黑料网

One of the most common cancers in women is cervical cancer, which will cause 604,000 new cases and 342,000 deaths by 2020. It is the only cancer that can be cured with primary prevention methods like early detection, timely treatment, and a fully effective 9-valent human papillomavirus (HPV) vaccine.

The persistent infection of the cervical epithelium with one of the 15 genotypes of the carcinogenic HPV is the root cause of nearly all cases of cervical cancer [1-3]. The four main stages of cervical cancer development are as follows: infection of the metaplastic epithelium at the cervical transformation zone, persistent HPV infection, the development of cervical precancer from persistently infected epithelium, and invasion through the epithelium's basement membrane. Girls and young women may be protected from HPV infection by receiving HPV vaccines. Yet, HPV antibody inclusion rate is extremely low as of now (even in a few created nations) and the recipients are restricted to young ladies matured <26 as far as 9-valent HPV immunizations. Because it is impossible to completely avoid risk, the American Cancer Society recommends that vaccinated women be screened in the same way as unvaccinated women. Women still place a high value on having their cervical cancers checked on a regular basis. Roughly 30% of cervical intraepithelial neoplasia (CIN) grade 3 sores form into intrusive diseases in 30 years or less. These lesions can be detected and treated in many ways thanks to their slow progression. Preventing cervical cancer at a low cost can be accomplished by screening and treating women's precancerous lesions [4]. Ideally, screening methods should be able to catch cervical cancer-causing early lesions while avoiding transient HPV infections and benign abnormalities, which can lead to overtreatment and other risks of screening. The detection rate of cervical cancer has increased, and the mortality rate has decreased, as screening methods have continued to improve; However, low- and middle-income nations account for the majority of deaths. Despite recent advancements in screening programs that work, many of them are unable to be implemented or maintained due to inadequate health infrastructure. Also, manual screening doesn't always work 100% of the time, so some related lesions can't be diagnosed right away. Therefore, the primary obstacle to early cervical cancer diagnosis is the development of a screening method that is both more accurate and less expensive.

AI has demonstrated promising application value in the diagnosis of various diseases in recent years, including the imaging diagnosis of tumors, the classification of skin tumors, and the diagnosis and classification of retinal diseases. Complex algorithms are used by AI to automatically recognize images, extract features, learn classification, and process data. Utilizing artificial intelligence (AI) in the early screening and diagnosis of cervical cancer is advantageous for addressing the issue of limited human resources and increasing diagnostic precision. The goal of this article was to talk about the most recent AI technologies and show how they can be used to screen for and catch cervical cancer early. Additionally, this review offers suggestions for future research and discusses the issues that need to be addressed.

Two Methods for Cervical Cancer Screening and Diagnosis The following three screening methods are recommended by the most recent World Health Organization guidelines for the early detection of cervical cancer: HPV testing, cytology (counting customary pap smear and fluid-based cytology smear), and visual investigation with acidic corrosive (Through). Since VIA is only utilized when the other two are unavailable, we concentrated on the first two methods. Test samples for HPV testing and cytology are brushed, exfoliated cells taken from the cervix. HPV testing distinguishes high-risk sorts of HPV disease in the cervix, while cytological assessment utilizes a magnifying lens to recognize cells taken from the cervix for conceivable cervical malignant growth or precancerous sores, shows the development of cervical malignant growth screening strategies. Be that as it may, colposcopydirected biopsy stays as the highest quality level for cervical malignant growth analysis, trailed by arranging as indicated by the clinical assessment and imaging results.

Introduction to Cytology A conventional Pap smear (CPS) is a manual screening procedure that uses a microscope to identify and classify exfoliated cervical cells by color and cytoplasmic and nucleus characteristics (18). Fluid based cytology (LBC) can further develop planning procedures . The LBC specimen is more uniformly distributed than the CPS specimen, is easier to preserve and remove artificially, and is better fixed in glass slides.

Introduction to Colposcopy is the use of a specific instrument to magnify the fully exposed cervix by 5 to 40 times in real time for a visual evaluation of the cervix, particularly the transformation area, to detect CIN or squamous intraepithelial lesion (SIL) and invasive cancer. A colposcopy-directed biopsy of the thought site is performed to decide if further treatment, like conization or cryotherapy, is required, which is significant in patients with high-grade CIN or more extreme illness .

Methodology for early screening and finding of cervical malignant growth

As indicated by the most recent suggestions of the American Malignant growth Society on cervical disease screening, ladies with a cervix matured ≥25 years are prescribed to go through cervical disease screening. A primary HPV test should be given to women between the ages of 25 and 65 every five years. Co-testing (HPV testing combined with cytology) or cytology evaluation can be performed every three years if a primary HPV test is unavailable.

