Analysis of the agreement between colposcopic impression and histopathological diagnosis of cervical biopsy in a single tertiary center of Chengdu

Artificial intelligence in colposcopic examination: A promising tool to assist junior colposcopists

Introduction

Well-trained colposcopists are in huge shortage worldwide, especially in low-resource areas. Here, we aimed to evaluate the Colposcopic Artificial Intelligence Auxiliary Diagnostic System (CAIADS) to detect abnormalities based on digital colposcopy images, especially focusing on its role in assisting junior colposcopist to correctly identify the lesion areas where biopsy should be performed.

Materials and methods

This is a hospital-based retrospective study, which recruited the women who visited colposcopy clinics between September 2021 to January 2022. A total of 366 of 1,146 women with complete medical information recorded by a senior colposcopist and valid histology results were included. Anonymized colposcopy images were reviewed by CAIADS and a junior colposcopist separately, and the junior colposcopist reviewed the colposcopy images with CAIADS results (named CAIADS-Junior). The diagnostic accuracy and biopsy efficiency of CAIADS and CAIADS-Junior were assessed in detecting cervical intraepithelial neoplasia grade 2 or worse (CIN2+), CIN3+, and cancer in comparison with the senior and junior colposcipists. The factors influencing the accuracy of CAIADS were explored.

Results

For CIN2 + and CIN3 + detection, CAIADS showed a sensitivity at ~80%, which was not significantly lower than the sensitivity achieved by the senior colposcopist (for CIN2 +: 80.6 vs. 91.3%, p = 0.061 and for CIN3 +: 80.0 vs. 90.0%, p = 0.189). The sensitivity of the junior colposcopist was increased significantly with the assistance of CAIADS (for CIN2 +: 95.1 vs. 79.6%, p = 0.002 and for CIN3 +: 97.1 vs. 85.7%, p = 0.039) and was comparable to those of the senior colposcopists (for CIN2 +: 95.1 vs. 91.3%, p = 0.388 and for CIN3 +: 97.1 vs. 90.0%, p = 0.125). In detecting cervical cancer, CAIADS achieved the highest sensitivity at 100%. For all endpoints, CAIADS showed the highest specificity (55-64%) and positive predictive values compared to both senior and junior colposcopists. When CIN grades became higher, the average biopsy numbers decreased for the subspecialists and CAIADS required a minimum number of biopsies to detect per case (2.2-2.6 cut-points). Meanwhile, the biopsy sensitivity of the junior colposcopist was the lowest, but the CAIADS-assisted junior colposcopist achieved a higher biopsy sensitivity.

Conclusion

Colposcopic Artificial Intelligence Auxiliary Diagnostic System could assist junior colposcopists to improve diagnostic accuracy and biopsy efficiency, which might be a promising solution to improve the quality of cervical cancer screening in low-resource settings.

Colposcopic accuracy in diagnosing squamous intraepithelial lesions: a systematic review and meta-analysis of the International Federation of Cervical Pathology and Colposcopy 2011 terminology

BACKGROUND

Colposcopy is an important tool in diagnosing cervical cancer, and the International Federation of Cervical Pathology and Colposcopy (IFCPC) issued the latest version of the guidelines in 2011. This study aims to systematically assess the accuracy of colposcopy in predicting low-grade squamous intraepithelial lesions or worse (LSIL+) / high-grade squamous intraepithelial lesions or worse (HSIL+) under the 2011 IFCPC terminology.

METHODS

We performed a systematic review and meta-analysis, following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We searched for studies about the performance of colposcopy in diagnosing cervical intraepithelial neoplasia under the new IFCPC colposcopy terminology from PubMed, Embase, Web of Science and the Cochrane database. Data were independently extracted by two authors and an overall diagnostic performance index was calculated under two colposcopic thresholds.

RESULTS

Totally, fifteen articles with 22,764 participants in compliance with the criteria were included in meta-analysis. When colposcopy was used to detect LSIL+, the combined sensitivity and specificity were 0.92 (95% CI 0.88-0.95) and 0.51 (0.43-0.59), respectively. When colposcopy was used to detect HSIL+, the combined sensitivity and specificity were 0.68 (0.58-0.76) and 0.93 (0.88-0.96), respectively.

CONCLUSION

In accordance with the 2011 IFCPC terminology, the accuracy of colposcopy has improved in terms of both sensitivity and specificity. Colposcopy is now more sensitive with LSIL+ taken as the cut-off value and is more specific to HSIL+. These findings suggest we are avoiding under- or overdiagnosis both of which impact on patients' well-being.

Application of EfficientNet-B0 and GRU-based deep learning on classifying the colposcopy diagnosis of precancerous cervical lesions

BACKGROUND

Colposcopy is indispensable for the diagnosis of cervical lesions. However, its diagnosis accuracy for high-grade squamous intraepithelial lesion (HSIL) is at about 50%, and the accuracy is largely dependent on the skill and experience of colposcopists. The advancement in computational power made it possible for the application of artificial intelligence (AI) to clinical problems. Here, we explored the feasibility and accuracy of the application of AI on precancerous and cancerous cervical colposcopic image recognition and classification.

