Manual Microscopic Examination Research Articles

Utility of digital cell imaging as an aid to enhance pathologist peripheral blood smear review for routine and on call practice: A verification study

Abstract Introduction/Objective Microscopic examination of peripheral blood smear (PBS) provides important information to support the diagnosis of hematological diseases. However, manual microscopic examination of PBS is labor-intensive and cannot be done remotely. CellaVision is digital cell imaging technology that provides pre-classification and digital imaging of the red blood cells (RBCs), white blood cells (WBCs), and platelets, myeloid, lymphoid and erythroid precursors. Several studies have evaluated the performance of this technology in the hematology laboratory as tool to assist laboratory technologists; our study evaluates its use for pathologist review of peripheral blood smears both during routine and on call practice. Methods/Case Report Our study included 30 PBS and two hematopathologists with greater than 10 years’ experience in performing PBS review. We compared pathologist’s own impression of microscopic PBS review to their own impression using CellaVision software (intraobserver comparison). Additionally, pathologist to pathologist impression was compared for both methods (interobserver comparison). All 30 cases revealed abnormal findings including, but not limited to: pancytopenia, leukocytosis, leukopenia, anemia, thrombocytopenia, and thrombocytosis. Of the 30 cases, 5 “on-call” were cases when the pathologist had to come on–site for pathologist PBS review. Results (if a Case Study enter NA) The pathologists own PBS interpretation (intraobserver) by microscopy and CellaVision had a concordance of 96.7%. The pathologist to pathologist interpretation of both methods had a concordance of 96.7%.The one discrepant case was a smear with platelet clumping and normal platelet count. On chart review of this case, this discrepancy would have minimal impact on patient care. Conclusion Pathologist use of CellaVision in their routine practice may result in speedier PBS review, particularly in cases with leukopenia, and better communication with laboratory technologists. CellaVision is less reliable in identifying platelet clumps and should not be used in isolation in evaluating thrombocytopenia. Used with other relevant data, CellaVision has the potential to preempt in person microscopic review of PBS when on-call in certain situations.

Acoustofluidic-based microscopic examination for automated and point-of-care urinalysis.

Urinalysis is a heavily used diagnostic test in clinical laboratories; however, it is chronically held back by urine sediment microscopic examination. Current instruments are bulky and expensive to be widely adopted, making microscopic examination a procedure that still relies on manual operations and requires large time and labor costs. To improve the efficacy and automation of urinalysis, this study develops an acoustofluidic-based microscopic examination system. The system utilizes the combination of acoustofluidic manipulation and a passive hydrodynamic mechanism, and thus achieves a high throughput (1000 μL min-1) and a high concentration factor (95.2 ± 2.1 fold) simultaneously, fulfilling the demands for urine examination. The concentrated urine sample is automatically dispensed into a hemocytometer chamber and the images are then analyzed using a machine learning algorithm. The whole process is completed within 3 minutes with detection accuracies of erythrocytes and leukocytes of 94.6 ± 3.5% and 95.1 ± 1.8%, respectively. The examination outcome of urine samples from 50 volunteers by this device shows a correlation coefficient of 0.96 compared to manual microscopic examination. Our system offers a promising tool for automated urine microscopic examination, thus it has potential to save a large amount of time and labor in clinical laboratories, as well as to promote point-of-care urine testing applications in and beyond hospitals.

Urinary renal tubular epithelial cells and casts as predictors of renal outcomes in patients with biopsy-proven diabetic nephropathy.

