Colposcopy audit for improving quality of service in areas with a high incidence of cervical cancer

Development of a prognostic prediction support system for cervical intraepithelial neoplasia using artificial intelligence-based diagnosis

Objective

Human papillomavirus subtypes are predictive indicators of cervical intraepithelial neoplasia (CIN) progression. While colposcopy is also an essential part of cervical cancer prevention, its accuracy and reproducibility are limited because of subjective evaluation. This study aimed to develop an artificial intelligence (AI) algorithm that can accurately detect the optimal lesion associated with prognosis using colposcopic images of CIN2 patients by utilizing objective AI diagnosis.

Methods

We identified colposcopic findings associated with the prognosis of patients with CIN2. We developed a convolutional neural network that can automatically detect the rate of high-grade lesions in the uterovaginal area in 12 segments. We finally evaluated the detection accuracy of our AI algorithm compared with the scores by multiple gynecologic oncologists.

Results

High-grade lesion occupancy in the uterovaginal area detected by senior colposcopists was significantly correlated with the prognosis of patients with CIN2. The detection rate for high-grade lesions in 12 segments of the uterovaginal area by the AI system was 62.1% for recall, and the overall correct response rate was 89.7%. Moreover, the percentage of high-grade lesions detected by the AI system was significantly correlated with the rate detected by multiple gynecologic senior oncologists (r=0.61).

Conclusion

Our novel AI algorithm can accurately determine high-grade lesions associated with prognosis on colposcopic images, and these results provide an insight into the additional utility of colposcopy for the management of patients with CIN2.

High-grade lesion occupancy in the uterovaginal area was significantly correlated with CIN2 patients’ prognosis. The number of high-grade lesions in 12 segments detected by an artificial intelligence (AI)-based system was comparable to that detected by senior colposcopists. The overall correct response rate of the AI algorithm for detecting high-grade lesions was 89.7%.

Is There a Place for the Introduction of Colposcopy Quality Standards?

OBJECTIVES

The aims of the study were to examine the degree of compliance with international quality measures for colposcopy in Israel, which does not currently have formal guidelines and to compare the achievement of quality measures between clinical settings.

METHODS

This is a retrospective cohort study, in a hospital, a community clinic, and 2 private colposcopy clinics in Israel, including women aged 18-70 years presenting for colposcopy after abnormal Pap results. Compliance was compared between clinical sites regarding 6 international standards: documentation of reason for referral, type of transformation zone, biopsy location, and grade; proportion of women with high-degree cytological abnormalities (atypical squamous cells - cannot exclude high grade squamous intraepithelial lesion and above) receiving a colposcopy within 4 weeks; and the positive predictive value of colposcopy to detect cervical intraepithelial neoplasia 2 and above.

RESULTS

Documentation of reason for referral (1.3% of target), transformation zone type (22.6% of target), biopsy location (18% of target), and lesion grade (31% of target) all failed to meet international standards, as did the proportion of patients with high-degree cytological abnormalities who underwent colposcopy within 4 weeks (32.9% of the target). The positive predictive value of colposcopy exceeded standards (30% above target). Differences existed between clinical settings.

CONCLUSIONS

In Israel, there is a considerable shortfall in performance and documentation of most international quality measures for colposcopy. Quality measures for cervical examinations and colposcopy should be considered for inclusion in the National Program for Quality Measures.

Is Age an Independent Predictor of High-Grade Histopathology in Women Referred for Colposcopy after Abnormal Cervical Cytology?

This study was conducted to determine whether advancing age is an independent predictor of increased risk of high-grade pathologies among women referred for colposcopy after abnormal cervical cytology. Medical records were reviewed for women with abnormal cervical cytology who underwent colposcopy at Khon Kaen University Hospital. Logistic regression was used to determine the independent impact of age on the risk of high-grade pathologies. Mean age of the women was 42.8 years. Of 482 women, 97 (20.1%) were postmenopausal, and 92 (19.1%) were nulliparous. The rate of high-grade pathologies included cervical intraepithelial neoplasia 2-3, 99 (20.5%), adenocarcinoma in situ, 4 (0.8%), cervical cancer, 30 (6.2%), and endometrial cancer, 1 (0.2%). The prevalence of significant lesions was 26.9% (95% CI, 23.1%-31.2%). In total, 31 women had cancers (6.4%; 95% CI, 4.4%-9.0%). When controlling for smear types and parity, age was noted to be a significant independent predictor of high-grade histopathology. Women older than 35-40 years were approximately 2 times as likely to have severe histopathology as the younger women. This study illustrates the substantial risk of underlying significant lesions especially invasive cancer in Thai women with abnormal cervical cytology. Age was a significant independent factor predicting the risk of high-grade pathologies.

Classification of images acquired with colposcopy using artificial neural networks

OBJECTIVE

To explore the advantages of using artificial neural networks (ANNs) to recognize patterns in colposcopy to classify images in colposcopy.

PURPOSE

Transversal, descriptive, and analytical study of a quantitative approach with an emphasis on diagnosis. The training test e validation set was composed of images collected from patients who underwent colposcopy. These images were provided by a gynecology clinic located in the city of Criciúma (Brazil). The image database (n = 170) was divided; 48 images were used for the training process, 58 images were used for the tests, and 64 images were used for the validation. A hybrid neural network based on Kohonen self-organizing maps and multilayer perceptron (MLP) networks was used.

RESULTS

After 126 cycles, the validation was performed. The best results reached an accuracy of 72.15%, a sensibility of 69.78%, and a specificity of 68%.

CONCLUSION

Although the preliminary results still exhibit an average efficiency, the present approach is an innovative and promising technique that should be deeply explored in the context of the present study.

Rationale and development of an on-line quality assurance programme for colposcopy in a population-based cervical screening setting in Italy

Background

Colposcopy, the key step in the management of women with abnormal Pap smear results, is a visual technique prone to observer variation, which implies the need for prolonged apprenticeship, continuous training, and quality assurance (QA) measures. Colposcopy QA programmes vary in level of responsibility of organizing subjects, geographic coverage, scope, model, and type of actions. The programmes addressing the clinical standards of colposcopy (quality of examination and appropriateness of clinical decisions) are more limited in space and less sustainable over time than those focused on the provision of the service (resources, accessibility, etc.). This article reports on the protocol of a QA programme targeting the clinical quality of colposcopy in a population-based cervical screening service in an administrative region of northern Italy.

Methods/design

After a situation analysis of local colposcopy audit practices and previous QA initiatives, a permanent web-based QA programme was developed. The design places more emphasis on providing education and feedback to participants than on testing them. The technical core is a log-in web application accessible on the website of the regional Administration. The primary objectives are to provide (1) a practical opportunity for retraining of screening colposcopists, and (2) a platform for them to interact with colposcopists from other settings and regions through exchange and discussion of digital colposcopic images. The retraining function is based on repeated QA sessions in which the registered colposcopists log-in, classify a posted set of colpophotographs, and receive on line a set of personal feedback data. Each session ends with a plenary seminar featuring the presentation of overall results and an interactive review of the test set of colpophotographs. This is meant to be a forum for an open exchange of views that may lead to more knowledge and more diagnostic homogeneity. The protocol includes the criteria for selection of colpophotographs and the rationale for colposcopic gold standards.

Discussion

This programme is an ongoing initiative open to further developments, in particular in the area of basic training. It uses the infrastructure of the internet to give a novel solution to technical problems affecting colposcopy QA in population-based screening services.