USE OF ARTIFICIAL INTELLIGENCE IN EARLY DETECTION OF CERVICAL INTRAEPITHELIAL NEOPLASIA FROM COLPOSCOPIC IMAGES
Main Article Content
Keywords
Artificial Intelligence; Colposcopy; Cervical Intraepithelial Neoplasia; Early Detection
Abstract
Cervical intraepithelial neoplasia (CIN) is a precancer lesion that is closely associated with human papillomavirus (HPV) long-standing infection. Prevention of cervical cancer development is more important in early diagnosis. Colposcopy is the most common diagnostic test, but it continues to require a subjective decision on the part of the clinician. Artificial intelligence (AI) provides an opportunity to enhance the quality of diagnosis by means of automatic analysis of colposcopy images.
Objectives: The purpose of this study is to assess the performance of artificial intelligence in the detection of cervical intraepithelial neoplasia through colposcopy images, compare the accuracy of the diagnosis with the traditional approaches, and evaluate how it can be used to improve the detection of the disease and its outcome.
Study design: A Observational Study.
Place and duration of study: This observational study was conducted at the Qazi hussain Ahmed Medical Complex Nowshera from October 2024 to March 2025
Methods: 100 patients undergoing colposcopy were enrolled in the current study. Images of high-resolution colposcopy were collected and processed in an AI system based on a convolutional neural network (CNN). Histopathological results which are regarded as a gold standard were compared with the diagnoses. The values of sensitivity, specificity and accuracy were computed. The statistical analysis consisted of determining mean age, standard deviation (SD), and p-values of chi-square and t-test to determine the significance of AI versus conventional diagnostic performance.
Results: 100 women who were referred to colposcopy on the basis of abnormal cytology participated in the study. The average age of the patients was 37.5 years with standard deviation of 8.4. CIN was confirmed in 64 patients and 36 were benign in histopathology. The AI system was found to be 92 percent sensitive, 88 percent specific, and 90 percent accurate, whereas the conventional colposcopy interpretation was found to be 82 percent sensitive and 75 percent specific. It showed statistically significant difference (p = 0.02) between the AI model and the human interpretation. These results suggest that AI may help clinicians with fewer interobserver variations and higher diagnostic reliability in the detection of CIN.
Conclusion: AI shows good promise in identifying cervical intraepithelial neoplasia on colposcopy images at an early stage, which is better than traditional interpretation in terms of sensitivity, specificity and overall accuracy. Its use can decrease observer bias, improve diagnostic consistency, and improve the timely management of cervical precancerous lesions. The adoption of AI in clinical practice may assist gynecologists, particularly in resource-poor environments, to provide a beneficial supplement to regular colposcopy to enhance cervical cancer prevention approaches.
References
2. Bedell SL, Goldstein LS, Goldstein AR, Goldstein AT. Cervical Cancer Screening: Past, Present, and Future. Sexual medicine reviews. 2020;8:28-37. doi: https://doi.org/10.1016/j.sxmr.2019.09.005.
3. Chen M, Xue P, Li Q, Shen Y, Ye Z, Wang H, et al. Enhancing colposcopy training using a widely accessible digital education tool in China. American journal of obstetrics and gynecology. 2023;229:538.e1-.e9. doi: https://doi.org/10.1016/j.ajog.2023.07.043.
4. Chen X, Pu X, Chen Z, Li L, Zhao KN, Liu H, et al. Application of EfficientNet-B0 and GRU-based deep learning on classifying the colposcopy diagnosis of precancerous cervical lesions. Cancer medicine. 2023;12:8690-9. doi: https://doi.org/10.1002/cam4.5581.
5. De Rosa N, Lavitola G, Della Corte L, Bifulco G. Diagnostic Accuracy of Endocervicoscopy in Identifying and Grading Cervical Intraepithelial Neoplasia Lesion. Gynecologic and obstetric investigation. 2020;85:196-205. doi: https://doi.org/10.1159/000506801.
6. Fachetti-Machado G, Figueiredo-Alves RR, Moreira MAR. Performance of three colposcopic images for the identification of squamous and glandular cervical precursor neoplasias. Archives of gynecology and obstetrics. 2022;305:1319-27. doi: https://doi.org/10.1007/s00404-021-06284-4.
7. Fu Z, Fan Y, Wu C, Yan P, Ye Y, Yang H, et al. Clinical efficacy and mechanism for focused ultrasound (FUS) in the management of cervical intraepithelial neoplasia 1 (CIN1). International journal of hyperthermia : the official journal of European Society for Hyperthermic Oncology, North American Hyperthermia Group. 2020;37:339-45. doi: https://doi.org/10.1080/02656736.2020.1749316.
8. Ito Y, Miyoshi A, Ueda Y, Tanaka Y, Nakae R, Morimoto A, et al. An artificial intelligence-assisted diagnostic system improves the accuracy of image diagnosis of uterine cervical lesions. Molecular and clinical oncology. 2022;16:27. doi: https://doi.org/10.3892/mco.2021.2460.
9. Khan S, Qadir M, Khalid A, Ashraf S, Ahmad I. Characterization of cervical tissue using Mueller matrix polarimetry. Lasers in medical science. 2023;38:46. doi: https://doi.org/10.1007/s10103-023-03712-6.