Urine sediment examination is a time-tested and non-invasive diagnostic tool. Thisstudy investigated the characteristics of urine sediment and its association with severity and renal outcomes in diabetic nephropathy (DN) patients. A total of 201 biopsy-proven diabetic nephropathy patients (according to the pathological classification of diabetic nephropathyproposed by the Renal Pathology Society in 2010)who underwent manual urine sediment microscopic examination were included. We compared the clinicopathologicalcharacteristics of diabetic nephropathy patients with and without urinary renal tubular epithelial cells (RTECs) or renal tubular epithelial cell casts. The predictive value of urinary renal tubular epithelial cells or renal tubular epithelial cell casts for renal outcomes in diabetic nephropathy was analyzed. Fifty of 201 (24.9%) diabetic nephropathy patients had renal tubular epithelial cells or renal tubular epithelial cell casts in urine sediment. Diabetic nephropathy patients with renal tubular epithelial cells or renal tubular epithelial cell casts in urine sediment had a significantly higher level of proteinuria [6.0 (3.1, 9.7)vs.3.6 (1.8, 6.8) g/24h, p = 0.001], higher serum creatinine [227.9 (151.6, 338.1)vs.177.0 (104.4, 288.4) μmol/L, p = 0.016] and lower estimated glomerular filtration rate (eGFR) [25.8 (15.8, 44.8)vs.35.7 (19.9, 65.0) mL/min/1.73m2, p = 0.025] than those without.Cox regression analysis demonstrated that the presence of urinaryrenal tubular epithelial cells or renal tubular epithelial cell casts was independently associated with the development of end-stage kidney disease (ESKD) in diabetic nephropathy patients [HR 1.670, 95% CI (1.042, 2.676), p = 0.033]. Adding the presence of urinary renal tubular epithelial cells or renal tubular epithelial cell casts to the predictive model could improve the effectiveness of the model for predicting the risk of ESKD within one year after renal biopsy. The presence of urinary renal tubular epithelial cells or renal tubular epithelial cell casts was associated with more severe kidney injury and worse renal outcomes in patients with diabetic nephropathy, thus perhaps providing a noninvasive biomarker for predicting diabetic nephropathy.

AI DRIVEN BLOOD CELL ANALYSIS FOR MALARIA AND BLOOD CANCER SCREENING

Malaria and blood Cancer is one of the deadliest diseases cross the globe. This is caused by the bite of female Anopheles mosquito that transmits the Plasmodium parasites. Some current malaria detection techniques include manual microscopic examination and RDT. These approaches are vulnerable to human mistakes. Early detection of malaria can help in reducing the death rates across the globe.Deep Learning can emerge as a highly beneficial solution in the diagnosis of disease. This model gives a faster and cheaper method for detecting plasmodium parasites. It Was Designed to Identify Malaria Parasites and Cancerous Cell Presence in Blood Using Images of Blood Samples which got Tested with Giemsa-stain. Convolutional neural networks (CNNs) are then trained on these extracted features to accurately classify blood cell images into disease positive or disease-negative categories. The CNN detects malaria parasites in microscopic images by classifying them into parasitized and healthy cells, enabling the detection of malaria parasites. The proposed model consists of three convolutional layers and fully connected layers each. The neural network presented is a cascade of several convolutional layers having multiple filters present in layers, which yields the exceptionally good accuracy as per the available resources. The experiment results demonstrate a significant improvement in malaria parasite and blood cancer recognition. Compared to traditional manual microscopy, the proposed system is more accurate and faster. Finally, this study demonstrates the need to provide robust and efficient solutions by leveraging state-of-the-art technologies to combat malaria and blood cancer. Keywords : Deep Learning, CNN, Image analysis , InceptionresnetV2

Verification and Application Evaluation of Intelligent Audit Rules for The UN9000 Urine Analysis System

Objective: To apply and verify the application of intelligent audit rules for urine analysis by Cui et al. Method: A total of 1139 urine samples of hospitalized patients in Tai’an Central Hospital from September 2021 to November 2021 were randomly selected, and all samples were manually microscopic examined after the detection of the UN9000 urine analysis line. The intelligent audit rules (including the microscopic review rules and manual verification rules) were validated based on the manual microscopic examination and manual audit, and the rules were adjusted to apply to our laboratory. The laboratory turnaround time (TAT) before and after the application of intelligent audit rules was compared. Result: The microscopic review rate of intelligent rules was 25.63% (292/1139), the true positive rate, false positive rate, true negative rate, and false negative rate were 27.66% (315/1139), 6.49% (74/1139), 62.34% (710/1139) and 3.51% (40/1139), respectively. The approval consistency rate of manual verification rules was 84.92% (727/856), the approval inconsistency rate was 0% (0/856), the interception consistency rate was 12.61% (108/856), and the interception inconsistency rate was 0% (0/856). Conclusion: The intelligence audit rules for urine analysis by Cui et al. have good clinical applicability in our laboratory.

Deep learning-assisted smartphone-based quantitative microscopy for label-free peripheral blood smear analysis.

