Implementation and evaluation of a new automated interactive image analysis system

Artificial Intelligence-driven Digital Cytology-based Cervical Cancer Screening: Is the Time Ripe to Adopt This Disruptive Technology in Resource-constrained Settings? A Literature Review

Cervical cancer is still a public health scourge in the developing countries due to the lack of organized screening programs. Though liquid-based cytology methods improved the performance of cervical cytology, the interpretation still suffers from subjectivity. Artificial intelligence (AI) algorithms have offered objectivity leading to better sensitivity and specificity of cervical cancer screening. Whole slide imaging (WSI) that converts a glass slide to a virtual slide provides a new perspective to the application of AI, especially for cervical cytology. In the recent years, there have been a few studies employing various AI algorithms on WSI images of conventional or LBC smears and demonstrating differing sensitivity/specificity or accuracy at detection of abnormalities in cervical smears. Considering the interest in AI-based screening modalities, this well-timed review intends to summarize the progress in this field while highlighting the research gaps and providing future research directions.

A Deep Learning Model for Cervical Cancer Screening on Liquid-Based Cytology Specimens in Whole Slide Images

Simple Summary

In this pilot study, we aimed to investigate the use of deep learning for the classification of whole-slide images of liquid-based cytology specimens into neoplastic and non-neoplastic. To do so, we used a large training and test sets. Overall, the model achieved good classification performance in classifying whole-slide images, demonstrating the promising potential use of such models for aiding the screening processes for cervical cancer.

Abstract

Liquid-based cytology (LBC) for cervical cancer screening is now more common than the conventional smears, which when digitised from glass slides into whole-slide images (WSIs), opens up the possibility of artificial intelligence (AI)-based automated image analysis. Since conventional screening processes by cytoscreeners and cytopathologists using microscopes is limited in terms of human resources, it is important to develop new computational techniques that can automatically and rapidly diagnose a large amount of specimens without delay, which would be of great benefit for clinical laboratories and hospitals. The goal of this study was to investigate the use of a deep learning model for the classification of WSIs of LBC specimens into neoplastic and non-neoplastic. To do so, we used a dataset of 1605 cervical WSIs. We evaluated the model on three test sets with a combined total of 1468 WSIs, achieving ROC AUCs for WSI diagnosis in the range of 0.89–0.96, demonstrating the promising potential use of such models for aiding screening processes.

Comparison of Diagnostic Rates and Concordance with Subsequent Surgical Resections between Conventional Smear and ThinPrep Preparations versus ThinPrep Only in Thyroid Fine Needle Aspiration (T-FNA) Specimens

BACKGROUND

Thyroid fine needle aspiration (T-FNA) is a mainstay in management of thyroid nodules. However, the preparation of T-FNA specimens varies across institutions. Prior studies have compared diagnostic rates between different specimen preparations of T-FNA specimens and their associated advantages and disadvantages. However, few have compared the rates of all diagnostic categories of The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) between liquid-based preparations (LBPs) and a combination of LBP and conventional smear (CS) preparations. Our study compares TBSRTC diagnostic rates between these 2 cohorts and correlates cytologic diagnoses with subsequent thyroid resections to evaluate rates of neoplasia (RON) and malignancy (ROM).

METHODS

584 consecutive thyroid FNA specimens were collected and stratified by preparation type (ThinPrep [TP] vs. CS & TP). Diagnostic rates for each TBSRTC diagnostic category were calculated. The institution's electronic medical records database was searched for histologic diagnoses of previously sampled thyroid nodules to evaluate the RON and ROM.

RESULTS

Of 584 thyroid FNA specimens, 73 (12.5%) and 511 (87.5%) were evaluated by TP only and CS & TP, respectively, reflecting the predominance of rapid on-site evaluation (ROSE) with CS for T-FNAs at our institution. Of the TP only and CS & TP cohorts, 29 (39.7%) and 98 (19.2%) had subsequent resections, respectively. The frequency of non-diagnostic cases was lower in the CS & TP cohort (12.7% vs. 26%). While the diagnostic rate of follicular lesion of undetermined significance was similar for both cohorts, SFN categorization was only utilized in the CS & TP cohort (1.5% vs. 0%). Although RON and ROM were similar between cohorts in many of the TBSRTC categories, there was a higher RON associated with non-diagnostic specimens in the TP only cohort when the denominator included all non-diagnostic cases.

