Does the ThinPrep Imaging System increase the detection of high-risk HPV-positive ASC-US and AGUS? The Women and Infants Hospital experience with over 200,000 cervical cytology cases

Digital cytology: Current status and future prospects

AIMS

In this paper, we reviewed the basic principle and the currentstatus of digital cytopathology.

MATERIALS AND METHODS

We reviewed the published papers on digitalcytology and analysed its future prospects.

RESULTS

Virtualcytology using digital platform is being increasingly used to render diagnosisrather than conventional glass slide microscopy. Whole slide imaging (WSI)offers the prospect of true virtual microscopy and in the near future, may evenreplace glass slides in routine practice. It may be pivotal in diagnosing andtraining pathology graduates faster and more accurately.

CONCLUSION

The digital cytopathology is a promising field and may have great impact indiagnosis, research and teaching.

Diagnostic utility of ThinPrep Imaging System® for detecting atypical glandular cells in cervical smear samples

INTRODUCTION

The ThinPrep Imaging System® (TIS) is an automated system that has now been used for over 20 years in the primary screening of ThinPrep liquid-based cervical samples. Although there are a lot of publications about the diagnostic utility of this method in squamous cell lesions, which has advantages such as time-saving and standardization, there are only a few publications on this issue in glandular cell lesions in the literature. We aimed in this study to investigate the diagnostic utility of the system in the detection of premalignant and malignant glandular lesions in cervical smears.

MATERIAL AND METHOD

Our study was conducted retrospectively, and a total of 126 cervical smear samples between 2010 and 2022 that have histological confirmation of endometrial adenocarcinoma (EAC), endocervical adenocarcinoma (ECAC), or adenocarcinoma in situ (AİS), were included. These samples were re-evaluated by manual and TIS by two experienced pathologists, and the results were compared in terms of sensitivity.

RESULTS

We found out that 70 of the 126 smear samples have atypical glandular cells. We detect 48 cases (48/70) (sensitivity 68.5%) in manual examination, however TIS successfully determined 66 cases (66/70) (sensitivity 94.3%). In 4 cases (5.7%) TIS could not detect the atypical cells within the 22 areas.

CONCLUSION

TIS is quite an effective method with a high sensitivity for detecting atypical glandular cells in cervical smears, like detecting squamous cell anomalies. Imposing this system in our laboratory and using them appropriately, save us time and help to ensure standardization. Additionally, it may be a good way to adopt artificial intelligence and digital pathology in today's world.

Robust whole slide image analysis for cervical cancer screening using deep learning

Computer-assisted diagnosis is key for scaling up cervical cancer screening. However, current recognition algorithms perform poorly on whole slide image (WSI) analysis, fail to generalize for diverse staining and imaging, and show sub-optimal clinical-level verification. Here, we develop a progressive lesion cell recognition method combining low- and high-resolution WSIs to recommend lesion cells and a recurrent neural network-based WSI classification model to evaluate the lesion degree of WSIs. We train and validate our WSI analysis system on 3,545 patient-wise WSIs with 79,911 annotations from multiple hospitals and several imaging instruments. On multi-center independent test sets of 1,170 patient-wise WSIs, we achieve 93.5% Specificity and 95.1% Sensitivity for classifying slides, comparing favourably to the average performance of three independent cytopathologists, and obtain 88.5% true positive rate for highlighting the top 10 lesion cells on 447 positive slides. After deployment, our system recognizes a one giga-pixel WSI in about 1.5 min.

Computer-assisted diagnosis is key for scaling up cervical cancer screening, but current algorithms perform poorly on whole slide image analysis and generalization. Here, the authors present a WSI classification and top lesion cell recommendation system using deep learning, and achieve comparable results with cytologists.

Cervical cancer: Automation of Pap test screening

BACKGROUND

Cervical cancer progresses slowly, increasing the chance of early detection of pre-neoplastic lesions via Pap exam test and subsequently preventing deaths. However, the exam presents both false-negatives and false-positives results. Therefore, automatic methods (AMs) of reading the Pap test have been used to improve the quality control of the exam. We performed a literature review to evaluate the feasibility of implementing AMs in laboratories.

