Label-Free White Blood Cell Classification Using Refractive Index Tomography and Deep Learning
40
Citation
50
Reference
10
Related Paper
Citation Trend
Abstract:
Objective and Impact Statement. We propose a rapid and accurate blood cell identification method exploiting deep learning and label-free refractive index (RI) tomography. Our computational approach that fully utilizes tomographic information of bone marrow (BM) white blood cell (WBC) enables us to not only classify the blood cells with deep learning but also quantitatively study their morphological and biochemical properties for hematology research. Introduction. Conventional methods for examining blood cells, such as blood smear analysis by medical professionals and fluorescence-activated cell sorting, require significant time, costs, and domain knowledge that could affect test results. While label-free imaging techniques that use a specimen's intrinsic contrast (e.g., multiphoton and Raman microscopy) have been used to characterize blood cells, their imaging procedures and instrumentations are relatively time-consuming and complex. Methods. The RI tomograms of the BM WBCs are acquired via Mach-Zehnder interferometer-based tomographic microscope and classified by a 3D convolutional neural network. We test our deep learning classifier for the four types of bone marrow WBC collected from healthy donors (n=10): monocyte, myelocyte, B lymphocyte, and T lymphocyte. The quantitative parameters of WBC are directly obtained from the tomograms. Results. Our results show >99% accuracy for the binary classification of myeloids and lymphoids and >96% accuracy for the four-type classification of B and T lymphocytes, monocyte, and myelocytes. The feature learning capability of our approach is visualized via an unsupervised dimension reduction technique. Conclusion. We envision that the proposed cell classification framework can be easily integrated into existing blood cell investigation workflows, providing cost-effective and rapid diagnosis for hematologic malignancy.Keywords:
White blood cell
Monocyte
This study was designed to compare the complete blood count values of opioid users (N = 61) and healthy subjects (N = 61), particularly monocyte-to-lymphocyte ratio (MLR) and platelet-to-lymphocyte ratio (PLR). PLR, MLR, and percentage of monocyte (MONO%) were significantly lower in opioid use disorder (OUD) group (P = 0.012, P = 0.005, P = 0.000). The area under the ROC curve of MLR and PLR levels for OUD was 0.349 and 0.368. MONO% correlated with substance use duration. Measurements like lymphocyte-related ratios and MONO% in opioid use can be important in substance monitoring, detection, and differentiation of acute and chronic conditions.
Monocyte
Cite
Citations (45)
Cite
Citations (5)
Monocyte
Cite
Citations (0)
Objective: This study aims to investigate the level of peripheral blood mononuclear cells and their ratios which may point to the immunological mechanisms involved in the etiopathogenesis of ASD.Method: The complete blood count parameters of the 45 ASD cases were compared with those of healthy controls.Childhood Autism Rating Scale (CARS) was performed to measure the disease severity.Results: The monocytes of ASD group were significantly higher; and the lymphocyte-to-monocyte ratio (LMR) was lower than the controls'. LMR and neutrophil-to-lymphocyte ratio were found to be predictors of ASD. The decrease in LMR (B: −0.744; P=0.035; CI: −1.431 to −0.056) and the increase in age (B: 0.432; P=0.045; CI: 0.011–0.853) were related to high CARS scores in linear regression analyses.Conclusions: The results of this study support the role of altered immune cell counts and ratios in ASD. A high monocyte level and low LMR may have diagnostic values in autism.