CONCLUSION

The combination of CS and LBP may potentially decrease the non-diagnostic rate of T-FNA specimens as well as the number of passes required for diagnosis, particularly with ROSE. Evaluation of morphologic features highlighted in conventional smears may facilitate diagnostic categorization in the "suspicious for follicular neoplasm" category.

Imaging and SERS Study of the Au Nanoparticles Interaction with HPV and Carcinogenic Cervical Tissues

In this work, gold NPs were prepared by the Turkevich method, and their interaction with HPV and cancerous cervical tissues were studied by scanning electron microscopy, energy-dispersive x-ray spectroscopy, confocal and multiphoton microscopy and SERS. The SEM images confirmed the presence and localization of the gold NPs inside of the two kinds of tissues. The light absorption of the gold NPs was at 520 nm. However, it was possible to obtain two-photon imaging (red emission region) of the gold NPs inside of the tissue, exciting the samples at 900 nm, observing the morphology of the tissues. The infrared absorption was probably due to the aggregation of gold NPs inside the tissues. Therefore, through the interaction of gold nanoparticles with the HPV and cancerous cervical tissues, a surface enhanced Raman spectroscopy (SERS) was obtained. As preliminary studies, having an average of 1000 Raman spectra per tissue, SERS signals showed changes between the HPV-infected and the carcinogenic tissues; these spectral signatures occurred mainly in the DNA bands, potentially offering a tool for the rapid screening of cancer.

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.

Computational Cytology: Lessons Learned from Pap Test Computer-Assisted Screening

BACKGROUND

In the face of rapid technological advances in computational cytology including artificial intelligence (AI), optimization of its application to clinical practice would benefit from reflection on the lessons learned from the decades-long journey in the development of computer-assisted Pap test screening.

SUMMARY

The initial driving force for automated screening in cytology was the overwhelming number of Pap tests requiring manual screening, leading to workflow backlogs and incorrect diagnoses. Several companies invested resources to address these concerns utilizing different specimen processing techniques and imaging systems. However, not all companies were commercially prosperous. Successful implementation of this new technology required viable use cases, improved clinical outcomes, and an acceptable means of integration into the daily workflow of cytopathology laboratories. Several factors including supply and demand, Food and Drug Administration (FDA) oversight, reimbursement, overcoming learning curves and workflow changes associated with the adoption of new technology, and cytologist apprehension, played a significant role in either promoting or preventing the widespread adoption of automated screening technologies. Key Messages: Any change in health care, particularly those involving new technology that impacts clinical workflow, is bound to have its successes and failures. However, perseverance through learning curves, optimizing workflow processes, improvements in diagnostic accuracy, and regulatory and financial approval can facilitate widespread adoption of these technologies. Given their history with successfully implementing automated Pap test screening, cytologists are uniquely positioned to not only help with the development of AI technology for other areas of pathology, but also to guide how they are utilized, regulated, and managed.

Prevention of Cervical Cancer: Guideline of the DGGG and the DKG (S3 Level, AWMF Register Number 015/027OL, December 2017) - Part 1 with Introduction, Screening and the Pathology of Cervical Dysplasia

Aims Annual opportunistic screening for cervical carcinoma has been carried out in Germany since 1971. The creation of this S3 guideline meets an important need, outlined in the National Cancer Plan, with regard to screening for cervical cancer, as the guideline aims to provide important information and support for planned organized screening for cervical cancer in Germany. Methods With the financial support of German Cancer Aid, 21 professional societies developed evidence-based statements and recommendations (classified using the GRADE system) for the screening, management and treatment of precancerous conditions of the cervix. Two independent scientific institutes compiled systematic reviews for this guideline. Recommendations The first part of this short summary presents the pathological basis and considers various questions related to screening for cervical cancer. As also reported in earlier reviews, the meta-analysis by Kleijnen Systematic Reviews showed that HPV-based screening offers better protection against invasive cervical cancer compared to cytology-based screening. The authors of this guideline therefore recommend - in accordance with the guideline of the Joint National Committee of Germany (Gemeinsamer Bundesauschuss, G-BA) - that women aged 35 and above should be examined at regular intervals (at least every 3 years) and undergo HPV-based screening. Co-testing can also be carried out. Women between the ages of 20 and 35 should have cytological screening every 2 years.