Hematologists evaluate alterations in blood cell enumeration and morphology to confirm peripheral blood smear findings through manual microscopic examination. However, routine peripheral blood smear analysis is both time-consuming and labor-intensive. Here, we propose using smartphone-based autofluorescence microscopy (Smart-AM) for imaging label-free blood smears at subcellular resolution with automatic hematological analysis. Smart-AM enables rapid and label-free visualization of morphological features of normal and abnormal blood cells (including leukocytes, erythrocytes, and thrombocytes). Moreover, assisted with deep-learning algorithms, this technique can automatically detect and classify different leukocytes with high accuracy, and transform the autofluorescence images into virtual Giemsa-stained images which show clear cellular features. The proposed technique is portable, cost-effective, and user-friendly, making it significant for broad point-of-care applications.

Detection of circulating plasma cells in peripheral blood using deep learning-based morphological analysis.

The presence of circulating plasma cells (CPCs) is an important laboratory indicator for the diagnosis, staging, risk stratification, and progression monitoring of multiple myeloma (MM). Early detection of CPCs in the peripheral blood (PB) followed by timely interventions can significantly improve MM prognosis and delay its progression. Although the conventional cell morphology examination remains the predominant method for CPC detection because of accessibility, its sensitivity and reproducibility are limited by technician expertise and cell quantity constraints. This study aims to develop an artificial intelligence (AI)-based automated system for a more sensitive and efficient CPC morphology detection. A total of 137 bone marrow smears and 72 PB smears from patients with at Zhongshan Hospital, Fudan University, were retrospectively reviewed. Using an AI-powered digital pathology platform, Morphogo, 305,019 cell images were collected for training. Morphogo's efficacy in CPC detection was evaluated with additional 184 PB smears (94 from patients with MM and 90 from those with other hematological malignancies) and compared with manual microscopy. Morphogo achieved 99.64% accuracy, 89.03% sensitivity, and 99.68% specificity in classifying CPCs. At a 0.60 threshold, Morphogo achieved a sensitivity of 96.15%, which was approximately twice that of manual microscopy, with a specificity of 78.03%. Patients with CPCs detected by AI scanning had a significantly shorter median progression-free survival compared with those without CPC detection (18months vs. 34months, p< .01). Morphogo is a highly sensitive system for the automated detection of CPCs, with potential applications in initial screening, prognosis prediction, and posttreatment monitoring for MM patients. Diagnosing and monitoring multiple myeloma (MM), a type of blood cancer, requires identifying and quantifying specific cells called circulating plasma cells (CPCs) in the blood. The conventional method for detecting CPCs is manual microscopic examination, which is time-consuming and lacks sensitivity. This study introduces a highly sensitive CPC detection method using an artificial intelligence-based system, Morphogo. It demonstrated remarkable sensitivity and accuracy, surpassing conventional microscopy. This advanced approach suggests that early and accurate CPC detection is achievable by morphology examination, making efficient CPC screening more accessible for patients with MM. This innovative system has the potential to be used in the diagnosis and risk assessment of MM.

Consistency Analysis of the UF-1500 and UF-5000 Automated Urine Particle Analyzers.

The aim was to evaluate the consistency of the results between the UF-1500 and UF-5000, fully automated urine particle analyzers. A total of 554 randomly selected inpatient and outpatient urine samples were collected for analysis using the UF-1500, the UF-5000, and by manual microscopic examination. The coincidence rate, intraday repeatability, and interday reproducibility were evaluated on the UF-1500 and UF-5000. To analyze the review flags from the UF-1500, the UF-1500 results were compared to manual microscopy as the gold standard. The repeatability of red blood cells (RBCs), white blood cells (WBCs), epithelial cells (ECs), casts, and bacteria using the UF-1500 and UF-5000 is expressed as the relative standard deviations of the intraday and inter-day measurements. For the UF-1500, the relative standard deviation values ranged from 5.9% to 12.6% and 4.9% to 17.2% for the low and 1.6% to 9.3% and 2.3% to 16.9% for the high samples, respectively. The correlation co-efficient for RBCs, WBCs, ECs, SECs, casts, crystals, and bacteria for the UF-1500 were 0.981, 0.993, 0.968, 0.963, 0.821, 0.783, and 0.992, respectively. Review samples from the UF-1500 were confirmed by microscopic examination. Review flags for all 554 samples included 3 samples with "DEBRIS High" and 23 samples with "RBCs/YLC Abnormal classification". The identification of various urine components by both instruments meets laboratory requirements. These two instruments with different performances have specific characteristics and should be used based upon the needs of each laboratory.