METHODS

This work reviewed scientific publications regarding automated cytology from the last 15 years. The terms used were "Papanicolaou test" and "Automated cytology screening" in Portuguese, English, and Spanish, in the three scientific databases (SCIELO, PUBMED, MEDLINE).

RESULTS

Of the resulting 787 articles, 34 were selected for a complete review, including three AMs: ThinPrep Imaging System, FocalPoint GS Imaging System and CytoProcessor. In total, 1 317 148 cytopathological slides were evaluated automatically, with 1 308 028 (99.3%) liquid-based cytology slides and 9120 (0.7%) conventional cytology smears. The AM diagnostic performances were statistically equal to or better than those of the manual method. AM use increased the detection of cellular abnormalities and reduced false-negatives. The average sample rejection rate was ≤3.5%.

CONCLUSION

AMs are relevant in quality control during the analytical phase of cervical cancer screening. This technology eliminates slide-handling steps and reduces the sample space, allowing professionals to focus on diagnostic interpretation while maintaining high-level care, which can reduce false-negatives. Further studies with conventional cytology are needed. The use of AM is still not so widespread in cytopathology laboratories.

Performance of A Convolutional Neural Network in Screening Liquid Based Cervical Cytology Smears

Context

Cervical cancer is the second most common cancer in women. The liquid based cervical cytology (LBCC) is a useful tool of choice for screening cervical cancer.

Aims

To train a convolutional neural network (CNN) to identify abnormal foci from LBCC smears.

Settings and Design

We have chosen retrospective study design from archived smears of patients undergoing screening from cervical cancer by LBCC smears.

Materials and Methods

2816 images, each of 256 × 256 pixels, were prepared from microphotographs of these LBCC smears, which included 816 "abnormal" foci (low grade or high grade squamous intraepithelial lesion) and 2000 'normal' foci (benign epithelial cells and reactive changes). The images were split into three sets, Training, Testing, and Evaluation. A convolutional neural network (CNN) was developed with the python programming language. The CNN was trained with the Training dataset; performance was assayed concurrently with the Testing dataset. Two CNN models were developed, after 20 and 10 epochs of training, respectively. The models were then run on the Evaluation dataset.

Statistical Analysis Used

A contingency table was prepared from the original image labels and the labels predicted by the CNN.

Results

Combined assessment of both models yielded a sensitivity of 95.63% in detecting abnormal foci, with 79.85% specificity. The negative predictive value was high (99.19%), suggesting potential utility in screening. False positives due to overlapping cells, neutrophils, and debris was the principal difficulty met during evaluation.

Conclusions

The CNN shows promise as a screening tool; however, for its use in confirmatory diagnosis, further training with a more diverse dataset will be required.

Implementation of a real-time reference and calibration grid platform for improved screening - mapping in Pap test slides

Cervical cancer screening based on the Papanicolaou (Pap) test is a widely applied but not always efficient practice for detecting Human Papillomavirus (HPV) mediated lesions, partially due to a non-systematic and inadequate screening process. Our aim was to introduce an inexpensive easy-to-use direct screening platform for improved detection of abnormal cells indicative of underlying cervical neoplasia as well as persisting HPV infection. By employing a novel, efficient technique of laser-based micromachining, we achieved the fabrication of spatial grids on commercially available coverslips allowing visual segmentation of the slide for efficient screening. Abnormal and formerly diagnosed as negative for intraepithelial lesion or malignancy (NILM) Pap test slides (n = 200) were analyzed by conventional and grid-based screening. Grid-based microscopy led to a more reliable diagnosis compared to the conventional by identifying an increased number of abnormal cells (P = 0.001). It decreased borderline ASCUS, AGC diagnosis, increasing LSIL, HSIL and in situ AdenoCa detection rates closely related with biopsy (P = 0.015; kappa = 0.978). Concerning the set of NILM diagnoses in rapid re-screening, the method upgraded six cases (n = 6) to LSIL (P = 0.001). The proposed technical solution offers a calibration and orientation visual aid during the on-site screening process providing significant advantages compared to expensive digital imaging techniques.

