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

Validation of AI-assisted ThinPrep® Pap test screening using the GeniusTM Digital Diagnostics System

Advances in whole-slide imaging and artificial intelligence present opportunities for improvement in Pap test screening. To date, there have been limited studies published regarding how best to validate newer AI-based digital systems for screening Pap tests in clinical practice. In this study, we validated the Genius™ Digital Diagnostics System (Hologic) by comparing the performance to traditional manual light microscopic diagnosis of ThinPrep® Pap test slides. A total of 319 ThinPrep® Pap test cases were prospectively assessed by six cytologists and three cytopathologists by light microscopy and digital evaluation and the results compared to the original ground truth Pap test diagnosis. Concordance with the original diagnosis was significantly different by digital and manual light microscopy review when comparing across: (i) exact Bethesda System diagnostic categories (62.1% vs 55.8%, respectively, p = 0.014), (ii) condensed diagnostic categories (76.8% vs 71.5%, respectively, p = 0.027), and (iii) condensed diagnoses based on clinical management (71.5% vs 65.2%, respectively, p = 0.017). Time to evaluate cases was shorter for digital (M = 3.2 min, SD = 2.2) compared to manual (M = 5.9 min, SD = 3.1) review (t(352) = 19.44, p < 0.001, Cohen's d = 1.035, 95% CI [0.905, 1.164]). Not only did our validation study demonstrate that AI-based digital Pap test evaluation had improved diagnostic accuracy and reduced screening time compared to light microscopy, but that participants reported a positive experience using this system.

The Current and Future Impact of Lymph Node Fine-Needle Aspiration Cytology on Patient Care

Lymph node (LN) fine-needle aspiration cytology (FNAC) is a common diagnostic procedure for lymphadenopathies. Despite the qualities and potentialities of LN-FNAC, the number of possible pathologies and the variety of clinical contexts represent a challenge and require a continuous upgrading of the procedure according to the emerging clinical requests and new technologies. This study presents an overview of the current and future impact of LN-FNAC on the care of patients with lymphadenopathy.

Evolution of Pathological Techniques for the Screening of Cervical Cancer: A Comprehensive Review

The evolutionary journey of cervical cancer screening has been a major medical success story, considering the substantial role it has played in dwindling the disease burden. Through sustained collaborative efforts within the medical community, significant advances have been made from the humble yet path-breaking conventional Pap smear to the current automated screening systems and human papillomavirus (HPV) molecular testing. With the integration of artificial intelligence into screening techniques, we are currently at the precipice of circumventing the pitfalls of manual cytology readings and improving the efficiency of the screening systems by a significant margin. Despite the technological milestones traversed, the high logistics and operational cost, besides the technical know-how of operating the automated systems, can pose a major practical challenge in the widespread adoption of these advanced techniques in cervical cancer screening programs. This would suggest the need to adopt strategies that are tailored to the demands and needs of the different settings keeping their limitations in mind. This review aims to take the reader through the entire evolutionary journey of cervical cancer screening programs, highlight the individual merits and demerits of each technique, and discuss the recommendations from the major global guidelines.

Pros and cons of artificial intelligence implementation in diagnostic pathology

The rapid introduction of digital pathology has greatly facilitated development of artificial intelligence (AI) models in pathology that have shown great promise in assisting morphological diagnostics and quantitation of therapeutic targets. We are now at a tipping point where companies have started to bring algorithms to the market, and questions arise whether the pathology community is ready to implement AI in routine workflow. However, concerns also arise about the use of AI in pathology. This article reviews the pros and cons of introducing AI in diagnostic pathology.

Evaluating Urine Cytology Slide Digitization Efficiency: A Comparative Study Using an Artificial Intelligence-Based Heuristic Scanning Simulation and Multiple Z-Plane Scanning

INTRODUCTION

Digitizing cytology slides presents challenges because of their three-dimensional features and uneven cell distribution. While multi-Z-plane scan is a prevalent solution, its adoption in clinical digital cytopathology is hindered by prolonged scanning times, increased image file sizes, and the requirement for cytopathologists to review multiple Z-plane images.

