The current state of digital cytology and artificial intelligence (AI): global survey results from the American Society of Cytopathology Digital Cytology Task Force.

INTRODUCTION: The integration of whole slide imaging (WSI) and artificial intelligence (AI) with digital cytology has been growing gradually. Therefore, there is a need to evaluate the current state of digital cytology. This study aimed to determine the current landscape of digital cytology via a survey conducted as part of the American Society of Cytopathology (ASC) Digital Cytology White Paper Task Force. MATERIALS AND METHODS: A survey with 43 questions pertaining to the current practices and experiences of WSI and AI in both surgical pathology and cytology was created. The survey was sent to members of the ASC, the International Academy of Cytology (IAC), and the Papanicolaou Society of Cytopathology (PSC). Responses were recorded and analyzed. RESULTS: In total, 327 individuals participated in the survey, spanning a diverse array of practice settings, roles, and experiences around the globe. The majority of responses indicated there was routine scanning of surgical pathology slides (n = 134; 61%) with fewer respondents scanning cytology slides (n = 150; 46%). The primary challenge for surgical WSI is the need for faster scanning and cost minimization, whereas image quality is the top issue for cytology WSI. AI tools are not widely utilized, with only 16% of participants using AI for surgical pathology samples and 13% for cytology practice. CONCLUSIONS: Utilization of digital pathology is limited in cytology laboratories as compared to surgical pathology. However, as more laboratories are willing to implement digital cytology in the near future, the establishment of practical clinical guidelines is needed.

Discovery and technical validation of high-performance methylated DNA markers for the detection of cervical lesions at risk of malignant progression in low- and middle-income countries.

BACKGROUND: Cervical cancer remains a leading cause of death, particularly in developing countries. WHO screening guidelines recommend human papilloma virus (HPV) detection as a means to identify women at risk of developing cervical cancer. While HPV testing identifies those at risk, it does not specifically distinguish individuals with neoplasia. We investigated whether a quantitative molecular test that measures methylated DNA markers could identify high-risk lesions in the cervix with accuracy. RESULTS: Marker discovery was performed in TCGA-CESC Infinium Methylation 450 K Array database and verified in three other public datasets. The panel was technically validated using Quantitative Multiplex-Methylation-Specific PCR in tissue sections (N = 252) and cervical smears (N = 244) from the USA, South Africa, and Vietnam. The gene panel consisted of FMN2, EDNRB, ZNF671, TBXT, and MOS. Cervical tissue samples from all three countries showed highly significant differential methylation in squamous cell carcinoma (SCC) with a sensitivity of 100% [95% CI 74.12-100.00], and specificity of 91% [95% CI 62.26-99.53] to 96% [95% CI 79.01-99.78], and receiver operating characteristic area under the curve (ROC AUC) = 1.000 [95% CI 1.00-1.00] compared to benign cervical tissue, and cervical intraepithelial neoplasia 2/3 with sensitivity of 55% [95% CI 37.77-70.84] to 89% [95% CI 67.20-98.03], specificity of 93% [95% CI 84.07-97.38] to 96% [95% CI 79.01-99.78], and a ROC AUC ranging from 0.793 [95% CI 0.68-0.89] to 0.99 [95% CI 0.97-1.00] compared to CIN1. In cervical smears, the marker panel detected SCC with a sensitivity of 87% [95% CI 77.45-92.69], specificity 95% [95% CI 88.64-98.18], and ROC AUC = 0.925 [95% CI 0.878-0.974] compared to normal, and high-grade squamous intraepithelial lesion (HSIL) at a sensitivity of 70% (95% CI 58.11-80.44), specificity of 94% (95% CI 88.30-97.40), and ROC AUC = 0.884 (95% CI 0.822-0.945) compared to low-grade intraepithelial lesion (LSIL)/normal in an analysis of pooled data from the three countries. Similar to HPV-positive, HPV-negative cervical carcinomas were frequently hypermethylated for these markers. CONCLUSIONS: This 5-marker panel detected SCC and HSIL in cervical smears with a high level of sensitivity and specificity. Molecular tests with the ability to rapidly detect high-risk HSIL will lead to timely treatment for those in need and prevent unnecessary procedures in women with low-risk lesions throughout the world. Validation of these markers in prospectively collected cervical smear cells followed by the development of a hypermethylated marker-based cervical cancer detection test is warranted.