Automatic screening of cervical cells using block image processing

Background

Cervical cancer is the second leading cause of female-specific cancer-related deaths after breast cancer, especially in developing countries. However, the incidence of the disease may be significantly decreased if the patient is diagnosed in the pre-cancerous lesion stage or earlier. In recent years, computer-based algorithms are widely used in cervical cancer screening. Most of the proposed algorithms follow the procedure of segmentation, feature extraction, and then classification. Nevertheless, few of the existing segmentation methods are as flexible and robust as the human visual system, and the complexity of the algorithms makes it difficult for clinical application.

Methods

In this study, a computer-assisted analytical approach is proposed to identify the existence of suspicious cells in a whole slide cervical cell image (WSCCI). The main difference between our method and the conventional algorithm is that the image is divided into blocks with certain size instead of segmented cells, which can greatly reduce the computational complexity. Via data analysis, some texture and color histogram features show significant differences between blocks with and without suspicious cells. Therefore these features can be used as the input of the support vector machine classifier. 1100 non-background blocks (110 suspicious blocks) are trained to build a model, while 1040 blocks (491 non-background blocks) from 12 other WSCCIs are tested to verify the feasibility of the algorithm.

Results

The experimental results show that the accuracy of our method is about 98.98 %. More importantly, the sensitivity, which is more fatal in cancer screening, is 95.0 % according to the images tested in the study, while the specificity is 99.33 %.

Conclusion

The analysis of the algorithm is based on block images, which is different from conventional methods. Although some analysis work should be done in advance, the later processing speed will be greatly enhanced with the establishment of the model. Furthermore, since the algorithm is based on the actual WSCCI, the method will be of directive significance for clinical screening.

Comparative analysis of cervical cytology screening methods and staining protocols for detection rate and accurate interpretation of ASC-H: Data from a high-volume laboratory in Turkey

BACKGROUND

This study evaluated the effectiveness of the ThinPrep(®) Imaging System (TIS) and ThinPrep(®) Pap Stain (TPPS). A comparative analysis was conducted to determine the detection rates of atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion (ASC-H), the ASC:squamous intraepithelial lesion (SIL) ratio, biopsy follow-up for ASC-H in terms of the screening method used (manual screening [MS] vs. TIS screening [TISS]) and the staining protocol (regular Pap stain [RPS] vs. TPPS).

METHODS

This study was performed over two periods. The RPS period included manually screened slides, whereas the TPPS period included TIS + manually screened slides. All data from the study periods were compared using statistical analysis.

RESULTS

The detection rate of ASC-H was significantly higher during the TPPS period than during the RPS period (0.49% vs. 0.23%); this finding is in contrast to the insignificant difference between the screening method periods. The positive predictive value (PPV) of ASC-H cytodiagnosis for cervical intraepithelial neoplasia of grade 2 or more severe histologies was significantly different between manually screened and TIS slides (22.10% vs. 38.55%), in contrast to an insignificant difference between RPS and TPPS periods (37.14% vs. 29.77%).

CONCLUSION

Implementation of the TIS did not change the ASC-H detection rates appreciably. However, the new technology improved PPV for ASC-H cytodiagnosis and enabled the detection of true disease. Our laboratory statistics indicate that the TPPS is not a superior staining protocol and did not increase our diagnostic accuracy for ASC-H compared with RPS.

The clinical impact of including pictures in Papanicolaou test reports

INTRODUCTION

Although the inclusion of representative photomicrographic images in pathology reports remains uncommon, the frequency has increased in recent years. The impact of the inclusion of pictures has not previously been examined in peer-reviewed literature.

MATERIALS AND METHODS

We compared the Papanicolaou (Pap) test interpretations produced by our laboratory before and after the introduction of a requirement to include pictures in reports with interpretations of low-grade squamous intraepithelial lesion (LSIL); atypical squamous cells, cannot rule out high-grade squamous intraepithelial lesion; high-grade squamous intraepithelial lesion; atypical glandular cells; or carcinoma. Atypical squamous cells of undetermined significance (ASC-US) did not require a picture, creating an incentive to interpret borderline cases as ASC-US instead of LSIL. We compared 1810 Pap tests from before the picture requirement to 1807 after.