Monocyte
Cite
Citations (33)
Objective: In this study, in patients with moderate coronary lesions evaluated in coronary angiography, fractional flow reserve by lesion severity, we aimed to determine the relationship between neutrophil/lymphocyte ratio, platelet/lymphocyte ratio, lymphocyte/monocyte ratio, and monocyte/high-density lipoprotein cholesterol ratio, which has been recently expressed as a predictor of cardiovascular disease risk. Methods: Stenosis with a fractional flow reserve of <0.80 was considered functionally severe. According to fractional flow reserve lesion severity, a total of 131 patients were analyzed, with fractional flow reserve > 0.8 (group 1) and fractional flow reserve < 0.8 (group 2). Patients with acute coronary syndrome, severe arrhythmia, hemodynamic instability, history of previous revascularization, severe renal and hepatic failure, active infection, malignancy, hematologic disease, familial history of hyperlipidemia, rheumatologic disease, life expectancy <1 year, and age <18 and >90 years were excluded from the study. Results: There was a statistically significant difference between monocyte/high-density lipoprotein cholesterol ratio, neutrophil /lymphocyte ratio, lymphocyte/monocyte ratio, and platelet/lymphocyte ratio, and fractional flow reserve groups (P <.001). Univariate and multivariate regression analyses were applied among the factors affecting the severity of the lesion detected in fractional flow reserve. Monocyte/high-density lipoprotein cholesterol ratio (odds ratio, 1.25; 95% CI, 1.05-1.47; P =.004), neutrophil/lymphocyte ratio (odds ratio, 3.15; 95% CI, 1.51-6.57; P <.001), hemoglobin A1c (odds ratio, 11.5; 95% CI, 2.76-48.4; P =.001), and lymphocyte/monocyte ratio (odds ratio, 0.27; 95% CI, 0.16-0.44; P =.002) were found to be independent predictors. Conclusions: In this study, we would like to emphasize that simple, fast, and low-cost methods such as monocyte/high-densit y lipoprotein cholesterol ratio, neutrophil/lymphocyte ratio, lymphocyte/monocyte ratio, and platelet/lymphocyte ratio can be parameters related to lesion severity detected in fractional flow reserve. These parameters can be widely used as they are easily accessible and repeatable.
Monocyte
Cite
Citations (0)
배경: 인듐은 휴대전화, TV를 비롯한 디지털기기에서 디스플레이패널의 핵심소재로 사용되는 금속이다. 과거 인듐은 인체에 무해한 금속으로 여겨졌으나 인듐노출로 인한 폐질환자와 사망자 발생을 계기로 일부 국가에서 그 유해성에 대한 연구가 진행되고 있다. 우리나라는 디스플레이 패널 세계시장점유율 1위로 세계 디스플레이 패널의 약 50%를 공급하고 있는 세계 제1의 인듐 소비 국가이다. 하지만 인듐 노출에 의한 건강영향 연구는 미비한 상황이며 폐질환 예방체계 및 검진기준 또한 마련되지 못한 실정이다. 따라서 본 연구를 통해 국내 인듐취급 근로자를 대상으로 백혈구 수치 변화를 확인하고 검진기준 마련을 위한 기초자료를 제공하고자 하였다. 방법: 국내에서 인듐을 취급하는 디스플레이 제조 관련 공장의 남성근로자 156명을 대상으로 하였다. 혈청 인듐은 ICP-MS (bruker), KL-6는 ELISA (EIDIA), 백혈구 분별계수는 자동혈구계수기(sysmex XE-2100)를 이용하여 분석하였다. 자료분석은 version18.0 SPSS Statistics 프로그램을 이용하였고 통계적 유의수준은 p＜0.05로 하였다. 결과: 혈청인듐의 일본참고치 3 μg/L 초과자에서 neutrophil, lymphocyte, monocyte, eosinophil이 의미있는 변화를 보였으며 인듐농도의 증가에 따라 neutrophil의 증가경향과 lymphocyte의 감소경향이 뚜렷하였다(p for trend＜0.05). KL-6의 일본참고치 500 U/mL 초과자에서는 neutrophil, lymphocyte, monocyte의 변화가 의미 있었으며 KL-6의 농도 증가에 따라 neutrophil은 증가하였으나, lymphocyte, monocyte, eosinophil은 뚜렷한 감소경향을 보였다(p for trend＜0.05). 혈청 인듐 또는 KL-6를 독립변수로, 백혈구 분별계수 결과 및 백분율을 독립변수로 회귀분석을 실시한 결과 혈청인듐은 영향력이 낮았으나 KL-6는 neutrophil의 증가와 lymphocyte, monocyte의 감소에 영향을 주는 것으로 분석되었다. 결론: 백혈구 총 수치와 분별계수결과는 모두 정상범위이나 일본의 참고치를 초과한 그룹에서 neutrophils의 증가, lymphocyte와 monocyte의 의미 있는 감소가 확인되었다. 이는 연구대상자에서 neutrophil을 증가시킬 수 있는 감염, 스테로이드 약물치료 등이 없었으며 혈청 인듐 농도 및 간질성폐질환의 마커인 KL-6의 농도 변화에 따라 의미 있는 변화를 보였으므로 직업적 인듐노출로 인한 백혈구 수치의 변화라고 판단된다. 고농도의 인듐노출에서도 변화의 폭이 작아 진단지표로 활용가치는 낮으나 특발성폐섬유증, 폐석면증 등의 질환에서 neutrophil이 증가하였다는 보고가 있으므로 인듐에 의한 간질성 폐질환예방을 염두에 둔 진단항목 개발과 인듐에 의한 폐질환발생기전 파악 등의 후속연구 진행이 요구된다.
