A Cervical Cancer Prediction Model Using REPTree Classifier

Abstract

<jats:p>Cervical cancer is the foremost gynecological disease globally. In this manuscript, we build up a Cervical Cancer prediction model that can aid medical experts in envisaging Cervical Cancer condition based on the clinical data of patients. At the outset, we choose 32 imperative clinical attributes viz., age, hormonal contraceptives, number of sexual partners, STDs: AIDS, first sexual intercourse (age), STDs: HIV, number of pregnancies, STDs: Hepatitis B, smokes etc., in addition to four classes (Hinselmann, Schiller, Cytology and Biopsy). Secondly, we build up a prediction model by means of REPTree classifier for classifying Cervical Cancer based on these clinical attributes against unpruned, and pruned error pruning approach. As a final point, it is concluded that the precision of unpruned REPTree classifier with Pruned REPTree classifier approach is better than the Pruned REPTree classifier approach. The outcome acquired that which illustrates that age, hormonal contraceptives, first sexual intercourse (age), STDs: genital herpes, number of pregnancies and smokes are the foremost predictive attributes which provides enhanced classification in opposition to the supplementary attributes.</jats:p>

Diagnosis of Cervical Cancer and Pre-Cancerous Lesions by Artificial Intelligence: A Systematic Review

“…Multivariate adjustment and multiple-regression techniques were introduced for prediction (that is, for estimating the predicted value of a certain outcome as a function of given values of independent variables) [ 82 ]. AI studies using machine learning principles have focused on algorithms to predict cervical cancer [ 55 , 56 , 57 , 58 , 59 , 60 , 61 , 62 , 63 , 83 , 84 , 85 , 86 , 87 , 88 , 89 , 90 , 91 , 92 , 93 , 94 , 95 ]. The most important predictors of cervical cancer were age, age at first sexual intercourse, number of sexual partners, pregnancies, smoking, period of smoking (years), hormonal contraceptives, period of use of hormonal contraceptives (years), IUD, period of use of IUD (years), STDs, period of STDs (years), Schiller, Hinselmann, cytology, the presence of 15 high-risk HPV genotypes [ 55 , 56 , 57 , 58 , 60 , 84 ], social status, marital status, personal health level, education level, and the number of caesarean deliveries [ 63 ].…”

“…AI studies using machine learning principles have focused on algorithms to predict cervical cancer [ 55 , 56 , 57 , 58 , 59 , 60 , 61 , 62 , 63 , 83 , 84 , 85 , 86 , 87 , 88 , 89 , 90 , 91 , 92 , 93 , 94 , 95 ]. The most important predictors of cervical cancer were age, age at first sexual intercourse, number of sexual partners, pregnancies, smoking, period of smoking (years), hormonal contraceptives, period of use of hormonal contraceptives (years), IUD, period of use of IUD (years), STDs, period of STDs (years), Schiller, Hinselmann, cytology, the presence of 15 high-risk HPV genotypes [ 55 , 56 , 57 , 58 , 60 , 84 ], social status, marital status, personal health level, education level, and the number of caesarean deliveries [ 63 ]. However, according to Mudawi et al, certain characteristics of the patient samples, including the quantity of alcohol consumed and the presence of HIV and HSV2, could not be regarded as reliable predictors [ 95 ].…”

“…However, according to Mudawi et al, certain characteristics of the patient samples, including the quantity of alcohol consumed and the presence of HIV and HSV2, could not be regarded as reliable predictors [ 95 ]. Based on the data shown in Table 1 , the accuracy of the algorithms in predicting cervical cancer varied from 70% to 100% [ 55 , 56 , 57 , 58 , 59 , 60 , 62 , 83 , 84 , 85 , 86 , 87 , 88 , 89 , 90 , 91 , 92 , 93 , 94 , 95 ]. The application of AI for the prediction of cervical cancer is shown in Table 1 .…”