Evaluation of the fully automated urine particle analyzer UF-1500.

This study primarily assessed the performance of the UF-1500, the novel and compact model of the fully automated urine particle analyzer and evaluated its performance against the existing UF-5000 instrument. A total of 648 residual urine specimens were randomly collected and examined using both the UF-1500 and UF-5000 instruments as well as manual microscopy. For each parameter, the concordance rates and detection accuracy of the UF-1500 against manual microscopy were compared with the UF-5000. The concordance rates between the UF-1500 and manual microscopy were 75.3%-98.5%. The UF-1500 concordance rates within one group agreement were observed to be >90%, for all parameters except for YLCs. The differences within one group agreement between the UF-1500 and manual microscopy were insignificant, in comparison to the UF-5000, with exceptions noted for ECs and YLCs. The sensitivity and specificity of the UF-1500 for RBCs, WBCs, Squa.ECs, and BACT exceeded 80%, while the positive predictive values of ECs and CASTs were below 70%. The UF-1500 exhibited a performance that was comparable to the existing instrument, the UF-5000, and was suitable to be introduced in clinical practice. For the samples with suspected false-positive or false-negative results, a manual microscopic examination is required for accurate testing.

Comparison of the performance of two automatic cell morphology analyzers for peripheral-blood leukocyte morphology analysis: Mindray MC-100i and Sysmex DI-60.

To compare the morphological classification ability of peripheral-blood leukocytes of the automatic cell morphology analyzers MC-100i and DI-60. (1) MC-100i and DI-60 were used to analyze leukocytes in 432 venous blood samples collected from three tertiary hospitals across China. The preclassification results were compared with the results reported by senior morphological experts (postclassification results) to evaluate the accuracy, sensitivity, specificity, and consistency of leukocyte preclassification for both instruments. (2) In 200 of the 432 blood samples, morphological experts conducted manual microscopic examination for various types of leukocytes. The correlation between the MC-100i and DI-60 leukocyte postclassification results and the expert microscopist results were analyzed. (1) MC-100i preclassified leukocytes and nucleated red blood cells (RBCs). Compared with the postclassification results, the total leukocyte preclassification accuracy of MC-100i was 97.16%, while that of DI-60 was 87.24%. The sensitivity of MC-100i to abnormal cells (including blasts, promyelocytes, neutrophilic myelocytes, neutrophilic metamyelocytes, reactive lymphocytes, abnormal promyelocytes, plasma cells, abnormal lymphocytes and nucleated RBCs) was 90.24%, which was significantly higher than the 50.72% sensitivity of DI-60. (2) Comparing the postclassification results with manual microscopy, except for reactive lymphocytes and basophils, the MC-100i and DI-60 results had good correlations with various leukocyte types and nucleated RBCs (r > 0.85), and MC-100i was better than DI-60 in the recognition of basophils. Both MC-100i and DI-60 have good detection ability for five normal types of leukocytes in peripheral blood. MC-100i has significantly better detection sensitivity for abnormal cells in peripheral blood than DI-60.

Artificial intelligence-based tools applied to pathological diagnosis of microbiological diseases.

Infectious diseases still threaten the global community, especially in resource-limited countries. An accurate diagnosis is paramount to proper patient and public health management. Identification of many microbes still relies on manual microscopic examination, a time-consuming process requiring skilled staff. Thus, artificial intelligence (AI) has been exploited for identification of microorganisms. A systematic search was carried out using electronic databases looking for studies dealing with the application of AI to pathology microbiology specimens. Of 4596 retrieved articles, 110 were included. The main applications of AI regarded malaria (54 studies), bacteria (28), nematodes (14), and other protozoa (11). Most publications examined cytological material (95, 86%), mainly analyzing images acquired through microscope cameras (65, 59%) or coupled with smartphones (16, 15%). Various deep-learning strategies were used for the analysis of digital images, achieving highly satisfactory results. The published evidence suggests that AI can be reliably utilized for assisting pathologists in the detection of microorganisms. Further technologic improvement and availability of datasets for training AI-based algorithms would help expand this field and widen its adoption, especially for developing countries.

Robust Image Processing Framework for Intelligent Multi-Stage Malaria Parasite Recognition of Thick and Thin Smear Images.

