Evaluation of AI-assisted colposcopy for detecting high-risk subtypes of human papillomavirus in CIN2

Cervical cancer is the second most common cancer in women, and the role of human papillomavirus (HPV) testing in its etiology is becoming increasingly important. We aimed to evaluate the performance of an artificial intelligence (AI)-assisted colposcopy system in detecting high-risk human papillomavirus (HR-HPV) subtypes in cervical intraepithelial neoplasia (CIN) 2 patients. We conducted a post-hoc analysis of a previous observational study that developed an AI algorithm for colposcopic images of patients with CIN2. Out of 78 patients with HR-HPVs (HPV 16, 18, 31, 33, 35, 45, 52 and 58), 60 (76.9%) had positive AI colposcopy results. The accuracy, sensitivity and specificity of the AI-assisted colposcopy system in detecting HR-HPVs (16, 18, 31, 33, 35, 45, 52 and 58) were 0.689, 0.769 and 0.537, respectively. This study is the first to focus on using AI to detect high-risk subtypes. The AI algorithm accurately detects the characteristics of HR-HPVs. It can be used to detect high-risk CIN types in patients with cervical dysplasia based only on colposcopic imaging findings and is potentially a valuable tool for follow-up.

Colposcopy; Artificial intelligence; Cervical intraepithelial neoplasia; Deep learning; Human papillomavirus

[1] Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal for Clinicians. 2021; 71: 209–249.

[2] Chu D, Liu T, Yao Y. Implications of viral infections and oncogenesis in uterine cervical carcinoma etiology and pathogenesis. Frontiers in Microbiology. 2023; 14: 1194431.

[3] Onuki M, Matsumoto K, Iwata T, Yamamoto K, Aoki Y, Maenohara S, et al. Human papillomavirus genotype contribution to cervical cancer and precancer: implications for screening and vaccination in Japan. Cancer Science. 2020; 111: 2546–2557.

[4] Lukic A, De Vincenzo R, Ciavattini A, Ricci C, Senatori R, Ruscito I, et al. Are we facing a new colposcopic practice in the HPV vaccination era? Opportunities, challenges, and new perspectives. Vaccines. 2021; 9: 1081.

[5] Hoste G, Vossaert K, Poppe WAJ. The clinical role of HPV testing in primary and secondary cervical cancer screening. Obstetrics and Gynecology International. 2013; 2013: 610373.

[6] Singer A, Monaghan JM, Quek SC, Deery ARS. Human papillomaviruses in pathogenesis of lower genital tract neoplasia. In Singer A, Monaghan JM, Quek SC, Deery ARS (eds.) Lower Genital Tract Precancer: Colposcopy, Pathology and Treatment (pp. 15–33). 2nd edn. Blackwell Publishing: London. 2000.

[7] 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.

[8] Fan A, Wang C, Zhang L, Yan Y, Han C, Xue F. Diagnostic value of the 2011 international federation for cervical pathology and colposcopy terminology in predicting cervical lesions. Oncotarget. 2018; 9: 9166–9176.

[9] 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.

[10] Massad LS, Einstein MH, Huh WK, Katki HA, Kinney WK, Schiffman M, et al. 2012 updated consensus guidelines for the management of abnormal cervical cancer screening tests and cancer precursors. Obstetrics & Gynecology. 2013; 121: 829–846.

[11] Kondo K, Uenoyama A, Kitagawa R, Tsunoda H, Kusumoto-Matsuo R, Mori S, et al. Genotype distribution of human papillomaviruses in Japanese women with abnormal cervical cytology. The Open Virology Journal. 2012; 6: 277–283.

[12] Matsumoto K, Oki A, Furuta R, Maeda H, Yasugi T, Takatsuka N, et al. Predicting the progression of cervical precursor lesions by human papillomavirus genotyping: a prospective cohort study. International Journal of Cancer. 2011; 128: 2898–2910.

[13] Gravitt PE, Peyton CL, Alessi TQ, Wheeler CM, Coutlée F, Hildesheim A, et al. Improved amplification of genital human papillomaviruses. Journal of Clinical Microbiology. 2000; 38: 357–361.

[14] Pimple SA, Mishra GA. Global strategies for cervical cancer prevention and screening. Minerva Ginecologica. 2019; 71: 313–320.

[15] Bhatla N, Singhal S. Primary HPV screening for cervical cancer. Best Practice & Research Clinical Obstetrics & Gynaecology. 2020; 65: 98–108.

[16] Bedell SL, Goldstein LS, Goldstein AR, Goldstein AT. Cervical cancer screening: past, present, and future. Sexual Medicine Reviews. 2020; 8: 28–37.

[17] Zhou Q, Gong Y, Qiu X, Sui L, Zhang H, Wang Y, et al. Visual appearance of the uterine cervix differs on the basis of HPV type status in high-grade squamous intraepithelial lesion: the results of a reliable method. BMC Women’s Health. 2022; 22: 24.

[18] Rema PN, Mathew A, Thomas S. Performance of colposcopic scoring by modified international federation of cervical pathology and colposcopy terminology for diagnosing cervical intraepithelial neoplasia in a low-resource setting. South Asian Journal of Cancer. 2019; 08: 218–220.

[19] Scheungraber C, Koenig U, Fechtel B, Kuehne-Heid R, Duerst M, Schneider A. The colposcopic feature ridge sign is associated with the presence of cervical intraepithelial neoplasia 2/3 and human papillomavirus 16 in young women. Journal of Lower Genital Tract Disease. 2009; 13: 13–16.

[20] Zaal A, Louwers JA, Berkhof J, Kocken M, Ter Harmsel WA, Graziosi GC, et al. Agreement between colposcopic impression and histological diagnosis among human papillomavirus type 16-positive women: a clinical trial using dynamic spectral imaging colposcopy. BJOG: an International Journal of Obstetrics & Gynaecology. 2012; 119: 537–544.

[21] Sone K, Toyohara Y, Taguchi A, Miyamoto Y, Tanikawa M, Uchino-Mori M, et al. Application of artificial intelligence in gynecologic malignancies: a review. Journal of Obstetrics and Gynaecology Research. 2021; 47: 2577–2585.