Application of AI in Early Cervical Cancer Screening HPV Typing and Detection Consistent high-risk HPV infection can result in cervical cancer [5-7]. HPV testing can distinguish HPV disease and assist with screening high-risk populaces. Women who have HPV DNA-positive results or positive cervical smear results will be able to assess their risk of developing cervical cancer more easily thanks to HPV genotyping. Artificial intelligence learning innovation utilizes research connected with HPV testing to further develop exactness and broaden the utilization of HPV testing in cervical malignant growth screening.

Cervical cytology screening programs that are based on cytology have reduced the incidence of cervical cancer in many Western nations. Cytology screening for high-grade cervical precancerous lesions has a high specificity, but it has a lower sensitivity (50–70%) and requires careful microscopy observation by experienced cytologists. Each procedure is time-consuming, labor-intensive, and prone to error. Additionally, there is a low level of cytological reproducibility, resulting in low accuracy. Additionally, changing the observer results in subjective and inconsistent outcomes. As a result, the researchers hope to create automated image analysis techniques to alleviate these strains.

In 1992, PAPNET was the first commercial automatic screening system. The framework was supported as a technique for re-evaluating slides that were passed judgment on negative by cytologists [8]. The thin prep imaging system was approved by FAD in 2004 as a commercial screening product. The proprietary algorithm enables the system to select the 22 most important fields of view (FOVs), and cytotechnologists must manually screen the entire slide if abnormal cells are detected. The framework works on the awareness and effectiveness of screening. In 2008, the Focal point GS imaging system was developed later. It divided the risk into 10 FOVs of cervical cells that were most likely to be abnormal to increase efficiency.

Segmentation of Cervical Cells The following are the five stages of a typical automatic smear analysis system: picture securing, preprocessing, division, include extraction, and characterization. For the automatic analysis of a smear, AI technology is used in the segmentation and classification stages, which helps increase screening efficiency.

The most vital phase in cytological conclusion is the exact recognizable proof of cells and their primary parts. Accurate segmentation is a requirement for screening solutions since the diagnostic criteria for cervical cytology are primarily based on abnormalities in the nuclear and cytoplasm [9]. Hepatoma cells, human metaphase II oocytes, and pluripotent stem cells have all been successfully segmented using AI, according to ongoing research on its use in cell segmentation. It has also been made available for the purpose of automatically segmenting cervical cells, and positive outcomes have been reported. For instance, Chankong et al. used fuzzy c-means clustering to create whole-cell segmentation by dividing single-cell images into the nucleus, cytoplasm, and background. Using supervised learning, some researchers were able to segment the cells' overlapped cytoplasm in images from cervical smears by extracting adaptive shape from cytoplasmic contour fragments and shape statistics. The results of the experiments indicate that this method is always more effective than the most advanced ones. Division model on pictures from pap smear slide likewise was investigated which has been accomplished through utilizing core limitation to order typical and strange cells .

Conclusion

AI can be used to treat, predict prognosis, and prevent cervical cancer, in addition to the early screening and diagnosis discussed in this paper. For better treatment decision-making, more research on treatment and prediction is required in the future. This will work with cervical disease annihilation programs around the world. Moreover, as the occurrence of cervical adenocarcinoma and other uncommon neurotic sorts expands, simulated intelligence ought to be applied for the early conclusion of such sicknesses later. Additionally, AI can be used to noninvasively distinguish cervical cancer from other diseases. The prediction of cervical cancer, enhancements to cervical cancer screening and diagnosis, optimization of staging systems, and improved patient prognosis will all be significantly enhanced by further development of AI technologies.

1. Schüler S, Ponnath M, Engel J, Ortmann O (2013) . Arch Gynecol Obstet 287:1187-1204.

       , ,

2. Franceschi S, La Vecchia C, Negri E, Guarneri S, Montella M ,et al. (1994) . Hum Reprod 9:1673-1675. 

       , ,

3. Cusidó M, Fábregas R, Pere BS, Escayola C, Barri PN (2007) . Gynecol Endocrinol 23:373-376.

       , ,

4. Leibowitz D, Hoffman J (2000) . J Obstet Gynecol Neonatal Nurs.29:201-210. 

              , ,

5. Holzer H, Casper R, Tulandi T (2006) . Fertil Steril85:277-284.

       , ,

6. Gips H, Hormel P, Hinz V (1996) . Andrologia 28:3-7. 

               , ,

7. Elias RT, Pereira N, Palermo GD (2017) . J Assist Reprod Genet 34:1233.

       , ,

8. Karakji EG, Tsang BK (1995) Regulation of rat granulosa cell plasminogen activator system: Influence of interleukin-1 beta and ovarian follicular development. Biol Reprod.53:1302-1310. 

       , ,

9. Shapiro BS, Daneshmand ST, Restrepo H, Garner FC, Aguirre M, et al. (2011) . Fertil Steril.   95:826–8.