{"id":3473,"date":"2025-08-31T11:12:59","date_gmt":"2025-08-31T09:12:59","guid":{"rendered":"https:\/\/neuraldesigner.com\/learning\/cervical-cancer-prognosis\/"},"modified":"2026-02-11T14:17:37","modified_gmt":"2026-02-11T13:17:37","slug":"cervical-cancer-prognosis","status":"publish","type":"learning","link":"https:\/\/www.neuraldesigner.com\/learning\/examples\/cervical-cancer-prognosis\/","title":{"rendered":"Cervical cancer prognosis using machine learning"},"content":{"rendered":"\t\t
\n\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t
<\/section>
Early detection of cervical cancer is crucial for increasing survival rates, minimizing treatment intensity, and reducing mortality and healthcare costs. Here, we create a prognosis model based on machine learning that focuses on the initial screening stages of cervical cancer. This prognosis model will aid in decision-making in daily clinical practice.<\/section>
\u00a0<\/section>

To build the model, we use data collected by the Cervical Pathology Unit in the health area of Palencia (Spain) from women participating in a cervical cancer early detection program.<\/p><\/section>

We use the data science and machine learning platform Neural Designer<\/a> to solve this example.\u00a0<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t

\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\n\t\t\t\t\t\t\n\t\t\t\t<\/path><\/svg>\t\t\t<\/span>\n\t\t\t\t\t\t\t\t\tDownload<\/span>\n\t\t\t\t\t<\/span>\n\t\t\t\t\t<\/a>\n\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Contents<\/h2>
  1. Introduction<\/a>.<\/li>
  2. Application type<\/a>.<\/li>
  3. Data set<\/a>.<\/li>
  4. Neural network<\/a>.<\/li>
  5. Training strategy<\/a>.<\/li>
  6. Model selection<\/a>.<\/li>
  7. Testing analysis<\/a>.<\/li>
  8. Model deployment<\/a>.<\/li><\/ol>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t
    \n\t\t\t\t\t\t\t\t\t

    1. Introduction<\/h2>

    Cervical cancer remains one of the leading causes of cancer-related mortality among women worldwide. Therefore, improving prognosis prediction plays a key role in treatment planning and patient management.<\/p>

    In this example, we use machine learning to estimate a prognosis value based on clinical variables. Specifically, the model predicts a continuous outcome that helps stratify patients into risk groups.<\/p>

    Although predictive models provide valuable support, clinicians must always interpret results within the broader medical context.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t

    \n\t\t\t\t\t\t\t\t\t

    2. Application type<\/h2>

    The variable to be predicted is continuous (PROGNOSIS).\u00a0Therefore, this is an\u00a0approximation<\/a>\u00a0project.<\/p>

    The basic goal of this study is to model the patient\u2019s prognosis as a function of the input variables.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t

    \n\t\t\t\t\t\t\t\t\t

    3. Data set<\/h2>

    The\u00a0data set<\/a>\u00a0contains the relevant information to create the prognostic model.<\/p>

    Data source<\/h3>

    In this case, the\u00a0cervixcancer.csv<\/a>\u00a0contains the information for this study.<\/p>

    Each column represents a variable related to cervical cancer, and each row represents a patient from the Cervical Pathology Unit.<\/p>