METHODS

This study presents heuristic scan as a novel solution, using an artificial intelligence (AI)-based approach specifically designed for cytology slide scanning as an alternative to the multi-Z-plane scan. Both the 21 Z-plane scan and the heuristic scan simulation methods were used on 52 urine cytology slides from three distinct cytopreparations (Cytospin, ThinPrep, and BD CytoRich™ [SurePath]), generating whole-slide images (WSIs) via the Leica Aperio AT2 digital scanner. The AI algorithm inferred the WSI from 21 Z-planes to quantitate the total number of suspicious for high-grade urothelial carcinoma or more severe cells (SHGUC+) cells. The heuristic scan simulation calculated the total number of SHGUC+ cells from the 21 Z-plane scan data. Performance metrics including SHGUC+ cell coverage rates (calculated by dividing the number of SHGUC+ cells identified in multiple Z-planes or heuristic scan simulation by the total SHGUC+ cells in the 21 Z-planes for each WSI), scanning time, and file size were analyzed to compare the performance of each scanning method. The heuristic scan's metrics were linearly estimated from the 21 Z-plane scan data. Additionally, AI-aided interpretations of WSIs with scant SHGUC+ cells followed The Paris System guidelines and were compared with original diagnoses.

RESULTS

The heuristic scan achieved median SHGUC+ cell coverage rates similar to 5 Z-plane scans across three cytopreparations (0.78-0.91 vs. 0.75-0.88, p = 0.451-0.578). Notably, it substantially reduced both scanning time (137.2-635.0 s vs. 332.6-1,278.8 s, p < 0.05) and image file size (0.51-2.10 GB vs. 1.16-3.10 GB, p < 0.05). Importantly, the heuristic scan yielded higher rates of accurate AI-aided interpretations compared to the single Z-plane scan (62.5% vs. 37.5%).

CONCLUSION

We demonstrated that the heuristic scan offers a cost-effective alternative to the conventional multi-Z-plane scan in digital cytopathology. It achieves comparable SHGUC+ cell capture rates while reducing both scanning time and image file size, promising to aid digital urine cytology interpretations with a higher accuracy rate compared to the conventional single (optimal) plane scan. Further studies are needed to assess the integration of this new technology into compatible digital scanners for practical cytology slide scanning.

The evolution of cervical cancer screening

There are few medical success stories in history as significant as the reduction in cervical cancer incidence. Through the collaborative efforts of dedicated scientific pioneers, the past century has witnessed remarkable advancement that began with the detection of exfoliated cancer cells through cytologic examination to widespread implementation of cervical cancer screening programs to the discovery of the link between cervical cancer and human papillomavirus (HPV). Current screening methods apply HPV-based testing, and artificial intelligence-based screening systems utilizing digitalized cytology images are being used in a continuous effort to optimize the accuracy and efficiency of the Papanicolaou test. This review summarizes the major milestones in cervical cancer screening history to emphasize its evolution as the World Health Organization aims for the global elimination of cervical cancer.

Validation of full-remote reporting for cervicovaginal cytology: the Caltagirone-Acireale distributed lab

INTRODUCTION

Cervical cancer is the fourth most common cancer in women, and its prevention is based on vaccination and screening. Screening consists of molecular human papillomavirus (HPV) testing and cytologic analysis of cervical smears, which require expensive equipment and the interaction of numerous professionals such as biologists, cytologists, laboratory technicians, and pathologists.

MATERIALS AND METHODS

We centralize the cervical samples from more than 51 clinics in 1 main laboratory, where automated HPV testing is performed. HPV-positive cases are collected and used to prepare a liquid-based cytology slide, which is stained and immediately scanned. The resulting whole-slide images (WSIs) are immediately available in a remote laboratory where they are examined by experienced cytologists using virtual microscopy. This setup was validated by making each of the 3 readers independently diagnose 506 specimens in random order, using both conventional light microscopy (CLM) and WSIs, with a minimum wash-out period of 3 weeks and with a final discussion for all cases.

RESULTS

Intraobserver agreement among CLM and WSI ranged from 0.71 to 0.79, and interobserver agreement for the 3 readers compared with the consensus diagnosis was similar for the 2 modes of assessment. Readers subjectively felt confident in their WSI diagnosis for inadequate and negative cases, but less so in other cases. The perceived difficulty was slightly higher in WSI readings.