RESULTS

The number of cases upgraded by the pathologists from a cytotechnologist interpretation of negative or ASC-US to one of the positive interpretations requiring a picture decreased from 99 in the prepicture era to 80 in the picture era (P = 0.19). Conversely, the number of cases downgraded by the pathologists increased from 65 to 98 (P = 0.015). The ASC-US to LSIL (ASC-SIL) ratio for the laboratory went from 0.92 to 1.45.

CONCLUSIONS

The introduction of a picture requirement in Pap test reports significantly affected the practice of pathologists in our laboratory. The ASC-SIL ratio of the laboratory shifted toward the national mean in association with this change.

Screening for cervical cancer using automated analysis of PAP-smears

Cervical cancer is one of the most deadly and common forms of cancer among women if no action is taken to prevent it, yet it is preventable through a simple screening test, the so-called PAP-smear. This is the most effective cancer prevention measure developed so far. But the visual examination of the smears is time consuming and expensive and there have been numerous attempts at automating the analysis ever since the test was introduced more than 60 years ago. The first commercial systems for automated analysis of the cell samples appeared around the turn of the millennium but they have had limited impact on the screening costs. In this paper we examine the key issues that need to be addressed when an automated analysis system is developed and discuss how these challenges have been met over the years. The lessons learned may be useful in the efforts to create a cost-effective screening system that could make affordable screening for cervical cancer available for all women globally, thus preventing most of the quarter million annual unnecessary deaths still caused by this disease.

Increased rate of atypical squamous cells of undetermined significance and declining high-risk human papillomavirus rates following implementation of ThinPrep Imaging System are associated with increased nuclear chromasia

INTRODUCTION

The ThinPrep Imaging System ([Imager], Cytyc Corporation, Hologic, Inc.) for cervical cytology is commonly employed in laboratories. It uses a proprietary stain that allows measurement of cellular DNA content. We evaluated the initial effect of implementing the Imager stain on atypical squamous cells of undetermined significance (ASCUS) and squamous intraepithelial lesion (SIL) diagnoses, and high-risk human papillomavirus (hrHPV) rates.

MATERIALS AND METHODS

This study included 264 Imager-stained ASCUS cases from the first 5 months of Imager use. All cases were analyzed for hrHPV using Hybrid Capture II assay (Digene Corporation, Qiagen). ASCUS/SIL ratios and hrHPV-positive rates were calculated for the laboratory overall and for 3 cytopathologists. ASCUS/SIL ratios were also calculated for 5 cytotechnologists. These metrics were compared with those from the equivalent 5-month period 1 year prior. Slide review of all Imager-stained ASCUS cases was performed to assess for potential factors involved in ASCUS misinterpretation.

RESULTS

The proportion of ASCUS diagnoses increased after Imager stain introduction, from 1.79% to 3.14% (P < 0.001). The proportion of SIL diagnoses also increased, from 0.69% to 1.1% (P < 0.001). However, the hrHPV-positive rate declined from 61.3% to 53.6% (P = 0.104). Review of Imager-stained ASCUS cases showed marked nuclear hyperchromasia, particularly in benign squamous metaplastic cells.

CONCLUSIONS

Imager stain implementation, required for the ThinPrep Imaging System, led to significant increase in ASCUS rates while the hrHPV-positive rate declined. Cytopathologists will need to recalibrate thresholds for an ASCUS diagnosis when interpreting Imager-stained slides. In particular, nuclear hyperchromasia in squamous metaplastic cells should not be overdiagnosed as ASCUS.

Effect of Thin Prep(®) imaging system on laboratory rate and relative sensitivity of atypical squamous cells, high-grade squamous intraepithelial lesion not excluded and high-grade squamous intraepithelial lesion interpretations

Introduction:

Automated screening of Thin Prep^® Papanicolaou Tests has become increasingly common in clinical practice. Increased productivity has initiated laboratory use of the Thin Prep^® Imaging System (TIS). Increased sensitivity is a potential additional benefit of TIS. Published studies have shown an increase in discovery of dysplastic cells. This study evaluates the effect of TIS on the incidence of atypical squamous cells high-grade squamous intraepithelial lesion not excluded (ASC-H) and high-grade squamous intraepithelial lesion (HGSIL) results on Thin Prep^® Pap Tests by comparing TIS-assisted and manual screening findings and the diagnoses on subsequent follow-up in a screening population over a 1-year time period.

