they can help to find the quasi-focus position more quickly and accurately.