CONCLUSIONS

Interobserver agreement in cervicovaginal cytology is moderate and does not vary if the slides are examined conventionally or digitally. Despite higher reported subjective difficulty and lower confidence in the WSI diagnosis, we did not observe a deterioration in diagnostic performance using WSI compared with CLM.

Relationship between a deep learning model and liquid-based cytological processing techniques

OBJECTIVE

Artificial intelligence (AI)-based cytopathology studies conducted using deep learning have enabled cell detection and classification. Liquid-based cytology (LBC) has facilitated the standardisation of specimen preparation; however, cytomorphology varies according to the LBC processing technique used. In this study, we elucidated the relationship between two LBC techniques and cell detection and classification using a deep learning model.

METHODS

Cytological specimens were prepared using the ThinPrep and SurePath methods. The accuracy of cell detection and cell classification was examined using the one- and five-cell models, which were trained with one and five cell types, respectively.

RESULTS

When the same LBC processing techniques were used for the training and detection preparations, the cell detection and classification rates were high. The model trained on ThinPrep preparations was more accurate than that trained on SurePath. When the preparation types used for training and detection were different, the accuracy of cell detection and classification was significantly reduced (P < 0.01). The model trained on both ThinPrep and SurePath preparations exhibited slightly reduced cell detection and classification rates but was highly accurate.

CONCLUSIONS

For the two LBC processing techniques, cytomorphology varied according to cell type; this difference affects the accuracy of cell detection and classification by deep learning. Therefore, for highly accurate cell detection and classification using AI, the same processing technique must be used for both training and detection. Our assessment also suggests that a deep learning model should be constructed using specimens prepared via a variety of processing techniques to construct a globally applicable AI model.

"SMART" cytology: The next generation cytology for precision diagnosis

Cytology plays an important role in diagnosing and managing human diseases, especially cancer, as it is often a simple, low cost yet effective, and non-invasive or minimally invasive diagnostic tool. However, traditional morphology-based cytology practice has limitations, especially in the era of precision diagnosis. Recently there have been tremendous efforts devoted to apply computational tools and to perform molecular analysis on cytological samples for a variety of clinical purposes. Now is probably the appropriate juncture to integrate morphology, machine learning, and molecular analysis together and transform cytology from a morphology-driven practice to the next level - "SMART" Cytology. In this article we will provide a rather brief review of the relevant works for computational analysis on cytology samples, focusing on single-cell-based multiplex quantitative analysis of biomarkers, and introduce the conceptual framework of "SMART (Single cell, Multiplex, AI-driven, and Real Time)" Cytology.

Cervical cell classification with deep-learning algorithms

Cervical cancer is a serious threat to the lives and health of women. The accurate analysis of cervical cell smear images is an important diagnostic basis for cancer identification. However, pathological data are often complex and difficult to analyze accurately because pathology images contain a wide variety of cells. To improve the recognition accuracy of cervical cell smear images, we propose a novel deep-learning model based on the improved Faster R-CNN, shallow feature enhancement networks, and generative adversarial networks. First, we used a global average pooling layer to enhance the robustness of the data feature transformation. Second, we designed a shallow feature enhancement network to improve the localization and recognition of weak cells. Finally, we established a data augmentation network to improve the detection capability of the model. The experimental results demonstrate that our proposed methods are superior to CenterNet, YOLOv5, and Faster R-CNN algorithms in some aspects, such as shorter time consumption, higher recognition precision, and stronger adaptive ability. Its maximum accuracy is 99.81%, and the overall mean average precision is 89.4% for the SIPaKMeD and Herlev datasets. Our method provides a useful reference for cervical cell smear image analysis. The missed diagnosis rate and false diagnosis rate are relatively high for cervical cell smear images of different pathologies and stages. Therefore, our algorithms need to be further improved to achieve a better balance. We will use a hyperspectral microscope to obtain more spectral data of cervical cells and input them into deep-learning models for data processing and classification research. First, we sent training samples of cervical cells into our proposed deep-learning model. Then, we used the proposed model to train eight types of cervical cells. Finally, we utilized the trained classifier to test the untrained samples and obtained the classification results. Fig 1. Deep-learning cervical cell classification framework.

