Role of artificial intelligence in detecting cervical cancer
- Conditions
- Malignant neoplasm of cervix uteri, unspecified,
- Registration Number
- CTRI/2021/03/032299
- Lead Sponsor
- Ayesha Siddiqua
- Brief Summary
Cancer of the uterinecervix is the second most common cancer among Indian women, and accounts for25% of global mortality due to cervical cancer. This is due to higherincidence, and late diagnosis of cervical cancer. Some of the commonly used screening testsfor early detection of cervical cancer are Pap smear, HPV-DNA and visualinspection with acetic acid (VIA). Pap smear is the most commonly used methodfor cervical cancer screening. It depends on availability of laboratoryfacility and medical experts for smear preparation and analysis, which islacking in developing countries. VIA evaluates the cervix for acetowhite (AW)regions which are formed when precancerous lesions combine with acetic acid. Itdoes not depend on well-established laboratory facility and hence is acost-effective test. It is a simple test as it is based on visual examinationof the cervix after the application of acetic acid. Thus, it is well suited forscreening in developing countries.VIA is a suitabletest for low-resource settings as far as the simplicity of the test isconsidered, it requires skilled personnel for the evaluation of the acetowhiteregions. Depending on the skills of the ones who perform the test, accuracy ofVIA varies widely, with significantly higher sensitivity and specificity whenobserved by physicians than by nurses . An automated system for cervix image analysis might be able toovercome this problem, providing objective and repeatable evaluations. Advancesin digital imaging and image processing have facilitated the acquisition ofgood-quality cervix images during the VIA procedure. At present, digital imagesof the cervix are acquired using a digital camera or a digital colposcope.Acquisition of cervix images using the digital camera and subsequent analysisof the images by skilled personnel is called cervicography. Since images acquiredcan be stored and sent over the Internet, it enables remote diagnosis. Digitalcolposcopy facilitates the acquisition of high-magnification cervix images.However, it is expensive and hence is not feasible for screening inresource-poor settings. An android device with an inbuilt app to acquire imagesand provide instant results would be an obvious choice in resource-poorsettings. Our overall goal is to analyse sensitivity, specificity, positive predictive value, negative predictivevalue, accuracy and ROC of Easy CervicalOptimiser(EC0) device which facilitatesacquisition and processing of images with an inbuilt image processingalgorithm. Such a system would be useful to augment the capability of a layhealth worker performing the VIA examination to offer instant decision whetherthe individual requires further evaluation or not.
- Detailed Description
Not available
Recruitment & Eligibility
- Status
- Not Yet Recruiting
- Sex
- Female
- Target Recruitment
- 326
1.Women between 30- 65 years who are Subjected to routine cervical screening 2.Includes those scheduled for colposcopic evaluation for indications- -Positive Pap test (ASCUS/ASC-H/LSIL/HSIL/Invasive -carcinoma (Bethesda classification) -HR-HPV positive -Clinically suspicious cervix.
1.Obvious growth on cervix (Ca cervix) 2.Unmarried woman 3.Pregnant/ postpartum 4.Menstruating on the day of the visit 5.Genital prolapse grade 2 and beyond.
Study & Design
- Study Type
- Observational
- Study Design
- Not specified
- Primary Outcome Measures
Name Time Method 1.Comparison of VIA –AI using ECO device with colposcopy diagnosis in identification of cervical epithelial abnormalities for screen positive women 4 WEEKS 2.Comparison of VIA –AI using ECO device with histopathology reports 4 WEEKS 3.Comparison of VIA –AI using ECO device with screen test (Pap smear/ HPV) results 4 WEEKS
- Secondary Outcome Measures
Name Time Method 1.Analysis of sensitivity,specifivity of VIA-AI using ECO device with standard screening procedures of carcinoma cervix 1 YEAR
Trial Locations
- Locations (1)
Kasturba Hospital,Department of OBG,KMC Manipal
🇮🇳Udupi, KARNATAKA, India
Kasturba Hospital,Department of OBG,KMC Manipal🇮🇳Udupi, KARNATAKA, IndiaAyesha SiddiquaPrincipal investigator8790509701ayesha.siddiqua@learner.manipal.edu