A model to predict breast cancer survivability using logistic regression

Can we use logistic regression to predict breast cancer?

Because of the fact that the detection of breast cancer and prediction of the breast cancer level is important, numerous researches have been conducted in this area. These include prediction using logistic regression. Logistic regression is used for prediction by fitting data to the logistic curve.

How many samples are needed to create a logistic regression model?

Using the coefficient from 130 samples, the logistic regression model is created. The model is used to predict the data of 46 samples. One of the observed uncertain of breast cancer is predicted as positive of breast cancer. The entire observed negatives of breast cancer are predicted as uncertain of breast cancer.

What are the independent variables in a breast cancer assessment?

The independent variables are patient menopause (H1), first degree relative with breast cancer (H2), family member with other cancer (H3), patient’s previous history of breast trauma (H4), the presence of mass (M1), architectural distortion (M2), skin thickening (M3), and the presence of the calcification (M4).

What is breasts cancer?

Breast cancer is a disease in which the healthy cells of the tissue in the breast are invaded and mutated, which further grow in large numbers to form a malignant tumor. It can most likely occur at any age. Few risk factors that contribute to breast cancer.


Abstract and Figures

Background: Breast cancer is the most common type of cancer amongst women worldwide. Considering its high incidence, effective detection and prognosis of this type of cancer may have a significant effect on reducing expenditures.


References (16)

ResearchGate has not been able to resolve any citations for this publication.


How does early detection of breast cancer affect survival?

Early detection increases the survival chances by almost 99%. By realizing the symptoms and signs, this cancer can be detected early. Self-examination and clinical examination can be conducted for detection of lumps. Mammography is a technique where the breast is exposed to a small dose of radiation to capture the images of the affected area. It can detect cancer before the formation of lumps.


What are the evaluation metrics used in logistic regression?

The evaluation metrics used are accuracy, ROC, confusion matrix, precis ion-recall.


What is the proportion of predicted positives that is actually positive?

Precision: The proportion of predicted Positives that is actually positive : TP / (T P + FP)


What is negative correlation?

Negative correlated features represent cases when the value of one variable increases, the value of the other variable tends to decrease producing a downward slope.


What is a positive correlated feature?

Positively correlated features are the ones where the value of one variable increases as the other increases resulting in an upward slope indicating a positive correlation with all the data points falling on the line. Following are positively correlated features.


What are the risk factors for breast cancer?

Few risk factors that contribute to breast cancer. Genetic factors such as mother, father, sister or brother had been diagnosed with ovarian or breast cancer. · Early puberty, late menopause or child birth at an older age or no child birth can increase the risk.


Why is breast cancer detected so late?

Dense breast tissue can also surge the risk of breast cancer. Detection can also be delayed due to the compact nature of the tissue.


What are the variables in a logistic regression analysis?

Logistic regression analysis was performed using the variables from the mammogram results which are mass, architectural distortion, skin thickening, and calcification. A patient with mass detected on mammogram screening has probability of five times higher in getting breast cancer. Patients with architectural distortion or skin thickening has high probability of being afflicted with breast cancer. Also for patient with calcification detected, the probability of getting breast cancer is 18 times higher. Thus, a patient having any of the symptoms or a combination of these symptoms has greater probability of getting breast cancer. The study can assist radiologists to correctly diagnose breast cancer from using mammograms and referring to the patients’ history.


How does mass screening affect breast cancer?

The implementation of mass screening would result in increased caseloads for radiologists. This will increase chances of improper diagnosis. The prediction using logistic regression would aid the radiologist to detect the breast cancer.

Leave a Comment