Unveiling Breast Cancer Insights: Empowering Diagnosis with Data

Posted on Jun 7, 2023

Breast cancer is a complex and widely studied disease for good reason. Affecting millions, it is the most common cancer among women in the US. A breast cancer diagnosis can be overwhelming, but a user-friendly app can help patients become more informed and feel more in control while leading clinicians to more data-driven decisions that could improve patient outcomes.


This project aims to present an interactive analysis, data exploration, and application spin-off of a breast cancer data set. The data set contains measurements of cell nuclei extracted during a Fine Needle Aspiration biopsy conducted on 569 women patients. The Shiny app developed provides users with an intuitive interface to explore the dataset interactively. Users can browse through various variables and select specific attributes to investigate. Through dynamic visualizations, users can understand how different parameters contribute to a breast cancer diagnosis. The app also allows users to perform logistic regression to predict diagnostic outcomes.

Understanding Fine Needle Aspiration Measurements

Fine Needle Aspiration (FNA) is a standard diagnostic technique for breast cancer detection. It is a biopsy procedure in which a small quantity of breast tissue or fluid is extracted from an area of concern using a thin, hollow needle. This procedure aims to examine the sample for the presence of cancer cells. Traditionally, once the cells are removed, the sample is assessed by a pathologist or cytologist who provides the diagnostic results.

Exploring the Dataset:

The features in this data set are computed from digitized images of a fine needle aspiration (FNA) of a breast mass. The measurements capture essential information about the cell's nuclei characteristics, such as size, shape, and symmetry. The data set contains 10 attributes of the cell nuclei:

  1. Radius
  2.  Texture
  3.  Perimeter
  4.  Area
  5.  Smoothness
  6.  Compactness
  7.  Concavity
  8.  Concave points
  9.  Symmetry
  10.  Fractal dimension

This dataset analysis seeks to uncover significant relationships between these measurements and diagnostic outcomes.

Visualizing Data Patterns

Visualizations play a crucial role in data exploration. The app offers various visual tools, including scatter plots, histograms, and box plots, to help users identify patterns, outliers, and potential relationships within the dataset. These visual representations enable users to grasp the complexities of breast cancer measurements and their impact on diagnosis.

To make the app more interactive, the user can choose the parameter to be explored, the number of bins in the histogram, and select either a box plot or a violin plot to display.

Observing the plots of each variable helps understand how they relate to the diagnosis. For example, the 'area_mean' parameter above shows that benign masses are smaller and narrowly distributed, unlike malignant masses, which are larger and broadly distributed. Other parameters like 'fractal_dimension_mean' do not show a significant difference between benign and malignant mass sizes.

Another way to leverage visualizations is to investigate the relationships between two variables. For example, the app allows the user to see that as 'perimeter_mean' increases, so does 'concavity_mean.'

However, not all relationships between independent variables are desirable. Users can explore the correlation between independent variables using a correlation matrix. In a linear model, the correlation between independent variables is not desired because it can lead to wildly varying and possibly numerically unstable solutions. As the same information is contained within the two variables, including only one of them suffices. For example, we can see from the correlation matrix that 'perimiter_mean' and 'radius_mean' are highly correlated. Therefore, we only need to include one of them in the final model.

Predictive Modeling with Logistic Regression

One of the most exciting features of the app is the ability to perform logistic regression. Logistic regression is a powerful statistical technique used to predict binary outcomes, in this case, predict if the mass found in the breast is benign or malignant. The output of the regression is displayed on the right-hand side.

By selecting relevant variables and applying logistic regression algorithms, users can generate predictive models that estimate the likelihood of the diagnosis based on the FNA measurements. This empowers clinicians and researchers to make informed decisions and tailor treatment plans accordingly. Users are encouraged to interact with the app by selecting different variables and observing how they affect the final logistic regression model.

Based on these results, the best variables to use as predictors are:

  • texture_mean
  •  area_mean
  •  smoothness_mean
  •  concave.points_mean

Using these 4 predictor variables, this model has an accuracy of 95.57%. This accuracy was calculated on the test data. The data for this project was divided into 80% training data and 20% testing data.