Updates in Cervical Cytology: The 90-Year-Long Journey from Battle Creek to Today

Ninety years ago, at the Battle Creek conference, Papanicolaou introduced cervical exfoliative cytology. Since then, the "Pap test" has come a long way. The discovery of a causal relationship between cervical carcinoma and HPV infection opened the door for molecular testing and immunomarkers for HPV. The Clinical Laboratory Improvement Amendments, 1988, established quality assurance and quality control programs to monitor performance of cytology laboratories. The Bethesda System for reporting cervical cytology laid the foundations for cervical cytology education, implementation of management guidelines, and further research on cervical carcinogenesis. HPV vaccine penetration in both genders remains 62% or less.

Projected future impact of HPV vaccination and primary HPV screening on cervical cancer rates from 2017-2035: Example from Australia

Background

Many countries are transitioning from cytology-based to longer-interval HPV screening. Trials comparing HPV-based screening to cytology report an increase in CIN2/3 detection at the first screen, and longer-term reductions in CIN3+; however, population level year-to-year transitional impacts are poorly understood. We undertook a comprehensive evaluation of switching to longer-interval primary HPV screening in the context of HPV vaccination. We used Australia as an example setting, since Australia will make this transition in December 2017.

Methods

Using a model of HPV vaccination, transmission, natural history and cervical screening, Policy1-Cervix, we simulated the planned transition from recommending cytology every two years for sexually-active women aged 18–20 to 69, to recommending HPV screening every five years for women aged 25–74 years. We estimated rates of CIN2/3, cervical cancer incidence, and mortality for each year from 2005 to 2035, considering ranges for HPV test accuracy and screening compliance in the context of HPV vaccination (current coverage ~82% in females; ~76% in males).

Findings

Transient increases are predicted to occur in rates of CIN2/3 detection and invasive cervical cancer in the first two to three years following the screening transition (of 16–24% and 11–14% in respectively, compared to 2017 rates). However, by 2035, CIN2/3 and invasive cervical cancer rates are predicted to fall by 40–44% and 42–51%, respectively, compared to 2017 rates. Cervical cancer mortality rates are predicted to remain unchanged until ~2020, then decline by 34–45% by 2035. Over the period 2018–2035, switching to primary HPV screening in Australia is expected to avert 2,006 cases of invasive cervical cancer and save 587 lives.

Conclusions

Transient increases in detected CIN2/3 and invasive cancer, which may be detectable at the population level, are predicted following a change to primary HPV screening. This is due to improved test sensitivity bringing forward diagnoses, resulting in longer term reductions in both cervical cancer incidence and mortality. Fluctuations in health outcomes due to the transition to a longer screening interval are predicted to occur for 10–15 years, but cervical cancer rates will be significantly reduced thereafter due to the impact of HPV vaccination and HPV screening. In order to maintain confidence in primary HPV screening through the transitional phase, it is important to widely communicate that an initial increase in CIN2/3 and perhaps even invasive cervical cancer is expected after a national transition to primary HPV screening, that this phenomenon is due to increased prevalent disease detection, and that this effect represents a marker of screening success.

Reliability of reporting the presence of transformation zone material in Papanicolaou smears using an automated screening system

OBJECTIVE

This study was conducted to evaluate the reliability of using an interactive automated screening system (IAS) to determine the presence of endocervical and/or metaplastic cells (transformation zone material; TZM) in liquid-based cytology (LBC) Papanicolaou (Pap) smear slides.

MATERIALS AND METHODS

A series of LBC Pap smears was manually screened to determine the presence or absence of TZM. The results were compared to those obtained through IAS.

RESULTS

ThinPrep^® Imaging System (Imager)-assisted screening indicated that 22.4% of the slides did not contain TZM in cases in which previous manual screening revealed that 100% of the slides did contain TZM.

CONCLUSION

The reliability of determining TZM through IAS is limited. Manual screening of the slides appears to be necessary to report the presence of TZM accurately.