Malaria is a pressing medical issue in tropical and subtropical regions. Currently, the manual microscopic examination remains the gold standard malaria diagnosis method. Nevertheless, this procedure required highly skilled lab technicians to prepare and examine the slides. Therefore, a framework encompassing image processing and machine learning is proposed due to inconsistencies in manual inspection, counting, and staging. Here, a standardized segmentation framework utilizing thresholding and clustering is developed to segment parasites' stages of P. falciparum and P. vivax species. Moreover, a multi-stage classifier is designed for recognizing parasite species and staging in both species. Experimental results indicate the effectiveness of segmenting thick smear images based on Phansalkar thresholding garnered an accuracy of 99.86%. The employment of variance and new transferring process for the clustered members, enhanced k-means (EKM) clustering has successfully segmented all malaria stages with accuracy and an F1-score of 99.20% and 0.9033, respectively. In addition, the accuracies of parasite detection, species recognition, and staging obtained through a random forest (RF) accounted for 86.89%, 98.82%, and 90.78%, respectively, simultaneously. The proposed framework enables versatile malaria parasite detection and staging with an interactive result, paving the path for future improvements by utilizing the proposed framework on all others malaria species.

Consistency analysis of the Sysmex UF-5000 and Atellica UAS 800 urine sedimentation analyzers.

ObjectiveTo evaluate the consistency between the results of Sysmex UF‐5000 system and Atellica® UAS 800 Urine Sediment Analyzer.MethodsA total of 636 random urine samples were collected from inpatients and outpatients from March to September 2021. Urine was collected for analysis by the Sysmex UF‐5000, Atellica UAS 800 systems, and manual microscopic examination. The results of manual microscopy as the gold standard, the coincidence rate and false‐negative rate of Sysmex UF‐5000 and Atellica UAS 800 systems in the detection of red blood cells, white blood cells, and casts were calculated.ResultsThe coincidence rates of red blood cells, white blood cells, and cast, crystals, and other sediment components for the Sysmex UF‐5000 system were 85.37%, 87.89%, 91.67%, 88.36%, and 71.86%. The false‐negative rates were 28.47%, 3.75%, 68.97%, 37.25%, and 30.63%. The coincidence rates of red blood cells, white blood cells, and cast, crystals, and other sediment components for the Atellica UAS 800 system were 85.06%, 90.25%, 59.12%, 91.67%, and 67.45% and the false‐negative rates were 60.42%, 21.25%, 36.21%, 19.64%, and 35.80%.ConclusionTwo instruments are superior in the detection of red blood cells and white blood cells. The Atellica UAS 800 system with image review has a good coincidence rate in the identification of crystals and casts. The identification of various sediment components in urine by both instruments meets the laboratory requirements. Two instruments with different methodologies have their own characteristics, and we should reasonably use them according to the conditions of the laboratory.

Accurate species identification of food-contaminating beetles with quality-improved elytral images and deep learning.

Food samples are routinely screened for food-contaminating beetles (i.e., pantry beetles) due to their adverse impact on the economy, environment, public health and safety. If found, their remains are subsequently analyzed to identify the species responsible for the contamination; each species poses different levels of risk, requiring different regulatory and management steps. At present, this identification is done through manual microscopic examination since each species of beetle has a unique pattern on its elytra (hardened forewing). Our study sought to automate the pattern recognition process through machine learning. Such automation will enable more efficient identification of pantry beetle species and could potentially be scaled up and implemented across various analysis centers in a consistent manner. In our earlier studies, we demonstrated that automated species identification of pantry beetles is feasible through elytral pattern recognition. Due to poor image quality, however, we failed to achieve prediction accuracies of more than 80%. Subsequently, we modified the traditional imaging technique, allowing us to acquire high-quality elytral images. In this study, we explored whether high-quality elytral images can truly achieve near-perfect prediction accuracies for 27 different species of pantry beetles. To test this hypothesis, we developed a convolutional neural network (CNN) model and compared performance between two different image sets for various pantry beetles. Our study indicates improved image quality indeed leads to better prediction accuracy; however, it was not the only requirement for achieving good accuracy. Also required are many high-quality images, especially for species with a high number of variations in their elytral patterns. The current study provided a direction toward achieving our ultimate goal of automated species identification through elytral pattern recognition.