Comparison of non-invasive diagnostic modalities for ocular surface squamous neoplasia at a tertiary hospital, South Africa.

AIMS: The aim of the study is to assess non-invasive diagnostic modalities for ocular surface squamous neoplasia (OSSN) when compared to histology. METHODS: A prospective case-control study was conducted of patients presenting with conjunctival masses at a tertiary eye hospital in Johannesburg, South Africa. Patients completed an interview and had three non-invasive diagnostic tests: optical coherence tomography, impression cytology and methylene blue stain. A biopsy with histology was performed as the gold standard to confirm the diagnosis. RESULTS: One hundred and eighty-two conjunctival masses of 175 patients were evaluated. There were 135 lesions identified as OSSN on biopsy and 47 lesions were benign on histology. Optical coherence tomography had a sensitivity and specificity of 87.2% (95% CI: 80.0-92.5) and 75.6% (95% CI: 60.5-87.1), respectively, when an epithelial thickness cutoff of 140 um was used. Shadowing was found in 46% of cases due to leukoplakia or increased thickness of the mass. Cytology had a sensitivity of 72.4% (95% CI: 62.5-81.0) and a specificity of 74.3% (95% CI: 56.7-87.5). Twenty-seven per cent of cytology specimens were excluded from analysis due to inadequate cellularity. Methylene blue had a high sensitivity of 91.9% (95% CI: 85.9-95.9), but low specificity of 55.3% (95% CI: 40.1-69.8). CONCLUSION: Optical coherence tomography had a high sensitivity and specificity as a non-invasive test and liquid-based cytology performed well but had a lower sensitivity and specificity than with optical coherence tomography. Methylene blue performed well as a screening test, with a high sensitivity but low specificity.

Digital cytology part 1: digital cytology implementation for practice: a concept paper with review and recommendations from the American Society of Cytopathology Digital Cytology Task Force.

Digital cytology and artificial intelligence (AI) are gaining greater adoption in the cytopathology laboratory. However, peer-reviewed real-world data and literature are lacking regarding the current clinical landscape. The American Society of Cytopathology in conjunction with the International Academy of Cytology and the Digital Pathology Association established a special task force comprising 20 members with expertise and/or interest in digital cytology. The aim of the group was to investigate the feasibility of incorporating digital cytology, specifically cytology whole slide scanning and AI applications, into the workflow of the laboratory. In turn, the impact on cytopathologists, cytologists (cytotechnologists), and cytology departments were also assessed. The task force reviewed existing literature on digital cytology, conducted a worldwide survey, and held a virtual roundtable discussion on digital cytology and AI with multiple industry corporate representatives. This white paper, presented in 2 parts, summarizes the current state of digital cytology and AI practice in global cytology practice. Part 1 of the white paper presented herein is a review and offers best practice recommendations for incorporating digital cytology into practice. Part 2 of the white paper provides a comprehensive review of AI in cytology practice along with best practice recommendations and legal considerations. Additionally, the results of a global survey regarding digital cytology are highlighted.

Digital cytology part 2: artificial intelligence in cytology: a concept paper with review and recommendations from the American Society of Cytopathology Digital Cytology Task Force.

