Breast Cancer Research
- Sophie Mao
- Dec 18, 2021
- 3 min read
INTRODUCTION
Breast Cancer is one of the most common cancers found in women. There are over 200,000 new cases in the US per year. While mammograms can help detect this cancer, it is important to be able to identify it as early as possible before it spreads. This research paper is written to help identify the relationship between certain breast characteristics and breast cancer, so that people are able to detect this cancer at its early stages.
PROCEDURE
The following research was done based on data from the Breast Cancer Wisconsin (Diagnostic) Data Set published by the University of California Irvine. This CSV data set was analyzed using Tableau Public in the span of two months. Firstly, decisions were made about which breast characteristics were to be analyzed and which data visualization layouts were to be used. Then, data visualizations were eventually created following these decisions.
DATA
This CSV was created by collecting data from 569 people. Each person had to provide their ID--a number used to identify themselves with--and their diagnosis. Research was then conducted on each patient and information on their breasts’ radius, texture, perimeter, area, smoothness, compactness, concavity, concave points, symmetry, and fractal dimension. All data is rounded to four decimal points, and for each breast characteristic the mean, standard error (SE) , and average of the three largest values (worst) was provided.
Link to Data Set: https://www.kaggle.com/uciml/breast-cancer-wisconsin-data
RESULTS
From this data, data visualizations were created in hopes to find relationships and trends between certain breast characteristics and diagnoses. The following data visualizations were made:
1. Average Breast Compactness vs. Diagnosis
From this data visualization it is clear that there is a relationship between a person’s breast compactness and whether or not the person has breast cancer. The average compactness of a breast that contains benign tumors (B) is significantly less than the compactness of a breast that contains malignant tumors (M).
2. Area vs. Diagnosis
The next data visualization created was between the breast area and diagnosis. The SE is listed above each column. But even with the standard error taken into consideration, the average area for benign breast tumors and malignant breast tumors are still far off. From this data visualization, it is clear that larger breasts tend to have malignant tumors.
3. Symmetry vs. Diagnosis
Breast symmetry in this data set refers to the size difference between a person’s breasts measured in a percentage. This data visualization is trying to find the relationship between breast symmetry and cancer. The blue columns are the average breast symmetry for that cancer diagnosis, and the yellow are the average of the three largest values of breast symmetry for that diagnosis (worst). The SE is listed above each column. Even without taking the standard error into consideration, the difference in breast symmetry between the two diagnoses is small and taking the standard error into consideration means there would be practically no difference at all. These findings show that there is not a very distinct relationship between breast symmetry and breast cancer.
CONCLUSION
From the three data visualizations, the following conclusions are able to be drawn: larger breasts tend to have a higher chance of containing malignant tumors, more compact breasts tend to have a higher chance of containing malignant tumors, and symmetry between breasts does not seem to have much of a relationship with breast cancer. It is important to identify breast cancer early to increase chances of survival and have more treatment options, which is why being regularly tested and knowing what breast characteristics are more commonly associated with malignant tumors is important.
Bình luận