10. Kim J, Park CM, Kim SY, Cho A. Convolutional neural network-based classification of cervical intraepithelial neoplasias using colposcopic image segmentation for acetowhite epithelium. Scientific reports. 2022;12:17228. doi: https://doi.org/10.1038/s41598-022-21692-5.
11. Li Y, Chen J, Xue P, Tang C, Chang J, Chu C, et al. Computer-Aided Cervical Cancer Diagnosis Using Time-Lapsed Colposcopic Images. IEEE transactions on medical imaging. 2020;39:3403-15. doi: https://doi.org/10.1109/tmi.2020.2994778.
12. Li Y, Liu ZH, Xue P, Chen J, Ma K, Qian T, et al. GRAND: A large-scale dataset and benchmark for cervical intraepithelial Neoplasia grading with fine-grained lesion description. Medical image analysis. 2021;70:102006. doi: https://doi.org/10.1016/j.media.2021.102006.
13. Lycke KD, Kalpathy-Cramer J, Jeronimo J, de Sanjose S, Egemen D, Del Pino M, et al. Agreement on Lesion Presence and Location at Colposcopy. Journal of lower genital tract disease. 2024;28:37-42. doi: https://doi.org/10.1097/lgt.0000000000000786.
14. Mantoani PTS, Jammal MP, Caixeta JM, Cordeiro NA, Barcelos ACM, Murta EFC, et al. Association of lesion area measured by colposcopy and cervical neoplasia. Journal of obstetrics and gynaecology : the journal of the Institute of Obstetrics and Gynaecology. 2022;42:306-9. doi: https://doi.org/10.1080/01443615.2021.1904218.
15. Mantoani PTS, Vieira JF, Menchete TT, Jammal MP, Michelin MA, Barcelos ACM, et al. CIN Extension at Colposcopy: Relation to Treatment and Blood Parameters. Journal of obstetrics and gynaecology Canada : JOGC = Journal d'obstetrique et gynecologie du Canada : JOGC. 2022;44:255-60. doi: https://doi.org/10.1016/j.jogc.2021.10.008.
16. Pereira Pinto P, Zanine RM. Diagnostic value of p16 and Ki-67 expression in cervical glandular intraepithelial disease: A review. Annals of diagnostic pathology. 2023;62:152054. doi: https://doi.org/10.1016/j.anndiagpath.2022.152054.
17. Perkins R, Jeronimo J, Hammer A, Novetsky A, Guido R, Del Pino M, et al. Comparison of accuracy and reproducibility of colposcopic impression based on a single image versus a two-minute time series of colposcopic images. Gynecologic oncology. 2022;167:89-95. doi: https://doi.org/10.1016/j.ygyno.2022.08.001.
18. Ruan Y, Liu M, Guo J, Zhao J, Niu S, Li F. Evaluation of the accuracy of colposcopy in detecting high-grade squamous intraepithelial lesion and cervical cancer. Archives of gynecology and obstetrics. 2020;302:1529-38. doi: https://doi.org/10.1007/s00404-020-05740-x.
19. Sahlgren HAI, Elfgren K, Sparen P, Elfstrom MK. Colposcopic performance in a birth cohort previously eligible for human papillomavirus vaccination. American journal of obstetrics and gynecology. 2022;226:704.e1-.e9. doi: https://doi.org/10.1016/j.ajog.2021.11.1372.
20. Takahashi T, Matsuoka H, Sakurai R, Akatsuka J, Kobayashi Y, Nakamura M, et al. Development of a prognostic prediction support system for cervical intraepithelial neoplasia using artificial intelligence-based diagnosis. Journal of gynecologic oncology. 2022;33:e57. doi: https://doi.org/10.3802/jgo.2022.33.e57.
21. Uchita K, Kobara H, Yorita K, Shigehisa Y, Kuroiwa C, Nishiyama N, et al. Quality Assessment of Endoscopic Forceps Biopsy Samples under Magnifying Narrow Band Imaging for Histological Diagnosis of Cervical Intraepithelial Neoplasia: A Feasibility Study. Diagnostics (Basel, Switzerland). 2021;11:doi: https://doi.org/10.3390/diagnostics11020360.
22. Wu A, Xue P, Abulizi G, Tuerxun D, Rezhake R, Qiao Y. Artificial intelligence in colposcopic examination: A promising tool to assist junior colposcopists. Frontiers in medicine. 2023;10:1060451. doi: https://doi.org/10.3389/fmed.2023.1060451.
23. Yu H, Fan Y, Ma H, Zhang H, Cao C, Yu X, et al. Segmentation of the cervical lesion region in colposcopic images based on deep learning. Frontiers in oncology. 2022;12:952847. doi: https://doi.org/10.3389/fonc.2022.952847.
Article Sidebar
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
All articles published in JPTCP are licensed under Copyright Creative Commons Attribution-NonCommercial 4.0 International License.
Shomaila Ishaq
Trainee Registrar DHQ Hospital Orakzai
Afshan Rani
Trainee Medical Officer, Qazi Hussain Ahmad Medical Complex, Nowshera.
Javairia
Trainee Medical Officer Khyber Teaching Hospital Peshawar
Saba Khattak
Consultant Gynecologist DHQ Hospital Orakzai
Nisma
Medical Officer, Lady Reading Hospital Peshawar