BestCyte® primary screening of 500 ThinPrep Pap Test thin-layers: 3 Cytologists' Interobserver diagnostic concordance with predicate manual microscopy relative to Truth Reference diagnoses defining NILM, ASCUS+, LSIL+, and ASCH+ thresholds for specificity, sensitivity, and equivalency grading

Background

The BestCyte® Cell Sorter Imaging System (BestCyte) facilitates algorithmic discrimination of clinically relevant cells in Pap test cytopathology by classifying and projecting images of cells in galleries based on cytomorphology. Warranted is awareness of potential BestCyte advantages as measured through 3 cytologists' interobserver diagnostic concordance, specificity and sensitivity differentials, and equivalency grading relative to manual microscopy (MM).

Objectives

Using 500 MM-reported ThinPrep thin-layers, analyze: (1) cytologists' blinded BestCyte screening to raise Bethesda diagnoses; (2) correlate BestCyte and MM diagnoses (i.e., predicate) to establish Truth Reference Diagnoses (TRDx) from concordance between 4 possible diagnoses; (3) analyze cytologists' and MM predicate diagnoses through 4 diagnostic thresholds defined by TRDx: NILM (Negative) for specificity, and ASCUS+, LSIL+, and ASCH+ (Positive) for graded sensitivity (with abnormal cells decreasing in size with increasing dysplasia); and, (4) statistically determine cytologists' equivalency grading to MM using 95% Confidence Interval (CI) ranges.

Results

500 TRDx breakdown (n/%): NILM (241/48.2), ASCUS (79/15.8), ASCH (9/1.80), AGUS (2/0.40), LSIL (86/17.2), HSIL (68/13.6), CA (2/0.40), UNSAT (13/2.60). TRDx breakdown (n/%) per 4 of 4, 3 of 4, 2 of 4 diagnostic concordances: 264 (52.8%), 182 (36.4%), 54 (10.8%), respectively. No cases of discordant diagnoses were recorded. HSIL TRDx were established from 66.2% of 4 of 4 concordances, followed by NILM (59.3%), LSIL (46.5%), ASCUS (41.8%); antithetically, from 4.40% of 2 of 4 concordances. Specificity for MM predicate (NILM): 67.08%; for Cytologists 1, 2, and 3: 89.71%, 82.30%, 97.53%, respectively. For NILM threshold, cytologists revealed Significantly Superior equivalency to MM. Sensitivity for ASCUS+, LSIL+, and ASCH+ thresholds: MM (91.36%, 86.67%, 74.36%); Cytologist 1 (95.88%, 96.97%, 94.87%); Cytologist 2 (95.47%, 95.76%, 93.59%), Cytologist 3 (94.65%, 95.15%, 98.72%), respectively. Cytologists revealed Significantly Superior equivalency to MM for graded Positive thresholds; with Cytologist 3 for ASCUS+ being: Superior.

Conclusions

BestCyte detects and efficiently displays abnormal cells in strategic galleries standardizing objectivity by systematizing mosaics of cell-types for cytologists' consideration. BestCyte fosters consistent, enhanced cytologists' sensitivity values for the ASCUS+, LSIL+, and ASCH+ Positive thresholds relative to MM. Also, BestCyte facilitates improved specificity and superior equivalency grading to MM reflecting efficient screening, and reduced labor. Confident interpretations of small dysplastic epithelial cells characteristic of HSIL led to exceptional interobserver diagnostic concordance inferring BestCyte is primed for effective cervical cancer screening practice.

BestCyte® Cell Sorter Imaging System: Primary and Adjudicative Whole Slide Image Rescreening Review Times of 500 ThinPrep Pap Test Thin-layers - An Intra-observer, Time-Surrogate Analysis of Diagnostic Confidence Potentialities

Background

The novel Artificial Intelligence-driven BestCyte® Cell Sorter Imaging System (BestCyte) enables hybrid digital screening through classification and sorting of tiles depicting cells in 8 galleries or whole slide image (WSI) reviews.