    Data encoding and variable transformation<\/h2>

    The number of variables is 7, corresponding to the different tests for each patient. The number of cases is 197, the patients of the last three years with a definitive diagnosis. The following table shows the first 10 cases of this data set. We can observe that some missing values are represented by \u201cNA\u201d.<\/p>
    AGE<\/th>CYTOLOGY<\/th>HPV<\/th>BIOPSY<\/th>P16\/KI67<\/th>SMOKE<\/th>PROGNOSIS<\/th><\/tr>
    60<\/td>ASC-US<\/td>HPV 18<\/td>CIN I<\/td>NA<\/td>NA<\/td>CIN I<\/td><\/tr>
    22<\/td>L-SIL<\/td>HPV 16<\/td>CIN II-III<\/td>NA<\/td>NA<\/td>CIN II-III<\/td><\/tr>
    40<\/td>ASC-US<\/td>OTHER HIGH RISK<\/td>CIN II-III<\/td>+<\/td>NA<\/td>CIN II-III<\/td><\/tr>
    37<\/td>L-SIL<\/td>HPV 16<\/td>CIN I<\/td>\u2013<\/td>YES<\/td>CIN I<\/td><\/tr>
    31<\/td>L-SIL<\/td>OTHER HIGH RISK<\/td>CIN I<\/td>+<\/td>YES<\/td>NEGATIVE<\/td><\/tr>
    45<\/td>ASC-US<\/td>OTHER HIGH RISK<\/td>CIN II-III<\/td>NA<\/td>YES<\/td>CIN II-III<\/td><\/tr>
    28<\/td>ASC-US<\/td>OTHER HIGH RISK<\/td>CIN II-III<\/td>NA<\/td>YES<\/td>CIN II-III<\/td><\/tr>
    53<\/td>NORMAL<\/td>OTHER HIGH RISK<\/td>NEGATIVE<\/td>NA<\/td>YES<\/td>CIN I<\/td><\/tr>
    51<\/td>NORMAL<\/td>HPV 16<\/td>NEGATIVE<\/td>NA<\/td>NA<\/td>CIN-I<\/td><\/tr>
    42<\/td>NORMAL<\/td>HPV 16<\/td>NOT ACCURATE<\/td>NA<\/td>NA<\/td>NEGATIVE<\/td><\/tr><\/tbody><\/table>

    Neural networks work with numerical values. However, some of the variables in the table above have categorical values. Therefore, the first step is to assign numerical values to all categorical variables.<\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t

    \n\t\t\t\t\t\t\t\t\t

    Age<\/h4>

    Age is a numerical value that indicates the age of the patients. The variable AGE is numeric; therefore, it does not need modification.<\/p>

    Cytology<\/h4>

    Cytology is the primary tool for the prevention of cervical cancer. It is a diagnostic test based on the morphological study of cells. It is obtained by scraping or brushing the surface of the exocervix and endocervix, and it detects cervical cancer precursor lesions, which are ASC-US, ASC-H, L-SIL, H-SIL, and AGC.<\/p>

    The variable CYTOLOGY contains NORMAL, ASC-US, ASC-H, L-SIL, H-SIL, and AGC values. This variable is ordinal since its values are different degrees of the same manifestation, and we can order them by severity. The following table shows the numerical values assigned to the results of this test.<\/p>
    CYTOLOGY<\/th>ASSIGNED GRADES<\/th><\/tr>
    NORMAL<\/td>0<\/td><\/tr>
    ASC-US<\/td>1<\/td><\/tr>
    ASC-H<\/td>2<\/td><\/tr>
    L-SIL<\/td>3<\/td><\/tr>
    H-SIL<\/td>4<\/td><\/tr>
    AGC<\/td>5<\/td><\/tr><\/tbody><\/table>

    HPV<\/h4>

    Human papillomavirus. The HPV variable contains NEGATIVE, HPV 16, HPV 18, OTHER LOW RISK, and OTHER HIGH RISK. This variable is also considered ordinal. HPV 16 and HPV 18 represent the same risk, the maximum. The following table shows the numerical values assigned to the results of this test.<\/p>
    HPV<\/th>ASSIGNED GRADES<\/th><\/tr>
    NEGATIVE<\/td>0<\/td><\/tr>
    OTHER LOW RISK<\/td>1<\/td><\/tr>
    OTHER HIGH RISK<\/td>2<\/td><\/tr>
    HPV 16<\/td>3<\/td><\/tr>
    HPV 18<\/td>3<\/td><\/tr><\/tbody><\/table>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t
    \n\t\t\t\t\t\t\t\t\t

    Biopsy<\/h4>

    A test for a histological diagnosis of intraepithelial lesions and cervical cancer. It classifies the results as:<\/p>