Comparative Study of Data Augmentation Strategies for White Blood Cells Classification

2019 ◽  
Author(s):  
George Kolokolnikov ◽  
Andrey Samorodov
Keyword(s):  
Comparative Study ◽  
Blood Cells ◽  
Data Augmentation ◽  
White Blood Cells ◽  
Augmentation Strategies

Society of Physicians at Kazan University

1927 ◽  
Vol 23 (1) ◽  
pp. 144-145
Keyword(s):  
Comparative Study ◽  
Erythrocyte Sedimentation Rate ◽  
Sedimentation Rate ◽  
Clinical Picture ◽  
Blood Cells ◽  
White Blood Cells ◽  
Not Given ◽  
Blood Picture ◽  
Erythrocyte Sedimentation

Session 9/Xll 1926. Prof. N. K. Goryaev: To estimate the content of white blood cells by smear (Schtzung). The report will be printed in "Kaz. Med. Jour.". Drs. N. Zakharov, N. Kudryashev and M. Aksyantsev: Experience in comparative study of immunobiological reactions in tbc clinic. Report, to be printed in Kaz. Med. Jour. Prof. P. N. Nikolaev pointed out about the report that it is impossible to perform all reactions studied by the reporters at the patient's bedside and that the doctor should never forget the personality of the patient in his work. Prof. N. K. Goryaev noted that a big drawback of the report was the lack of description of the clinical picture of the cases studied, as well as the fact that in parallel with the reactions studied the blood picture was not given. To study erythrocyte sedimentation rate Prof. Goryaev considers the Linzeamier's methodology inconvenient. In addition, Drs. Mastbaum and Aksyantsev commented on the reports.

scholarly journals Improved Classification of White Blood Cells with the Generative Adversarial Network and Deep Convolutional Neural Network

2020 ◽  
Vol 2020 ◽  
pp. 1-12
Author(s):  
Khaled Almezhghwi ◽  
Sertan Serte
Keyword(s):  
Blood Cells ◽  
Data Augmentation ◽  
White Blood Cells ◽  
Training Data ◽  
Generative Adversarial Networks ◽  
Generative Adversarial Network ◽  
Human Errors ◽  
Morphological Variations ◽  
Adversarial Network

White blood cells (leukocytes) are a very important component of the blood that forms the immune system, which is responsible for fighting foreign elements. The five types of white blood cells include neutrophils, eosinophils, lymphocytes, monocytes, and basophils, where each type constitutes a different proportion and performs specific functions. Being able to classify and, therefore, count these different constituents is critical for assessing the health of patients and infection risks. Generally, laboratory experiments are used for determining the type of a white blood cell. The staining process and manual evaluation of acquired images under the microscope are tedious and subject to human errors. Moreover, a major challenge is the unavailability of training data that cover the morphological variations of white blood cells so that trained classifiers can generalize well. As such, this paper investigates image transformation operations and generative adversarial networks (GAN) for data augmentation and state-of-the-art deep neural networks (i.e., VGG-16, ResNet, and DenseNet) for the classification of white blood cells into the five types. Furthermore, we explore initializing the DNNs’ weights randomly or using weights pretrained on the CIFAR-100 dataset. In contrast to other works that require advanced image preprocessing and manual feature extraction before classification, our method works directly with the acquired images. The results of extensive experiments show that the proposed method can successfully classify white blood cells. The best DNN model, DenseNet-169, yields a validation accuracy of 98.8%. Particularly, we find that the proposed approach outperforms other methods that rely on sophisticated image processing and manual feature engineering.

scholarly journals Higher positive identification of malignant CSF cells using the cytocentrifuge than the Suta chamber

2016 ◽  
Vol 74 (10) ◽  
pp. 816-822 ◽  
Author(s):  
Sérgio Monteiro de Almeida ◽  
Indianara Rotta ◽  
Arnaldo José de Conto ◽  
Dario Antonelli Filho ◽  
Carlos Dabdoub Roda ◽  
...  
Keyword(s):  
Cerebrospinal Fluid ◽  
Comparative Study ◽  
Retrospective Analysis ◽  
Blood Cells ◽  
Cell Concentration ◽  
White Blood Cells ◽  
Malignant Cells ◽  
Normal Range ◽  
Positive Identification ◽  
Csf Cells