METHODS

The images were collected from 6002 colposcopy examinations of normal control, low-grade squamous intraepithelial lesion (LSIL), and HSIL. For each patient, the original, Schiller test, and acetic-acid images were all collected. We built a new neural network classification model based on the hybrid algorithm. EfficientNet-b0 was used as the backbone network for the image feature extraction, and GRU(Gate Recurrent Unit)was applied for feature fusion of the three modes examinations (original, acetic acid, and Schiller test).

RESULTS

The connected network classifier achieved an accuracy of 90.61% in distinguishing HSIL from normal and LSIL. Furthermore, the model was applied to "Trichotomy", which reached an accuracy of 91.18% in distinguishing the HSIL, LSIL and normal control at the same time.

CONCLUSION

Our results revealed that as shown by the high accuracy of AI in the classification of colposcopic images, AI exhibited great potential to be an effective tool for the accurate diagnosis of cervical disease and for early therapeutic intervention in cervical precancer.

Evaluating the Feasibility of Machine-Learning-Based Predictive Models for Precancerous Cervical Lesions in Patients Referred for Colposcopy

Background: Colposcopy plays an essential role in cervical cancer control, but its performance remains unsatisfactory. This study evaluates the feasibility of machine learning (ML) models for predicting high-grade squamous intraepithelial lesions or worse (HSIL+) in patients referred for colposcopy by combining colposcopic findings with demographic and screening results. Methods: In total, 7485 patients who underwent colposcopy examination in seven hospitals in mainland China were used to train, internally validate, and externally validate six commonly used ML models, including logistic regression, decision tree, naïve bayes, support vector machine, random forest, and extreme gradient boosting. Nine variables, including age, gravidity, parity, menopause status, cytological results, high-risk human papillomavirus (HR-HPV) infection type, HR-HPV multi-infection, transformation zone (TZ) type, and colposcopic impression, were used for model construction. Results: Colposcopic impression, HR-HPV results, and cytology results were the top three variables that determined model performance among all included variables. In the internal validation set, six ML models that integrated demographics, screening results, and colposcopic impression showed significant improvements in the area under the curve (AUC) (0.067 to 0.099) and sensitivity (11.55% to 14.88%) compared with colposcopists. Greater increases in AUC (0.087 to 0.119) and sensitivity (17.17% to 22.08%) were observed in the six models with the external validation set. Conclusions: By incorporating demographics, screening results, and colposcopic impressions, ML improved the AUC and sensitivity for detecting HSIL+ in patients referred for colposcopy. Such models could transform the subjective experience into objective judgments to help clinicians make decisions at the time of colposcopy examinations.

Which is the best management for women with normal cervical cytologic findings despite positivity for non-16/18 high risk human papillomaviruses?

Women who test positive for the human papillomavirus (HPV) but have normal cytology constitute the predominant subgroup of patients in the screening population in the post-vaccination era. The distribution of HPV genotypes changed dramatically, which was attributable to an increase in HPV vaccination coverage. These changes have created uncertainty about how to properly manage women with normal cytology, non-HPV16/18 infections, or persistent infections. Current recommendations include retesting and continued surveillance in the absence of HPV16/18 infection. However, these are not always applicable. The ability to implement genotyping or incorporate HPV16/18 with some additional high-risk HPV (HR-HPV) types for triage and management with the aim of identifying type-specific risks in this population could be acceptable. When the next set of guidelines is updated, generating potential triage strategies for detecting high-grade cervical lesions, such as the p16/Ki67 cytology assay and other alternatives that incorporate genotyping with newer tests, should be considered. Current clinical management is shifting to risk-based strategies; however, no specific risk threshold has been established in this population. Importantly, innovative triage testing should be evaluated in combination with primary screening and management. Furthermore, there is an untapped opportunity to coordinate HPV genotyping in combination with colposcopic characteristics to modify risk in this group. Hence, providing a more personalized schedule through the efficient application of risk stratification and improving the detection of pre-cancer and cancer is an option worth exploring.

Improving colposcopic accuracy for cervical precancer detection: a retrospective multicenter study in China

Background

Colposcopy alone can result in misidentification of high-grade squamous intraepithelial or worse lesions (HSIL +), especially for women with Type 3 transformation zone (TZ) lesions, where colposcopic assessment is particularly imprecise. This study aimed to improve HSIL + case identification by supplementing referral screening results to colposcopic findings.

Methods

This is an observational multicenter study of 2,417 women, referred to colposcopy after receiving cervical cancer screening results. Logistic regression analysis was conducted under uni- and multivariate models to identify factors which could be used to improve HSIL + case identification. Histological diagnosis was established as the gold standard and is used to assess accuracy, sensitivity, and specificity, as well as to incrementally improve colposcopy.