Digital cytology and artificial intelligence (AI) are gaining greater adoption in the cytology laboratory. However, peer-reviewed real-world data and literature are lacking in regard to the current clinical landscape. The American Society of Cytopathology in conjunction with the International Academy of Cytology and the Digital Pathology Association established a special task force comprising 20 members with expertise and/or interest in digital cytology. The aim of the group was to investigate the feasibility of incorporating digital cytology, specifically cytology whole slide scanning and AI applications, into the workflow of the laboratory. In turn, the impact on cytopathologists, cytologists (cytotechnologists), and cytology departments were also assessed. The task force reviewed existing literature on digital cytology, conducted a worldwide survey, and held a virtual roundtable discussion on digital cytology and AI with multiple industry corporate representatives. This white paper, presented in 2 parts, summarizes the current state of digital cytology and AI practice in global cytology practice. Part 1 of the white paper is presented as a separate paper which details a review and best practice recommendations for incorporating digital cytology into practice. Part 2 of the white paper presented here provides a comprehensive review of AI in cytology practice along with best practice recommendations and legal considerations. Additionally, the cytology global survey results highlighting current AI practices by various laboratories, as well as current attitudes, are reported.

Validation of the Xpert Breast Cancer STRAT 4 Assay on the GeneXpert instrument to Assess Hormone Receptor, Ki67, and HER2 Gene Expression Status in Breast Cancer Tissue Samples.

Breast cancer is the commonest cause of cancer-related mortality in African females where patients often present later and with advanced disease. Causes for delayed diagnosis include restricted diagnostic access and international controversy on interpretation of ancillary tests like immunohistochemistry (IHC). Fine needle aspirates (FNAC) are an attractive alternative although may have reduced sensitivity. The Xpert Breast Cancer STRAT4 (STRAT4) (CE-IVD*) assay (Cepheid, Sunnyvale) is a semi-quantitative reverse-transcription polymerase chain reaction assay which detects messenger RNA (mRNA) expression in breast samples for estrogen receptor ( ESR1 ), progesterone receptor ( PGR1 ), human epidermal growth factor receptor/Erb-B2 receptor tyrosine kinase 2 (HER2/ ERBB2 ) and the proliferation marker, MKi67 . We assessed the performance of this assay on both formalin-fixed paraffin-embedded (FFPE, n=31) and matched FNAC (n=20) samples from patients presenting with breast cancer to the Johannesburg academic hospitals. IHC and Fluorescent in situ hybridization analysis (performed on HER2-indeterminate samples) was compared with the mRNA expression of the corresponding target genes in FFPE samples, and mRNA expression on FNAC samples was compared with the FFPE results for both mRNA expression and IHC. Concordance between IHC/FISH and Xpert Breast Cancer STRAT4 in FFPE and FNAC samples using the Quick lysis (Q) method (a research-use-only modification of the validated FFPE-lysis method), showed an overall percentage agreement for ESR1 expression of 90.3% and 81.3%, and for PGR1 expression at 86.7% and 81.3% respectively in FFPE and FNAC samples. Concordance was lowest for Ki67 expression, using a binary IHC cutoff for Ki67 positivity at ≥20% staining) at 83.9% and 62.5%, for FFPE and FNAC samples, respectively. This suggests that the STRAT4 assay may be a useful ancillary test in determining HR and Ki67 status in FFPE samples and that use on FNAC samples may be feasible. Future studies should expand the sample numbers and establish locally relevant cutoffs.

Demographics, clinical presentation and risk factors of ocular surface squamous neoplasia at a tertiary hospital, South Africa.

AIMS: The aim of this study is to describe the demographic, presenting features and associated risk factors of ocular surface squamous neoplasia (OSSN) at a tertiary eye hospital in Johannesburg, South Africa. METHODS: An interventional prospective study of patients presenting with conjunctival masses was conducted. An electronic questionnaire was completed to document demographic data, presenting history, and associated risk factors. A slit lamp examination and photos were used to document and describe the clinical features at presentation. Cases (OSSN) and controls (benign lesions) were determined by histology. RESULTS: There were 130 cases and 45 controls. Median age was 44 years (IQR: 35-51) with an equal gender distribution in cases. The prevalence of HIV in cases was 74% and was strongly associated with OSSN (p < 0.001). Vascularisation, leukoplakia and pigmentation were clinical features that distinguished cases from controls. A fibrovascular morphology was strongly associated with a benign histology (p < 0.001), whereas leukoplakic and gelatinous morphologies were associated with OSSN. Conjunctival intra-epithelial neoplasia made up 82% of cases. CONCLUSION: Our study describes a sample of OSSN that is young and has no gender predisposition. The majority of cases presented with CIN lesions, rather than SCC reported in other African countries. HIV was the most significant risk factor in this study population.