Materials and Methods:

A compilation of all ASC-H and HGSIL cases was prepared by conducting a computerized search over a 1-year period (7/06-6/07). The accumulated cases include Thin Prep Pap tests that were both TIS and manually screened. Follow-up results of cytologic and histologic cervical specimens were obtained for a time period extending to 2010. Interpretation utilizing TIS was in place 10 months prior to the study's initiation.

Results:

During the study period 70,522 Pap tests were performed in our laboratory. One third (33%) of Pap tests were screened with assistance of TIS. Manual screening was performed on 47,380 Pap tests of which 153 (0.32%) were interpreted as ASC-H and 164 (0.35%) were interpreted as HGSIL. During the same time period automated screening (TIS) was performed on 23,111 Pap tests. Interpretation of 62 (0.27%) cases provided an ASC-H result, while 71 (0.31%) were HGSIL. Follow-up cervical dysplasia by colposcopic biopsy and cone biopsy was distributed proportionally between TIS and manual screening for both ASC-H and HGSIL categories. Cervical intraepithelial neoplasia (CIN II/III) was identified on follow-up biopsy of 41% TIS cases and 45% manually screened cases for ASC-H. In the HGSIL subset 71% of TIS cases and 69% manually screened cases showed CIN II/III on follow-up. TIS was 26% less sensitive relative to manual screening for ASC-H cases and 3% less sensitive for HGSIL.

Conclusion:

The similar rate of detection using TIS with an equal percentage of histologic correlation for ASC-H and HGSIL lesions on follow-up histology suggests patients screened by the TIS method are being sent for appropriate follow-up surveillance and treatment. A high-grade or possible high-grade lesion is as likely to be detected by TIS as by a manual screen. The similarities in relative sensitivity and specificity in a direct comparison between manual and TIS screening methodologies indicate that TIS compared to manual screening does not affect detection in patients with high-grade cervical lesions.

Human papillomavirus testing versus repeat cytology for triage of minor cytological cervical lesions

BACKGROUND

Atypical squamous cells of undetermined significance (ASCUS) and low-grade squamous intra-epithelial lesions (LSIL) are minor lesions of the cervical epithelium, detectable by cytological examination of cells collected from the surface of the cervix of a woman.Usually, women with ASCUS and LSIL do not have cervical (pre-) cancer, however a substantial proportion of them do have underlying high-grade cervical intra-epithelial neoplasia (CIN, grade 2 or 3) and so are at increased risk for developing cervical cancer. Therefore, accurate triage of women with ASCUS or LSIL is required to identify those who need further management.This review evaluates two ways to triage women with ASCUS or LSIL: repeating the cytological test, and DNA testing for high-risk types of the human papillomavirus (hrHPV) - the main causal factor of cervical cancer.

OBJECTIVES

Main objective To compare the accuracy of hrHPV testing with the Hybrid Capture 2 (HC2) assay against that of repeat cytology for detection of underlying cervical intraepithelial neoplasia of grade 2 or worse (CIN2+) or grade 3 or worse (CIN3+) in women with ASCUS or LSIL. For the HC2 assay, a positive result was defined as proposed by the manufacturer. For repeat cytology, different cut-offs were used to define positivity: Atypical squamous cells of undetermined significance or worse (ASCUS+), low-grade squamous intra-epithelial lesions or worse (LSIL+) or high-grade squamous intra-epithelial lesions or worse (HSIL+).Secondary objective To assess the accuracy of the HC2 assay to detect CIN2+ or CIN3+ in women with ASCUS or LSIL in a larger group of reports of studies that applied hrHPV testing and the reference standard (coloscopy and biopsy), irrespective whether or not repeat cytology was done.