Monocyte
Cite
Citations (0)
The effect of sera from patients with Hodgkin's disease (HD) on monocyte dependent ConA and MLC-activation of normal lymphocytes to DNA synthesis was studied. Lymphocyte stimulation was greatly enhanced in the presence of monocytes at a monocyte : lymphocyte ratio of less than or equal to 8 : 1. Higher ratios were usually suppressive. Some HD sera suppressed monocyte mediated enhancement of ConA and MLC-stimulation efficiently. The degree of inhibition by the individual HD serum remained similar in the absence of monocytes and at various monocyte : lymphocyte ratios. Pretreatment of monocytes or lymphocytes with HD serum had no effect. Inhibition was only noted when serum was present during the whole culture period. It is concluded that HD sera do not hamper the activity of monocytes to augment lymphocyte growth. The effect may be explained by direct effects of serum factors on lymphocytes.
Monocyte
Cite
Citations (1)
Deep learning has made spectacular achievements in analysing natural images, but it faces challenges for medical applications partly due to inadequate images.Aiming to classify malignant and benign pulmonary nodules using CT images, we explore different strategies to utilize the state-of-the-art deep convolutional neural networks (CNN).Experiments are conducted using the Lung Image Database Consortium image collection (LIDC-IDRI), which is a public database containing 1018 cases. Three strategies are implemented including to 1) modify some state-of-the-art CNN architectures, 2) integrate different CNNs and 3) adopt transfer learning. Totally, 11 deep CNN models are compared using the same dataset.Study demonstrates that, for the model modification scheme, a concise CifarNet performs better than the other modified CNNs with more complex architectures, achieving an area under ROC curve of AUC = 0.90. Integrated CNN models do not significantly improve the classification performance, but the model complexity is reduced. Transfer learning outperforms the other two schemes and ResNet with fine-tuning leads to the best performance with an AUC = 0.94, as well as the sensitivity of 91% and an overall accuracy of 88%.Model modification, model integration, and transfer learning can play important roles to identify and generate optimal deep CNN models in classifying pulmonary nodules based on CT images efficiently. Transfer learning is preferred when applying deep learning to medical imaging applications.
Transfer of learning
Cite
Citations (54)
Manual Fruit classification is the traditional way of classifying fruits. It is manual contact-labor that is time-consuming and often results in lesser productivity, inconsistency, and sometimes damaging the fruits (Prabha & Kumar, 2012). Thus, new technologies such as deep learning paved the way for a faster and more efficient method of fruit classification (Faridi & Aboonajmi, 2017). A deep convolutional neural network, or deep learning, is a machine learning algorithm that contains several layers of neural networks stacked together to create a more complex model capable of solving complex problems. The utilization of state-of-the-art pre-trained deep learning models such as AlexNet, GoogLeNet, and ResNet-50 was widely used. However, such models were not explicitly trained for fruit classification (Dyrmann, Karstoft, & Midtiby, 2016). The study aimed to create a new deep convolutional neural network and compared its performance to fine-tuned models based on accuracy, precision, sensitivity, and specificity.
Deep Neural Networks
Cite
Citations (2)
Deep learning is now an active research area. Deep learning has done a success in computer vision and image recognition. It is a subset of the Machine Learning. In Deep learning, Convolutional Neural Network (CNN) is popular deep neural network approach. In this paper, we have addressed that how to extract useful leaf features automatically from the leaf dataset through Convolutional Neural Networks (CNN) using Deep Learning. In this paper, we have shown that the accuracy obtained by CNN approach is efficient when compared to accuracy obtained by the traditional neural network.
Deep Neural Networks
Cite
Citations (7)