Are synthetic cytology images ready for prime time? A comparative assessment of real and synthetic urine cytology images.

INTRODUCTION: The use of synthetic data in pathology has, to date, predominantly been augmenting existing pathology data to improve supervised machine learning algorithms. We present an alternative use case-using synthetic images to augment cytology training when the availability of real-world examples is limited. Moreover, we compare the assessment of real and synthetic urine cytology images by pathology personnel to explore the usefulness of this technology in a real-world setting. MATERIALS AND METHODS: Synthetic urine cytology images were generated using a custom-trained conditional StyleGAN3 model. A morphologically balanced 60-image data set of real and synthetic urine cytology images was created for an online image survey system to allow for the assessment of the differences in visual perception between real and synthetic urine cytology images by pathology personnel. RESULTS: A total of 12 participants were recruited to answer the 60-image survey. The study population had a median age of 36.5 years and a median of 5 years of pathology experience. There was no significant difference in diagnostic error rates between real and synthetic images, nor was there a significant difference between subjective image quality scores between real and synthetic images when assessed on an individual observer basis. CONCLUSIONS: The ability of Generative Adversarial Networks technology to generate highly realistic urine cytology images was demonstrated. Furthermore, there was no difference in how pathology personnel perceived the subjective quality of synthetic images, nor was there a difference in diagnostic error rates between real and synthetic urine cytology images. This has important implications for the application of Generative Adversarial Networks technology to cytology teaching and learning.

Diagnostic accuracy of cervical cancer screening strategies for high-grade cervical intraepithelial neoplasia (CIN2+/CIN3+) among women living with HIV: A systematic review and meta-analysis.

Background: We systematically reviewed the diagnostic accuracy of cervical cancer screening and triage strategies in women living with HIV (WLHIV). Methods: Cochrane Library, Embase, Global Health and Medline were searched for randomised controlled trials, prospective or cross-sectional studies published from database inception to 15 July 2022 reporting diagnostic accuracy of tests in cervical cancer screening and triage of screen-positive WLHIV. Studies were included if they reported the diagnostic accuracy of any cervical cancer screening or triage strategies for the detection of histologically-confirmed high-grade cervical intraepithelial neoplasia (CIN2+/CIN3+) among WLHIV. Summary data were extracted from published reports. Authors were contacted for missing data where applicable. Sensitivity and specificity estimates for CIN2/3+ were pooled using models for meta-analysis of diagnostic accuracy data. Study quality was assessed using the QUADAS-2 tool for the quality assessment of diagnostic accuracy studies. PROSPERO registration:CRD42020189031. Findings: In 38 studies among 18,737 WLHIV, the majority (n=19) were conducted in sub-Saharan Africa. The pooled prevalence was 12.0% (95%CI:9.8-14.1) for CIN2+ and 6.7% (95%CI:5.0-8.4) for CIN3+. The proportion of screen-positive ranged from 3-31% (visual inspection using acetic acid[VIA]); 2-46% (high-grade squamous intraepithelial lesions, and greater [HSIL+] cytology); 20-64% (high-risk[HR]-HPV DNA). In 14 studies, sensitivity and specificity of VIA were variable limiting the reliability of pooled estimates. In 5 studies where majority had histology-confirmed CIN2+, pooled sensitivity was 56.0% (95%CI:45.4-66.1; I2=65%) for CIN2+ and 65.0% (95%CI:52.9-75.4; I2 =42%) for CIN3+; specificity for

A Randomized Clinical Trial of Human Papillomavirus Test-and-Treat as Compared to Cytology-Based Screening for Prevention of Cervical Cancer Among Women With Human Immunodeficiency Virus: AIDS Clinical Trials Group Protocol A5282.