The parameters selected for the final model depend on a few factors, correlation, as explained previously, VIF, AIC, and p-values as described below:

The Variance Inflation Factor (VIF) is used in regression analysis to gauge the level of multicollinearity present. Multicollinearity refers to the correlation between independent variables in a model. Generally, a VIF exceeding 5 or 10 is considered significant, suggesting difficulty in estimating the coefficient. However, it's important to note that high VIF values do not necessarily undermine the model's predictive quality. Detecting multicollinearity is crucial because, while it doesn't diminish the model's explanatory ability, it reduces the independent variables' statistical significance.

The Akaike Information Criterion (AIC) is a numerical score that helps select the most suitable model among several options for a given dataset. It assesses models in relation to one another, so AIC scores are meaningful only when comparing different models on the same dataset. A lower AIC score indicates a better-performing model.

A p-value under 5% shows that the independent variable is statistically significant in explaining the target variable. The final model should include only independent variables that affect the dependent variable.

The user is also encouraged to select different combinations of independent variables and observe how the selections affect the probability plot.

Empowering Research and Clinical Decision-Making

This interactive app is designed to assist researchers and empower clinicians in their decision-making process. By incorporating a user interface where clinicians can input a patient's FNA parameters information, the app has the potential to contribute to improved diagnostic accuracy.


With the interactive R Shiny app, users are invited to embark on a journey of breast cancer analysis. Users can gain valuable insights into breast mass diagnosis by exploring the dataset of fine needle aspiration measurements, visualizing data patterns, and utilizing logistic regression. This app aims to support clinical decision-making and improve patient outcomes in the fight against breast cancer.

About Author


Neiza is a data scientist passionate about uncovering the story behind the data. Neiza worked 10 years as a CPU designer at Intel. She earned a Bachelor and a Master of Science in Electrical Engineering from the University...
View all posts by neizalazo >

Leave a Comment

No comments found.

View Posts by Categories

Our Recent Popular Posts

View Posts by Tags

#python #trainwithnycdsa 2019 2020 Revenue 3-points agriculture air quality airbnb airline alcohol Alex Baransky algorithm alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus ames dataset ames housing dataset apartment rent API Application artist aws bank loans beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep boston safety Bundles cake recipe California Cancer Research capstone car price Career Career Day citibike classic cars classpass clustering Coding Course Demo Course Report covid 19 credit credit card crime frequency crops D3.js data data analysis Data Analyst data analytics data for tripadvisor reviews data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization database Deep Learning Demo Day Discount disney dplyr drug data e-commerce economy employee employee burnout employer networking environment feature engineering Finance Financial Data Science fitness studio Flask flight delay gbm Get Hired ggplot2 googleVis H20 Hadoop hallmark holiday movie happiness healthcare frauds higgs boson Hiring hiring partner events Hiring Partners hotels housing housing data housing predictions housing price hy-vee Income Industry Experts Injuries Instructor Blog Instructor Interview insurance italki Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter las vegas airport lasso regression Lead Data Scienctist Lead Data Scientist leaflet league linear regression Logistic Regression machine learning Maps market matplotlib Medical Research Meet the team meetup methal health miami beach movie music Napoli NBA netflix Networking neural network Neural networks New Courses NHL nlp NYC NYC Data Science nyc data science academy NYC Open Data nyc property NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time performance phoenix pollutants Portfolio Development precision measurement prediction Prework Programming public safety PwC python Python Data Analysis python machine learning python scrapy python web scraping python webscraping Python Workshop R R Data Analysis R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn seafood type Selenium sentiment analysis sentiment classification Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau teachers team team performance TensorFlow Testimonial tf-idf Top Data Science Bootcamp Top manufacturing companies Transfers tweets twitter videos visualization wallstreet wallstreetbets web scraping Weekend Course What to expect whiskey whiskeyadvocate wildfire word cloud word2vec XGBoost yelp youtube trending ZORI