SEARCH METHODS

We made a comprehensive literature search that included the Cochrane Register of Diagnostic Test Accuracy Studies; the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library), MEDLINE (through PubMed), and EMBASE (last search 6 January 2011). Selected journals likely to contain relevant papers were handsearched from 1992 to 2010 (December). We also searched CERVIX, the bibliographic database of the Unit of Cancer Epidemiology at the Scientific Institute of Public Health (Brussels, Belgium) which contains more than 20,000 references on cervical cancer.More recent searches, up to December 2012, targeted reports on the accuracy of triage of ASCUS or LSIL with other HPV DNA assays, or HPV RNA assays and other molecular markers. These searches will be used for new Cochrane reviews as well as for updates of the current review.

SELECTION CRITERIA

Studies eligible for inclusion in the review had to include: women presenting with a cervical cytology result of ASCUS or LSIL, who had undergone both HC2 testing and repeat cytology, or HC2 testing alone, and were subsequently subjected to reference standard verification with colposcopy and colposcopy-directed biopsies for histologic verification.

DATA COLLECTION AND ANALYSIS

The review authors independently extracted data from the selected studies, and obtained additional data from report authors.Two groups of meta-analyses were performed: group I concerned triage of women with ASCUS, group II concerned women with LSIL. The bivariate model (METADAS-macro in SAS) was used to assess the absolute accuracy of the triage tests in both groups as well as the differences in accuracy between the triage tests.

MAIN RESULTS

The pooled sensitivity of HC2 was significantly higher than that of repeat cytology at cut-off ASCUS+ to detect CIN2+ in both triage of ASCUS and LSIL (relative sensitivity of 1.27 (95% CI 1.16 to 1.39; P value < 0.0001) and 1.23 (95% CI 1.06 to 1.4; P value 0.007), respectively. In ASCUS triage, the pooled specificity of the triage methods did not differ significantly from each other (relative specificity: 0.99 (95% CI 0.97 to 1.03; P value 0.98)). However, the specificity of HC2 was substantially, and significantly, lower than that of repeat cytology in the triage of LSIL (relative specificity: 0.66 (95% CI 0.58 to 0.75) P value < 0.0001).

AUTHORS' CONCLUSIONS

HPV-triage with HC2 can be recommended to triage women with ASCUS because it has higher accuracy (significantly higher sensitivity, and similar specificity) than repeat cytology. When triaging women with LSIL, an HC2 test yields a significantly higher sensitivity, but a significantly lower specificity, compared to a repeat cytology. Therefore, practice recommendations for management of women with LSIL should be balanced, taking local circumstances into account.

The Positive Impact of Simultaneous Implementation of the BD FocalPoint GS Imaging System and Lean Principles on the Operation of Gynecologic Cytology

Context.—Our cytology laboratory, like many others, is under pressure to improve quality and provide test results faster while decreasing costs. We sought to address these issues by introducing new technology and lean principles.

Objective.—To determine the combined impact of the FocalPoint Guided Screener (GS) Imaging System (BD Diagnostics–TriPath, Burlington, North Carolina) and lean manufacturing principles on the turnaround time (TAT) and productivity of the gynecologic cytology operation.

Design.—We established a baseline measure of the TAT for Papanicolaou tests. We then compared that to the performance after implementing the FocalPoint GS Imaging System and lean principles. The latter included value-stream mapping, workflow modification, and a first in–first out policy.

Results.—The mean (SD) TAT for Papanicolaou tests before and after the implementation of FocalPoint GS Imaging System and lean principles was 4.38 (1.28) days and 3.20 (1.32) days, respectively. This represented a 27% improvement in the average TAT, which was statistically significant (P &lt; .001). In addition, the productivity of staff improved 17%, as evidenced by the increase in slides screened from 8.85/h to 10.38/h. The false-negative fraction decreased from 1.4% to 0.9%, representing a 36% improvement.

Conclusions.—In our laboratory, the implementation of FocalPoint GS Imaging System in conjunction with lean principles resulted in a significant decrease in the average TAT for Papanicolaou tests and a substantial increase in the productivity of cytotechnologists while maintaining the diagnostic quality of gynecologic cytology.