BACKGROUND: Cytology-based cervical cancer screening followed by confirmation and treatment of biopsy-proven high-grade squamous intraepithelial lesions (bHSIL) is difficult to implement in resource-constrained settings. We hypothesized that high-risk human papillomavirus (hrHPV) testing followed by immediate cryotherapy of women with hrHPV (HPV screen-and-treat) may improve outcomes. METHODS: Randomized, open-label, phase 2, multinational clinical trial enrolling women with human immunodeficiency virus (HIV) age 18 or older with cervical hrHPV and having no cervical lesions or lesions appropriate for cryotherapy. Women were randomized to immediate cryotherapy (Arm A) or cytology-based screening (Arm B). For Arm A, cervical biopsies were obtained followed by cervical cryotherapy, and in Arm B, women with abnormal cytology underwent colposcopy followed by loop electroexcision procedure (LEEP) if bHSIL was detected. Women were followed through 30 months. The primary outcome was time to bHSIL detected from Month 6 through study completion. RESULTS: In total, 288 women (145 in Arm A, 143 in Arm B) were randomized: median age 35 years, 84% on antiretroviral therapy, median CD4 501 cells/mm3. In Arm A, 39 (27%) of women had bHSIL at entry, and in Arm B, 88 (62%) had abnormal cytology, 22 (15%) were diagnosed with bHSIL, 12 (8%) underwent LEEP. In follow-up, 30 (21%) and 31 (22%) developed bHSIL; time to bHSIL was similar between arms (P=.94). The prevalence of hrHPV at Month 6 was similar between arms (61% and 70%, P=.13). CONCLUSIONS: HPV test-and-treat was not associated with improved bHSIL outcomes as compared to cytology-based screening. More effective treatment options are required to improve outcomes from screen-and-treat programs. CLINICAL TRIALS REGISTRATION: NCT01315363.

Is it real or not? Toward artificial intelligence-based realistic synthetic cytology image generation to augment teaching and quality assurance in pathology.

INTRODUCTION: Urine cytology offers a rapid and relatively inexpensive method to diagnose urothelial neoplasia. In our setting of a public sector laboratory in South Africa, urothelial neoplasia is rare, compromising pathology training in this specific aspect of cytology. Artificial intelligence-based synthetic image generation-specifically the use of generative adversarial networks (GANs)-offers a solution to this problem. MATERIALS AND METHODS: A limited, but morphologically diverse, dataset of 1000 malignant urothelial cytology images was used to train a StyleGAN3 model to create completely novel, synthetic examples of malignant urine cytology using computer resources within reach of most pathology departments worldwide. RESULTS: We have presented the results of our trained GAN model, which was able to generate realistic, morphologically diverse examples of malignant urine cytology images when trained using a modest dataset. Although the trained model is capable of generating realistic images, we have also presented examples for which unrealistic and artifactual images were generated-illustrating the need for manual curation when using this technology in a training context. CONCLUSIONS: We have presented a proof-of-concept illustration of creating synthetic malignant urine cytology images using machine learning technology to augment cytology training when real-world examples are sparse. We have shown that despite significant morphologic diversity in terms of staining variations, slide background, variations in the diagnostic malignant cellular elements, the presence of other nondiagnostic cellular elements, and artifacts, visually acceptable and varied results are achievable using limited data and computing resources.

COVID-19 pandemic impact on cytopathology practice in the post-lockdown period: An international, multicenter study.