ABSTRACT Objective To define how to best handle cerebrospinal fluid (CSF) specimens to obtain the highest positivity rate for the diagnosis of malignancy, comparing two different methods of cell concentration, sedimentation and cytocentrifugation. Methods A retrospective analysis of 411 CSF reports. Results This is a descriptive comparative study. The positive identification of malignant CSF cells was higher using the centrifuge than that using the Suta chamber (27.8% vs. 19.0%, respectively; p = 0.038). Centrifuge positively identified higher numbers of malignant cells in samples with a normal concentration of white blood cells (WBCs) (< 5 cells/mm3) and with more than 200 cells/mm3, although this was not statistically significant. There was no lymphocyte loss using either method. Conclusions Cytocentrifugation positively identified a greater number of malignant cells in the CSF than cytosedimentation with the Suta chamber. However, there was no difference between the methods when the WBC counts were within the normal range.

scholarly journals Diagnosis of Leukaemia in Blood Slides Based on a Fine-Tuned and Highly Generalisable Deep Learning Model

Sensors ◽  
2021 ◽  
Vol 21 (9) ◽  
pp. 2989
Author(s):  
Luis Vogado ◽  
Rodrigo Veras ◽  
Kelson Aires ◽  
Flávio Araújo ◽  
Romuere Silva ◽  
...  
Keyword(s):  
Blood Cells ◽  
Data Augmentation ◽  
White Blood Cells ◽  
Model Performance ◽  
Fine Tuning ◽  
Training Dataset ◽  
Current State ◽  
Unseen Data ◽  
Texture Characteristics ◽  
Tuning Methods

Leukaemia is a dysfunction that affects the production of white blood cells in the bone marrow. Young cells are abnormally produced, replacing normal blood cells. Consequently, the person suffers problems in transporting oxygen and in fighting infections. This article proposes a convolutional neural network (CNN) named LeukNet that was inspired on convolutional blocks of VGG-16, but with smaller dense layers. To define the LeukNet parameters, we evaluated different CNNs models and fine-tuning methods using 18 image datasets, with different resolution, contrast, colour and texture characteristics. We applied data augmentation operations to expand the training dataset, and the 5-fold cross-validation led to an accuracy of 98.61%. To evaluate the CNNs generalisation ability, we applied a cross-dataset validation technique. The obtained accuracies using cross-dataset experiments on three datasets were 97.04, 82.46 and 70.24%, which overcome the accuracies obtained by current state-of-the-art methods. We conclude that using the most common and deepest CNNs may not be the best choice for applications where the images to be classified differ from those used in pre-training. Additionally, the adopted cross-dataset validation approach proved to be an excellent choice to evaluate the generalisation capability of a model, as it considers the model performance on unseen data, which is paramount for CAD systems.

scholarly journals Mobile-Aware Deep Learning Algorithms for Malaria Parasites and White Blood Cells Localization in Thick Blood Smears

Algorithms ◽  
2021 ◽  
Vol 14 (1) ◽  
pp. 17
Author(s):  
Rose Nakasi ◽  
Ernest Mwebaze ◽  
Aminah Zawedde
Keyword(s):  
Computer Vision ◽  
Deep Learning ◽  
Blood Cells ◽  
Data Augmentation ◽  
Expert Knowledge ◽  
White Blood Cells ◽  
Single Shot ◽  
Good Precision ◽  
Malaria Parasites ◽  
Malaria Parasitemia