Time consumed by microscopic and nonmicroscopic tasks in image-assisted gynecologic screening: Implications for workload assessment

BACKGROUND

Gynecologic screening cytology is a complex task that includes microscopic activities and nonmicroscopic activities. The authors sought to determine the amount and percentage of time that cytotechnologists spend on those activities using the ThinPrep imaging system.

METHODS

In arm 1, a total of 550 consecutive unselected slides were reviewed by 11 cytotechnologists, and the time used for individual subtasks of the screening process was recorded. In arm 2, a total of 20 unselected slides were each screened by 10 different cytotechnologists (200 slides in total) and total screening times and full manual review (FMR) times were recorded.

RESULTS

In arm 1, cases with and without FMR required an average of 5.6 minutes and 3.0 minutes, respectively, to screen. Overall, review of fields of view (FOVs) took 95 seconds. FMR took an average of 2.6 minutes. The average screening times for FOV-only cases was significantly longer than the US Food and Drug Administration/Centers for Medicare and Medicaid Services (FDA/CMS) workload limit of 2.4 minutes (P = .005). However, in arm 2, the time needed to screen a case increased by an average of 1 minute compared with arm 1, including 1.1 minute for FOV-only cases and >2 minutes for FMR plus FOV cases. Approximately 100% of cases screened as FOV only exceeded the FDA/CMS workload limit of 2.4 minutes.

CONCLUSIONS

The FDA/CMS workload limits for FOV-only cases appears to significantly underestimate the time needed to screen those cases, but seems to be appropriate for the majority of FMR plus FOV cases. Approximately 60% and 30% of the time designated to screening slides was spent on nonmicroscopic activities for FOV-only cases and FMR cases, respectively. Cancer Cytopathol 2016;124:501-7. © 2016 American Cancer Society.

A comparison of ThinPrep Imager-assisted with manual screening, and its place in the New Zealand cervical cancer screening program

AIMS

Overseas studies have shown equivalent sensitivity and specificity between computer-assisted screening and manual screening, with increased screener productivity. This study was undertaken to test these in a New Zealand laboratory setting.

METHODS

A total of 9232 slides were read manually alone, and following ThinPrep Imager-assisted screening, and the results compared. Two senior screeners and a cytopathologist reviewed the slides with discordant results.

RESULTS

The detection rate for abnormalities was 7.30% for Imager-assisted screening and 7.83% for manual screening. The concordance in diagnosis of abnormalities ranged from 72.7% to 100% with the lowest concordance for high-grade abnormalities diagnosed by Imager-assisted screening. The rate of unsatisfactory smears with Imager-assisted screening is half that of manual screening. There was a screener productivity increase of 140%. In all but one case, abnormal cells were identified by the Imager but screeners varied in their interpretations.

CONCLUSIONS

Overall, Imager-assisted screening was as sensitive as manual screening, and more sensitive for high-grade lesions, with a halving of the rate of unsatisfactory smears. In the setting of the New Zealand cervical screening program, the initial screen by the Imager removes the need for a second, rapid rescreen required by the program for manual screening.

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.

A randomized trial comparing conventional cytology to liquid-based cytology and computer assistance

Liquid-based cytology (LBC) has replaced conventional cytology (CC) for cervical cancer screening in some countries. However, it remains unclear whether LBC is superior to CC. A randomized controlled trial was conducted between August 2007 and March 2009 in Germany to compare LBC, alone and in combination with computer-assisted imaging technology (CAS), to CC in the detection of histologically confirmed cervical intraepithelial neoplasia (CIN). The main outcome measures were detection rates, relative sensitivities, positive predictive values (PPVs) and relative PPVs comparing LBC without and with CAS to CC. Primary histological outcome was CIN2 or higher. Included were 20,627 women participating in opportunistic cervical cancer screening at 20 gynecologic practices. The practices were randomized weekly to use LBC (n = 11,331) or CC (n = 9,296). Patients with positive findings were invited to expert colposcopy. The relative sensitivity of LBC versus CC using the CIN2+ cut-off was 2.74 (95% confidence interval [CI] 1.66-4.53). The relative sensitivity of LBC/CAS versus CC for CIN2+ was 3.17 (95% CI 1.94-5.19). The PPV of LBC and CC for CIN2+ was 48% and 38%, respectively. The PPV ratio did not differ significantly from unity. Differences between LBC and CC were smaller in some sensitivity and subgroup analyses; however, relative sensitivity of LBC remained increased. LBC without and with CAS compared with CC under the field conditions of an opportunistic screening system had a significantly higher sensitivity for the detection of CIN without deterioration of PPVs. Additional use of CAS did not further improve sensitivity of LBC. © 2012 Wiley Periodicals, Inc.