BACKGROUND: In a previous worldwide survey, the authors showed a drastic reduction in the number of cytological specimens processed during the coronavirus disease 2019 "lockdown" period along with an increase in malignancy rates. To assess the continued impact of the pandemic on cytological practices around the world, they undertook a second follow-up worldwide survey collecting data from the post-lockdown period (2020). METHODS: Participants were asked to provide data regarding their cytopathology activity during the first 12 weeks of their respective national post-lockdown period (2020), which ranged from April 4 to October 31. Differences between the post-lockdown period and the corresponding 2019 period were evaluated, and the authors specifically focused on rates of malignant diagnoses. RESULTS: A total of 29 respondents from 17 countries worldwide joined the survey. Overall, a lower number of cytological specimens (n = 236,352) were processed in comparison with the same period in 2019 (n = 321,466) for a relative reduction of 26.5%. The overall malignancy rate showed a statistically significant increase (12,442 [5.26%] vs 12,882 [4.01%]; P < .001) during the same time period. Similar results were obtained if both malignancy and suspicious for malignancy rates were considered together (15,759 [6.58%] vs 16,011 [4.98%]; P < .001). CONCLUSIONS: The data showed a persistent reduction in the cytological specimen volume during the post-lockdown period (2020). However, the relative increase in the cytological workload in the late part of the post-lockdown is a promising finding of a slow return to normality.

The Utility of Unsupervised Machine Learning in Anatomic Pathology.

OBJECTIVES: Developing accurate supervised machine learning algorithms is hampered by the lack of representative annotated datasets. Most data in anatomic pathology are unlabeled and creating large, annotated datasets is a time consuming and laborious process. Unsupervised learning, which does not require annotated data, possesses the potential to assist with this challenge. This review aims to introduce the concept of unsupervised learning and illustrate how clustering, generative adversarial networks (GANs) and autoencoders have the potential to address the lack of annotated data in anatomic pathology. METHODS: A review of unsupervised learning with examples from the literature was carried out. RESULTS: Clustering can be used as part of semisupervised learning where labels are propagated from a subset of annotated data points to remaining unlabeled data points in a dataset. GANs may assist by generating large amounts of synthetic data and performing color normalization. Autoencoders allow training of a network on a large, unlabeled dataset and transferring learned representations to a classifier using a smaller, labeled subset (unsupervised pretraining). CONCLUSIONS: Unsupervised machine learning techniques such as clustering, GANs, and autoencoders, used individually or in combination, may help address the lack of annotated data in pathology and improve the process of developing supervised learning models.

The Dynamics of Pathology Dataset Creation Using Urine Cytology as an Example.

INTRODUCTION: Dataset creation is one of the first tasks required for training AI algorithms but is underestimated in pathology. High-quality data are essential for training algorithms and data should be labelled accurately and include sufficient morphological diversity. The dynamics and challenges of labelling a urine cytology dataset using The Paris System (TPS) criteria are presented. METHODS: 2,454 images were labelled by pathologist consensus via video conferencing over a 14-day period. During the labelling sessions, the dynamics of the labelling process were recorded. Quality assurance images were randomly selected from images labelled in previous sessions within this study and randomly distributed throughout new labelling sessions. To assess the effect of time on the labelling process, the labelled set of images was split into 2 groups according to the median relative label time and the time taken to label images and intersession agreement were assessed. RESULTS: Labelling sessions ranged from 24 m 11 s to 41 m 06 s in length, with a median of 33 m 47 s. The majority of the 2,454 images were labelled as benign urothelial cells, with atypical and malignant urothelial cells more sparsely represented. The time taken to label individual images ranged from 1 s to 42 s with a median of 2.9 s. Labelling times differed significantly among categories, with the median label time for the atypical urothelial category being 7.2 s, followed by the malignant urothelial category at 3.8 s and the benign urothelial category at 2.9 s. The overall intersession agreement for quality assurance images was substantial. The level of agreement differed among classes of urothelial cells - benign and malignant urothelial cell classes showed almost perfect agreement and the atypical urothelial cell class showed moderate agreement. Image labelling times seemed to speed up, and there was no evidence of worsening of intersession agreement with session time. DISCUSSION/CONCLUSION: Important aspects of pathology dataset creation are presented, illustrating the significant resources required for labelling a large dataset. We present evidence that the time taken to categorise urine cytology images varies by diagnosis/class. The known challenges relating to the reproducibility of the AUC (atypical) category in TPS when compared to the NHGUC (benign) or HGUC (malignant) categories is also confirmed.