Enumeration and speciation of coccidia affecting turkeys using flow cytometry method

Enumeration of Eimeria oocysts is a common practice in monitoring coccidiosis in turkeys; however, the conventional method of manual microscopic examination of Eimeria oocysts is time-consuming. Previously, we used flow cytometry ( FCM ) to enumerate and speciate coccidia affecting chickens and here, we extended those findings to turkey coccidia species, E. adenoides and E. meleagrimitis . Using FCM, a commercial vaccine containing these species was used to optimize the scatter-plot parameters, including Forward-(size) and Side-(shape/granularity) scatter Area, Height, and Width patterns for Eimeria . The two Eimeria species populations in the vaccine were accurately phenotyped and the gated populations were then sorted using a Cell sorter instrument to obtain pure oocyst suspensions. The individual Eimeria species identity of sorted oocysts was confirmed by PCR using species-specific primers. A significant ( P = 0.0466) correlation (R = 0.9893) in the total oocyst count between FCM and manual methods were observed. Furthermore, when FCM was employed to analyze farm fecal samples, the close similarities in the oocyst morphologies coupled with organic debris particulate interference prevented a precise separation of these 2 species resulting in a lack of oocyst count ( OPG ) correlation between the 2 methods. The OPG counts by FCM were much lower than the manual method; however, a partial OPG trend between the two methods was observed only at the early timepoint collections during a period of 35 d. Collectively, our findings showed that FCM can be used in the enumeration of turkey Eimeria oocysts with a potential scope for a more precise enumeration and speciation in field samples.

Automated Analysis of Blood Smear Images for Leukemia Detection: A Comprehensive Review

Leukemia, the malignancy of blood-forming tissues, becomes fatal if not detected in the early stages. It is detected through a blood smear test that involves the morphological analysis of the stained blood slide. The manual microscopic examination of slides is tedious, time-consuming, error-prone, and subject to inter-observer and intra-observer bias. Several computerized methods to automate this task have been developed to alleviate these problems during the past few years. However, no exclusive comprehensive review of these methods has been presented to date. Such a review shall be highly beneficial for novice readers interested in pursuing research in this domain. This article fills the void by presenting a comprehensive review of 149 papers detailing the methods used to analyze blood smear images and detect leukemia. The primary focus of the review is on presenting the underlying techniques used and their reported performance, along with their merits and demerits. It also enumerates the research issues that have been satisfactorily solved and open challenges still existing in the domain.

Detection of nucleated red blood cells using the Mindray BC-6800Plus hematology analyzer: a clinical performance evaluation.

With the development of instrument technology, the functions and detection methods of automatic blood cell analyzers have become more complex. To ensure optimal clinical applicability, it is crucial to select an automatic blood cell analyzer with excellent clinical detection performance. This study evaluated the latest Mindray BC-6800Plus automatic blood cell analyzer and assessed its performance in the detection of nucleated red blood cells (NRBCs). A total of 490 clinical blood samples were used to assess the performance of the instrument, including parameters such as precision, linearity, conformity rate of manual microscopic examination, carryover, and limit of quantitation. The instrument showed a small carryover (≤0.02) and excellent linearity (R2≥0.9986). The reproducibility of the sample tests was satisfactory, and the coefficient of variation (CV) of the test results [0.98-1.72% and 0.62-6.97% for white blood cells (WBCs) and NRBCs, respectively] were significantly lower than that declared by the manufacturer (2.5% and 20% for WBCs and NRBCs, respectively). Thus, the BC-6800Plus satisfies the requirements of clinical testing. Two separate Mindray BC-6800Plus machines were tested and found to be in good agreement with each other and with manual microscopy methods. Furthermore, WBC and NRBC counts were highly consistent with results obtained using the XN-9100 blood analyzer. The Mindray BC-6800Plus is an excellent analyzer that can provide timely and accurate reports for clinical laboratory detection of NRBC.

A Study for Accurate Reporting of Bacteria in Urine by Manual Microscopic Examination

A Study for Accurate Reporting of Bacteria in Urine by Manual Microscopic Examination

Evaluation of humoral pattern detection performance of Mindray BC-6000PLUS blood analyzer based on WS/T 662-2020.