Effective determination of malaria parasitemia is paramount in aiding clinicians to accurately estimate the severity of malaria and guide the response for quality treatment. Microscopy by thick smear blood films is the conventional method for malaria parasitemia determination. Despite its edge over other existing methods of malaria parasitemia determination, it has been critiqued for being laborious, time consuming and equally requires expert knowledge for an efficient manual quantification of the parasitemia. This pauses a big challenge to most low developing countries as they are not only highly endemic but equally low resourced in terms of technical personnel in medical laboratories This study presents an end-to-end deep learning approach to automate the localization and count of P.falciparum parasites and White Blood Cells (WBCs) for effective parasitemia determination. The method involved building computer vision models on a dataset of annotated thick blood smear images. These computer vision models were built based on pre-trained deep learning models including Faster Regional Convolutional Neural Network (Faster R-CNN) and Single Shot Multibox Detector (SSD) models that help process the obtained digital images. To improve model performance due to a limited dataset, data augmentation was applied. Results from the evaluation of our approach showed that it reliably detected and returned a count of parasites and WBCs with good precision and recall. A strong correlation was observed between our model-generated counts and the manual counts done by microscopy experts (posting a spear man correlation of ρ = 0.998 for parasites and ρ = 0.987 for WBCs). Additionally, our proposed SSD model was quantized and deployed on a mobile smartphone-based inference app to detect malaria parasites and WBCs in situ. Our proposed method can be applied to support malaria diagnostics in settings with few trained Microscopy Experts yet constrained with large volume of patients to diagnose.

Ultrastructural changes in murine lymphoma cells exposed to ultraviolet-A radiation

1991 ◽  
Vol 49 ◽  
pp. 104-105
Author(s):  
Delma P. Thomas ◽  
Dianne E. Godar
Keyword(s):  
Medical Devices ◽  
Blood Cells ◽  
White Blood Cells ◽  
Consumer Products ◽  
Ultrastructural Changes ◽  
Ultraviolet A ◽  
Uv Spectrum ◽  
Lymphoma Cells ◽  
Murine Lymphoma ◽  
Transmission Electron

Ultraviolet radiation (UVR) from all three waveband regions of the UV spectrum, UVA (320-400 nm), UVB (290-320 nm), and UVC (200-290 nm), can be emitted by some medical devices and consumer products. Sunlamps can expose the blood to a considerable amount of UVR, particularly UVA and/or UVB. The percent transmission of each waveband through the epidermis to the dermis, which contains blood, increases in the order of increasing wavelength: UVC (10%) < UVB (20%) < UVA (30%). To investigate the effects of UVR on white blood cells, we chose transmission electron microscopy to examine the ultrastructure changes in L5178Y-R murine lymphoma cells.

The Effect of Red Blood Cells on the ADP-induced Aggregation of Human Platelets in Flow Through Tubes

1990 ◽  
Vol 63 (01) ◽  
pp. 112-121 ◽  
Author(s):  
David N Bell ◽  
Samira Spain ◽  
Harry L Goldsmith
Keyword(s):  
Platelet Aggregation ◽  
Shear Rate ◽  
Red Blood Cells ◽  
Blood Cells ◽  
Platelet Rich Plasma ◽  
Mean Transit Time ◽  
White Blood Cells ◽  
Aggregate Size ◽  
Human Platelets ◽  
Induced Aggregation

SummaryThe effect of red blood cells, rbc, and shear rate on the ADPinduced aggregation of platelets in whole blood, WB, flowing through polyethylene tubing was studied using a previously described technique (1). Effluent WB was collected into 0.5% glutaraldehyde and the red blood cells removed by centrifugation through Percoll. At 23°C the rate of single platelet aggregtion was upt to 9× greater in WB than previously found in platelet-rich plasma (2) at mean tube shear rates Ḡ = 41.9,335, and 1,920 s−1, and at both 0.2 and 1.0 µM ADP. At 0.2 pM ADP, the rate of aggregation was greatest at Ḡ = 41.9 s−1 over the first 1.7 s mean transit time through the flow tube, t, but decreased steadily with time. At Ḡ ≥335 s−1 the rate of aggregation increased between t = 1.7 and 8.6 s; however, aggregate size decreased with increasing shear rate. At 1.0 µM ADP, the initial rate of single platelet aggregation was still highest at Ḡ = 41.9 s1 where large aggregates up to several millimeters in diameter containing rbc formed by t = 43 s. At this ADP concentration, aggregate size was still limited at Ḡ ≥335 s−1 but the rate of single platelet aggregation was markedly greater than at 0.2 pM ADP. By t = 43 s, no single platelets remained and rbc were not incorporated into aggregates. Although aggregate size increased slowly, large aggregates eventually formed. White blood cells were not significantly incorporated into aggregates at any shear rate or ADP concentration. Since the present technique did not induce platelet thromboxane A2 formation or cause cell lysis, these experiments provide evidence for a purely mechanical effect of rbc in augmenting platelet aggregation in WB.