Diagnostic accuracy of cervical cancer screening and screening-triage strategies among women living with HIV-1 in Burkina Faso and South Africa: A cohort study.

BACKGROUND: Cervical cancer screening strategies using visual inspection or cytology may have suboptimal diagnostic accuracy for detection of precancer in women living with HIV (WLHIV). The optimal screen and screen-triage strategy, age to initiate, and frequency of screening for WLHIV remain unclear. This study evaluated the sensitivity, specificity, and positive predictive value of different cervical cancer strategies in WLHIV in Africa. METHODS AND FINDINGS: WLHIV aged 25-50 years attending HIV treatment centres in Burkina Faso (BF) and South Africa (SA) from 5 December 2011 to 30 October 2012 were enrolled in a prospective evaluation study of visual inspection using acetic acid (VIA) or visual inspection using Lugol's iodine (VILI), high-risk human papillomavirus DNA test (Hybrid Capture 2 [HC2] or careHPV), and cytology for histology-verified high-grade cervical intraepithelial neoplasia (CIN2+/CIN3+) at baseline and endline, a median 16 months later. Among 1,238 women (BF: 615; SA: 623), median age was 36 and 34 years (p < 0.001), 28.6% and 49.6% ever had prior cervical cancer screening (p < 0.001), and 69.9% and 64.2% were taking ART at enrolment (p = 0.045) in BF and SA, respectively. CIN2+ prevalence was 5.8% and 22.4% in BF and SA (p < 0.001), respectively. VIA had low sensitivity for CIN2+ (44.7%, 95% confidence interval [CI] 36.9%-52.7%) and CIN3+ (56.1%, 95% CI 43.3%-68.3%) in both countries, with specificity for ≤CIN1 of 78.7% (95% CI 76.0%-81.3%). HC2 had sensitivity of 88.8% (95% CI 82.9%-93.2%) for CIN2+ and 86.4% (95% CI 75.7%-93.6%) for CIN3+. Specificity for ≤CIN1 was 55.4% (95% CI 52.2%-58.6%), and screen positivity was 51.3%. Specificity was higher with a restricted genotype (HPV16/18/31/33/35/45/52/58) approach (73.5%, 95% CI 70.6%-76.2%), with lower screen positivity (33.7%), although there was lower sensitivity for CIN3+ (77.3%, 95% CI 65.3%-86.7%). In BF, HC2 was more sensitive for CIN2+/CIN3+ compared to VIA/VILI (relative sensitivity for CIN2+ = 1.72, 95% CI 1.28-2.32; CIN3+: 1.18, 95% CI 0.94-1.49). Triage of HC2-positive women with VIA/VILI reduced the number of colposcopy referrals, but with loss in sensitivity for CIN2+ (58.1%) but not for CIN3+ (84.6%). In SA, cytology high-grade squamous intraepithelial lesion or greater (HSIL+) had best combination of sensitivity (CIN2+: 70.1%, 95% CI 61.3%-77.9%; CIN3+: 80.8%, 95% CI 67.5%-90.4%) and specificity (81.6%, 95% CI 77.6%-85.1%). HC2 had similar sensitivity for CIN3+ (83.0%, 95% CI 70.2%-91.9%) but lower specificity compared to HSIL+ (42.7%, 95% CI 38.4%-47.1%; relative specificity = 0.57, 95% CI 0.52-0.63), resulting in almost twice as many referrals. Compared to HC2, triage of HC2-positive women with HSIL+ resulted in a 40% reduction in colposcopy referrals but was associated with some loss in sensitivity. CIN2+ incidence over a median 16 months was highest among VIA baseline screen-negative women (2.2%, 95% CI 1.3%-3.7%) and women who were baseline double-negative with HC2 and VIA (2.1%, 95% CI 1.3%-3.5%) and lowest among HC2 baseline screen-negative women (0.5%, 95% CI 0.1%-1.8%). Limitations of our study are that WLHIV included in the study may not reflect a contemporary cohort of WLHIV initiating ART in the universal ART era and that we did not evaluate HPV tests available in study settings today. CONCLUSIONS: In this cohort study among WLHIV in Africa, a human papillomavirus (HPV) test targeting 14 high-risk (HR) types had higher sensitivity to detect CIN2+ compared to visual inspection but had low specificity, although a restricted genotype approach targeting 8 HR types decreased the number of unnecessary colposcopy referrals. Cytology HSIL+ had optimal performance for CIN2+/CIN3+ detection in SA. Triage of HPV-positive women with HSIL+ maintained high specificity but with some loss in sensitivity compared to HC2 alone.