Results

Multivariate analysis highlighted age, TZ types, referral screening, and colposcopists’ skills as independent factors. Across this sample population, diagnostic accuracies for detecting HSIL + increased from 72.9% (95%CI 71.1–74.7%) for colposcopy alone to 82.1% (95%CI 80.6–83.6%) after supplementing colposcopy with screening results. A significant increase in colposcopic accuracy was observed across all subgroups. Although, the highest increase was observed in women with a TZ3 lesion, and for those diagnosed by junior colposcopists.

Conclusion

It appears possible to supplement colposcopic examinations with screening results to improve HSIL + detection, especially for women with TZ3 lesions. It may also be possible to improve junior colposcopists’ diagnoses although, further psychological research is necessary. We need to understand how levels of uncertainty influence diagnostic decisions and what the concept of “experience” actually is and what it means for colposcopic practice.

Assessing colposcopic accuracy for high-grade squamous intraepithelial lesion detection: a retrospective, cohort study

Background

Inappropriate management of high-grade squamous intraepithelial lesions (HSIL) may be the result of an inaccurate colposcopic diagnosis. The aim of this study was to assess colposcopic performance in identifying HSIL+ cases and to analyze the associated clinical factors.

Methods

Records from 1130 patients admitted to Shenzhen Maternal and Child Healthcare Hospital from 12th January, 2018 up until 30th December, 2018 were retrospectively collected, and included demographics, cytological results, HPV status, transformation zone type, number of cervical biopsy sites, colposcopists’ competencies, colposcopic impressions, as well as histopathological results. Colposcopy was carried out using 2011 colposcopic terminology from the International Federation of Cervical Pathology and Colposcopy. Logistic regression modelling was implemented for uni- and multivariate analyses. A forward stepwise approach was adopted in order to identify variables associated with colposcopic accuracy. Histopathologic results were taken as the comparative gold standard.

Results

Data from 1130 patient records were collated and analyzed. Colposcopy was 69.7% accurate in identifying HSIL+ cases. Positive predictive value, negative predictive value, sensitivity and specificity of detecting HSIL or more (HSIL+) were 35.53%, 64.47%, 42.35% and 77.60%, respectively. Multivariate analysis highlighted the number of biopsies, cytology, and transformation zone type as independent factors. Age and HPV subtype did not appear to statistically correlate with high-grade lesion/carcinoma.

Conclusion

Evidence presented here suggests that colposcopy is only 69.7% accurate at diagnosing HSIL. Even though not all HSIL will progress into cancer it is considered pre-cancerous and therefore early identification will save lives. The number of biopsies, cytology and transformation zone type appear to be predictors of misdiagnosis and therefore should be considered during clinical consultations and by way of further research.

Optimizing the Detection of Occult Cervical Cancer: A Prospective Multicentre Study in China

Purpose

Early-stage cervical cancer is usually diagnosed by colposcopy-directed biopsy (CDB) and/or endocervical curettage (ECC), but some neglected lesions must be detected by conization because they are occult. This study aimed to explore the optimal method for detecting these "occult" cervical cancers.

Patients and Methods

A total of 1299 patients who were high-risk for early-stage cervical cancer from five centres in China were prospectively included. We evaluated the diagnostic performance of cytology, HPV testing, colposcopy and CDB&ECC for detecting "occult" cervical cancer and discussed the diagnostic importance of transformation zone (TZ) type, conization length and the proportion of cervical cone excision.

Results

The diagnostic agreement between colposcopy impression and conization was 64.5% and 72.4% between CDB&ECC and conization. Forty-two patients were finally diagnosed with pathologic cancer, and the sensitivities of cytology, colposcopy, CDB&ECC were 4.8%, 7.1%, and 47.4%, respectively. Twenty cases were neglected by CDB&ECC but further diagnosed as cancer by conization, considered to be occult cervical cancer, accounting for 1.6%. Cytologic high-grade squamous intraepithelial lesion (HSIL)+, positive HPV, biopsy HSIL+ and cervical TZ type 3 were considered risk factors for developing HSIL+, while colposcopy impression HSIL+ was not. There was a significant difference between cancerous and HSIL patients in the proportion of cervical cone excision (P<0.001), which was recognized as a risk factor (P<0.001) for detecting cancer, while the length of cervical cone excision was not. The average proportion was 0.62, and the minimal effective proportion was 0.56.

Conclusion

Since the incidence of occult cervical cancer neglected by CDB&ECC, colposcopy and cytology was far beyond expectations, conization is necessary, especially in patients with TZ type 3, high-grade cytology and biopsy results. As the cervical length varies in patients, the proportion of cervical cone excision might be a better indicator for detecting occult cervical cancer.