Prospective parallel randomized trial of the MultiCyte™ ThinPrep(®) imaging system: the Scottish experience

BACKGROUND

Computer-assisted screening of cervical liquid-based cytology (LBC) preparations using the ThinPrep® Imaging System (TIS) has shown improved qualitative and quantitative gains. The use of Multicyte™ has not been described in a well-established national screening programme with a low incidence of high-grade dyskaryosis.

OBJECTIVES

To assess the impact of computer-assisted screening within the Scottish Cervical Screening Programme (SCSP).

METHODS

Two groups of three laboratories, each sharing a ThinPrep® Imager, screened 79 366 slides randomized to test and 90 551 to control arms by laboratory accession. Screeners were not blinded. Standard laboratory reporting profiles of the SCSP, sensitivity, specificity and false-negative rates of all grades of LBC abnormalities with respect to final cytology reports, predictive value for cervical intraepithelial neoplasia grade 2 or worse (CIN2+) on histology; and screening rates were compared for both arms.

RESULTS

Inadequate and negative reporting rates were significantly lower and low-grade reporting rates significantly higher in the imager arm. Imager-assisted screening showed significantly better specificity than manual screening with respect to the final cytology result. There was no evidence of a significant difference in the detection of CIN2+ or CIN3 +. Positive, abnormal and total predictive values (high-grade, low-grade and all abnormal cytology found to be CIN2+, respectively) were similar in both arms. Productivity was significantly higher in the imager arm.

CONCLUSION

  Computer-assisted screening in a well established screening programme showed significantly improved productivity without loss of quality. These findings should inform future policy for cervical screening programmes.

Comparison of the sensitivity of conventional cytology and the ThinPrep Imaging System for 1,083 biopsy confirmed high-grade squamous lesions

Liquid-based cytology continues to be utilized as an adjunct to conventional cytology in most Australian laboratories, even though a direct-to-vial ThinPrep protocol has been introduced in many countries with established cervical screening programs. Manual screening of ThinPrep slides has been widely practiced for more than 10 years and the recent introduction of the ThinPrep Imaging System (TPI) has been reported as being more sensitive than the conventional smear (CS) in the identification of high-grade cervical disease.We report our experience with ThinPrep Imaging since its introduction into our routine gynecological cytology service. 87,284 split sample pairs reported using the Imaging System demonstrated a decrease in unsatisfactory reports (3.65% for CS and 0.87% for TPI) and an increase in possible high grade and definite high-grade squamous reports (1.57% for CS and 1.62% for TPI).For 1,083 biopsy confirmed high-grade lesions, the correct diagnosis of high grade or possible high-grade squamous disease was made on the ThinPrep imaged slide in 61.0% (661/1,083) of cases and on the CS in 59.4% (643/1,083). This was not statistically significant. When all abnormalities identified on cytology were considered, including possible low grade and definite low-grade abnormalities, the difference in sensitivity for Thinprep imaged slides of 96.0% (1,040/1,083) and CSs of 91.6% (992/1,083) was statistically significant.

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

Background:

Published reports have demonstrated that introduction of the ThinPrep Imaging System (Imager) to the cytology screening services has increased the detection rate of high-grade squamous intraepithelial lesions (HSILs). In accordance with recent clinical treatment guidelines, patients with atypical squamous or glandular cells of undetermined significance (ASC-US or AGUS) are often tested for high-risk HPV infection using the Hybrid Capture HPV DNA test. We took the opportunity to investigate whether the Imager had resulted in any significant differences in our diagnostic categories, as well as whether the Imager increased the detection of high-risk HPV-DNA-positive (HRHPV+) ASC-US or AGUS.

Materials and Methods:

Cytology cases with the diagnosis of ASC-US and AGUS were retrieved from the archival files of our institution during periods of 11 months prior to and 11 months after the introduction of the Imager. The total number of cases in each category was correlated with results of reflex high-risk HPV DNA testing when the latter were available. All AGUS diagnoses were correlated with subsequent biopsy follow-up. Statistical analyses were performed using the chi-Square test with Yate's Correction and Fisher's Exact test.