Objectives

(1) Analyze expenditures of time (minutes) for primary BestCyte cell sorter screening and adjudicative WSI rescreening of 500 blinded, randomized ThinPrep thin-layers to determine review times per Bethesda nomenclature; (2) Analyze review times for NILM qualifier diagnoses reflecting increasing interpretive complexity (i.e., Inflammation, Reactive/Repair, Bacterial cytolysis, Bacterial vaginosis, Atrophy, and Atrophic vaginitis); (3) Challenge accuracy of primary diagnoses (Downgraded, Upheld, and Upgraded) following adjudicative WSI rescreening to assess correlated review times as surrogate indicators of diagnostic confidence in BestCyte functionality (i.e., learning curve); and (4) Correlate primary and adjudicative diagnoses to calculate intra-observer reproducibility Kappa coefficients per Bethesda nomenclature.

Results

Of 500 thin-layers, the mean [primary/adjudicative rescreening review times (minutes)] were: Overall study [1.38/3.94], NILM [1.23/3.02], ASCUS [1.18/2.53], ASC-H [1.73/4.86], AGUS [1.84/6.34], LSIL [1.49/4.16], HSIL [1.52/4.10], CA [0.65/2.57]. Of 500 primary Bethesda diagnoses: 2 (0.40%) downgraded; 483 (96.6%) upheld; 15 (3.00%) upgraded after adjudicative WSI rescreening. Of 354 NILM diagnoses: 0 downgraded; 344 (97.2%) upheld; 10 (2.82%) upgraded. Of 34 ASCUS diagnoses: 2 (5.88%) downgraded; 28 (82.4%) upheld; 4 (11.8%) upgraded. Of 17 ASC-H diagnoses: 0 downgraded; 16 (94.1%) upheld; 1 (5.88%) upgraded. Of AGUS (n=1), LSIL (n=24), HSIL (n=52), CA (n=1), UNSAT (n=17): 100% upheld. Kappa coefficients with 95% (Confidence Intervals): Overall study 0.9305 (0.8983–0.9627), NILM 0.9429 (0.9110–0.9748), ASCUS 0.8378 (0.7393–0.9363), ASC-H 0.9112 (0.8113–0.9999), AGUS 1.0 (1.0–1.0), LSIL 0.9189 (0.8400–0.9978), HSIL 0.9894 (0.9685–0.9999), CA 1.0 (1.0–1.0), UNSAT 1.0 (1.0–1.0). Primary BestCyte cell image review time trends for NILM, ASCUS, LSIL, and HSIL, revealed plateaus relative to decreasing respective adjudicative WSI rescreening times.

Conclusions

Given innovative robustness, BestCyte accommodates interpretive fundamentals, enabling shorter ThinPrep thin-layer review times with optimal intra-observer concordance per Bethesda nomenclature through classifying, ranking, sorting, and displaying clinically relevant cells efficiently in galleries. BestCyte fosters continuously optimizing diagnostic confidence learning curves; may supplant manual microscopy for primary screening.

Advances in Imaging Modalities, Artificial Intelligence, and Single Cell Biomarker Analysis, and Their Applications in Cytopathology

Several advances in recent decades in digital imaging, artificial intelligence, and multiplex modalities have improved our ability to automatically analyze and interpret imaging data. Imaging technologies such as optical coherence tomography, optical projection tomography, and quantitative phase microscopy allow analysis of tissues and cells in 3-dimensions and with subcellular granularity. Improvements in computer vision and machine learning have made algorithms more successful in automatically identifying important features to diagnose disease. Many new automated multiplex modalities such as antibody barcoding with cleavable DNA (ABCD), single cell analysis for tumor phenotyping (SCANT), fast analytical screening technique fine needle aspiration (FAST-FNA), and portable fluorescence-based image cytometry analyzer (CytoPAN) are under investigation. These have shown great promise in their ability to automatically analyze several biomarkers concurrently with high sensitivity, even in paucicellular samples, lending themselves well as tools in FNA. Not yet widely adopted for clinical use, many have successfully been applied to human samples. Once clinically validated, some of these technologies are poised to change the routine practice of cytopathology.