White blood cells levels and PCOS: direct and indirect relationship with insulin resistance, but not with hyperandogenemia

2013 ◽  
Author(s):  
Olga Papalou ◽  
Sarantis Livadas ◽  
Athanasios Karachalios ◽  
Nektarios Benetatos ◽  
George Boutzios ◽  
...  
Keyword(s):  
Insulin Resistance ◽  
Blood Cells ◽  
White Blood Cells ◽  
Indirect Relationship

Endogenous Heparinoids Detected by anti-Xa Activity are Present in Blood during Acute Variceal Bleeding in Cirrhosis. A Prospective Study

2014 ◽  
Vol 23 (2) ◽  
pp. 187-194 ◽  
Author(s):  
Christos Triantos ◽  
Emmanuel Louvros ◽  
Maria Kalafateli ◽  
Anne Riddell ◽  
Ulrich Thalheimer ◽  
...  
Keyword(s):  
Hepatocellular Carcinoma ◽  
Variceal Bleeding ◽  
Blood Cells ◽  
Venous Pressure ◽  
White Blood Cells ◽  
International Normalized Ratio ◽  
Ulcer Bleeding ◽  
Factor X ◽  
Packed Red Blood Cells ◽  
Bacterial Peritonitis

Background & Aims: Endogenous heparinoids have been detected by thromboelastography and quantified by clotting based anti-Xa activity assays in patients with cirrhosis, but their presence in variceal bleeding has not been established yet.Methods: Clotting based anti-Xa activity was measured in A) 30 cirrhotics with variceal bleeding, B) 15 noncirrhotics with peptic ulcer bleeding, C) 10 cirrhotics without infection or bleeding, and D) 10 cirrhotics with hepatocellular carcinoma (HCC).Results: Anti-Xa activity was not detected in ulcer bleeders or in cirrhotics without infection or bleedingbut was present in seven (23%) variceal bleeders (median levels: 0.03 u/mL (0.01-0.07)) and was quantifiable for 3 days in six of seven patients. Four of seven variceal bleeders with anti-Xa activity present had HCC (p=0.023). Age, creatinine, platelet count and total infections the second day from admission were significantly correlated with the presence of measureable anti-Xa levels (p=0.014, 0.032, 0.004 and 0.019, respectively). In the HCC group, anti-Xa activity was present in three patients (30%) [median levels: 0.05 u/mL (0.01-0.06)].Conclusions: In this study, variceal bleeders and 30% of the patients with HCC had endogenous heparinoids that were detected by a clotting based anti-Xa activity assay, whereas there was no anti Xa activity present in patients with cirrhosis without infection, or bleeding or HCC, nor in those with ulcer bleeding. Thus, the anti-Xa activity is likely to be a response to bacterial infection and/or presence of HCC in cirrhosis.List of abbreviations: AFP, alpha-fetoprotein; aPTT, activated partial thromboplastin time; CP, Child-Pugh; FXa, activated factor X; GAGS, glycosaminoglycans; Hb, hemoglobin; HCC, hepatocellular carcinoma; HVPG, hepatic venous pressure gradient; INR, International normalized ratio; LMWHs, low molecular weight heparins; MELD, Model for End-stage Liver Disease; PPP, platelet-poor plasma; PRBC, packed red blood cells; PT, prothrombin time; SBP, sponataneous bacterial peritonitis; TEG, thromboelastography; WBC, white blood cells.

Sign in / Sign up

Export Citation Format

Share Document