Results:

A total of 108,371 and 104,555 of ThinPrep® Pap Test (TPPT) cases were reviewed during 11 months pre- and post-imager introduction. The ASC-US rate was 5.4% in the pre-Imager and 5.3% in the post-Imager period. The HPV reflex test was 38% and 34% positive respectively in the pre- and post-Imager period (P>0.124). Similarly, 0.14% and 0.12% AGUS were found in the pre- and post-Imager period. The positive HPV reflex test was 14% versus 23% (P = 0.1690). The abnormal biopsy follow-up rate in the AGUS category was increased from 20.9% in the pre-Imager period to 31% in the post-Imager period (P = 0.1471). The ASCUS/SIL ratios were 1.9 and 1.6 respectively.

Conclusions:

The ASC-US and AGUS rates did not change statistically before and after the introduction of the Imager in our cytology laboratory. Although use of the Imager did not increase detection of HPV+ ASC-US, it did appear to increase the detection rate of HPV+ AGUS and subsequent abnormal biopsy follow-up rates in all categories. However, the increase in the detection rate did not reach the point of statistical significance

Automated quantitative analysis of estrogen receptor expression in breast carcinoma does not differ from expert pathologist scoring: a tissue microarray study of 3,484 cases

BACKGROUND

Estrogen receptor (ER) expression is routinely assessed by immunohistochemistry (IHC) in breast carcinoma. Our study compares visual scoring of ER in invasive breast cancer by histopathologists to quantitation of staining using a fully automated system.

MATERIALS AND METHODS

A tissue microarray was constructed from 4,049 cases (3,484 included in analysis) of invasive breast carcinoma linked to treatment and outcome information. Slides were scored independently by two pathologists and scores were dichotomised, with ER positivity recognized at a cut-off of >1% positive nuclei. The slides were scanned and analyzed with an Ariol automated system.

RESULTS

Using data dichotomised as ER positive or negative, both visual and automated scores were highly consistent: there was excellent concordance between two pathologists (kappa = 0.918 (95%CI: 0.903-0.932)) and between two Ariol machines (kappa = 0.913 (95%CI: 0.897-0.928)). The prognostic significance of ER positivity was similar whether determined by pathologist or automated scoring for both the entire patient cohort and subsets of patients treated with tamoxifen alone or receiving no systemic adjuvant therapy. The optimal cut point for the automated scores using breast cancer disease-specific survival as an endpoint was >0.4% positive nuclei. The concordance between dextran-coated charcoal ER biochemical assay data and automated scores (kappa = 0.728 (95%CI: 0.69-0.75); 0.74 (95%CI: 0.71-0.77)) was similar to the concordance between biochemical assay and pathologist scores (kappa = 0.72 (95%CI: 0.70-0.75; 0.70 (95%CI: 0.67-0.72)).

CONCLUSION

Fully automated quantitation of ER immunostaining yields results that do not differ from human scoring against both biochemical assay and patient outcome gold standards.

Vibrational spectroscopy for cervical cancer pathology, from biochemical analysis to diagnostic tool

Cervical cancer is the second most common cancer in women worldwide with 80% of cases arising in the developing world. The mortality associated with cervical cancer can be reduced if this disease is detected at the early stages of development or at the pre-malignant state (cervical intraepithelial neoplasia, CIN). The aim of this study was to investigate the potential of Raman spectroscopy as a diagnostic tool to detect biochemical changes accompanying cervical cancer progression. Raman spectra were acquired from proteins, nucleic acids, lipids and carbohydrates in order to gain an insight into the biochemical composition of cells and tissues. Spectra were also obtained from histological samples of normal, CIN and invasive carcinoma tissue from 40 patients. Multivariate analysis of the spectra was carried out to develop a classification model to discriminate normal from abnormal tissue. The results show that Raman spectroscopy displays a high sensitivity to biochemical changes in tissue during disease progression resulting in an exceptional prediction accuracy when discriminating between normal cervical tissue, invasive carcinoma and cervical intraepithelial neoplasia (CIN). Raman spectroscopy shows enormous clinical potential as a rapid non-invasive diagnostic tool for cervical and other cancers.