Artificial Intelligence-Based Screening for Mycobacteria in Whole-Slide Images of Tissue Samples.

OBJECTIVES: This study aimed to develop and validate a deep learning algorithm to screen digitized acid fast-stained (AFS) slides for mycobacteria within tissue sections. METHODS: A total of 441 whole-slide images (WSIs) of AFS tissue material were used to develop a deep learning algorithm. Regions of interest with possible acid-fast bacilli (AFBs) were displayed in a web-based gallery format alongside corresponding WSIs for pathologist review. Artificial intelligence (AI)-assisted analysis of another 138 AFS slides was compared to manual light microscopy and WSI evaluation without AI support. RESULTS: Algorithm performance showed an area under the curve of 0.960 at the image patch level. More AI-assisted reviews identified AFBs than manual microscopy or WSI examination (P < .001). Sensitivity, negative predictive value, and accuracy were highest for AI-assisted reviews. AI-assisted reviews also had the highest rate of matching the original sign-out diagnosis, were less time-consuming, and were much easier for pathologists to perform (P < .001). CONCLUSIONS: This study reports the successful development and clinical validation of an AI-based digital pathology system to screen for AFBs in anatomic pathology material. AI assistance proved to be more sensitive and accurate, took pathologists less time to screen cases, and was easier to use than either manual microscopy or viewing WSIs.

Ocular surface squamous neoplasia: management and outcomes.

Ocular surface squamous neoplasia (OSSN) is the most common ocular tumour with an incidence ranging from 0.03 to 1.9 per 100,000 persons/year. The diagnosis is made on clinical suspicion and confirmed with anterior-segment optical coherence tomography (AS-OCT), cytology, or histology. The purpose of this review is to provide an overview of the management options available for OSSN and review their success and recurrence rates. Surgery is the gold standard for the management of small OSSN lesions. With the increased use of less invasive diagnostic modalities such as AS-OCT and cytology, there has been a move to use topical therapies for the management of OSSN. The most commonly used agents are interferon-α2b (IFN), mitomycin-C (MMC) and 5-fluorouracil (5FU). They have been shown to have similar resolution and recurrence rates but differ in cost and side effect profile. IFN has the lowest side effect profile, but is also the most expensive, whereas MMC has the greatest surface toxicity and is priced midway between the three. 5FU is the cheapest of the three topical agents with less surface toxicity than MMC. Radiotherapy is mostly employed as adjuvant therapy. Newer novel therapies are available but have not been widely adopted as mainstream therapy due to cost and lack of clinical evidence. OSSN has the benefit of many management options. No single modality has been shown to superior and some patients will need the use of combination therapy to achieve an optimal clinical outcome.