Manual microscopic examination is the gold standard of humoral cell count test. However, it has some limitations and cannot fully meet clinical needs. Compared with the manual method, the automatic blood cell analyzer has the advantages of a high degree of automation, minimal error, high speed, high precision, and easy standardization. This study intends to verify the detection performance of the body fluid model of the Mindray BC-6000PLUS automatic hematology analyzer. This study was performed in accordance with the International Committee for Standardization in Haematology (ICSH) Hematology Analyzer Evaluation Guide (version 2014) and the requirements of WS/T662-2020 "Clinical humoral examination technique". The humoral white blood cell-body fluid (WBC-BF), humoral red blood cell-body fluid (RBC-BF), monocyte (MN), polymorphonuclear (PMN) were measured to verify the performance indicators of the instrument, including background counting, intra-batch precision, accuracy, carrying contamination rate, and linear range. Referring to the WS/T514-2017 (Establishment and verification of detection capability for clinical laboratory measurement procedures), the limit of blank (LoB) and limit of detection (LoD) values of WBC-BF and RBC-BF in the humoral mode of the instrument were established. The blank count of WBC-BF and RBC-BF, and contamination of Mindray BC-6000PLUS analyzer were zero; the coefficient of variation (CV) of intra-batch precision at different levels of each item was less than 10%. There was a high correlation between instrument test results and manual microscopic examination results (r>0.95). The linear range of the instrument was wide, and the linear verification parameters was good (R2>0.999). The LoB value and LoD value of WBC-BF established by the instrument were 0×109/L and 0.004×109/L, respectively. The LoB value and LoD value of the RBC-BF established by the instrument were 0×1012/L and 0.004×1012/L, respectively. The lower detection limits of WBC-BF and RBC-BF were set as 0.004×109/L and 0.004×1012/L, respectively. All performance indicators of the Mindray BC-6000PLUS automatic blood analyzer met the requirements of the manufacturer's criteria. This instrument can fulfill the requirement of body fluid sample routine test in clinical practice.

Abstract PO-069: A deep learning model assists urine cytology reporting with computational estimates of the nuclear/cytoplasmic ratios of the urothelial cells based on the Paris System

Abstract The Paris System (TPS) for Reporting Urinary Cytology is an international system with standardized terminology and cytomorphologic criteria that aims for better risk stratification to identify high-grade urothelial carcinoma in urinary cytology specimens. Based on TPS, the nuclear/cytoplasmic ratio (N/C ratio) of the observed urothelial cells is a crucial criterion and the cutoff values of 0.5 and 0.7 are used for the diagnosis of Atypical Urothelial Cells (AUC) and High-Grade Urothelial Carcinoma (HGUC)/Suspicious for HGUC respectively. Although the N/C ratio is a clear and measurable criterion, there are still challenges for adopting TPS into clinical practice. As urine cytology mainly relies on manual microscopic examination by cytopathologists and cytotechnologists, determining the N/C ratio through the naked eyes results in intra- and inter-observer variations and therefore generates subjective, inconsistent, and unreproducible reports. To overcome the aforementioned issues, the authors developed a deep learning-based image analysis model to assist calculating the N/C ratio and classifying urothelial cells computationally. Deidentified urine cytology slides were digitalized as whole-slide images (WSIs) for analyzing. An “active learning” method was used to improve results over time with only a small amount of upfront training data required. First, a small amount of annotated WSIs were provided to train the initial computational model. Next, the experts reviewed the results predicted by the initial model, and provide feedback for the model to learn and improve. The above steps were repeated by experts reviewing the “improved model” and providing feedbacks, thus improving the initial/intermediate models continuously until a satisfactory result was achieved. Preliminary results showed that our model was able to determine the areas of cytoplasm and nucleus in urothelial cells as well as calculate the N/C ratios using the WSI. Moreover, it enumerates the AUC and HGUC cells on the WSI and provides quantification and distribution information on the whole slide scale. This computational model provides a significant advantage as an objective tool to assist urine cytology analysis of the N/C ratio with greater accuracy and reproducibility. Citation Format: Wei-Lei Yang, Chi-Bin Li, Yen-Chuan Ou, Yi-Sheng Lin, Tang-Yi Tsao, Ming-Chen Chang, Jen-Fan Hang, Tien-Jen Liu. A deep learning model assists urine cytology reporting with computational estimates of the nuclear/cytoplasmic ratios of the urothelial cells based on the Paris System [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-069.