The Breast Cancer Diagnosis App uses Machine Learning to predict whether a breast cancer tumor is malignant or benign based on cytological characteristics.
- You can access the app here
- This app was made for a webinar and a workshop about Streamlit
The Breast Cancer Wisconsin dataset is a widely-used dataset in the field of Machine Learning and medical research. It originates from the University of Wisconsin-Madison and was created by Dr. William H. Wolberg.
The dataset is designed to help develop predictive models for diagnosing breast cancer based on cytological characteristics of fine needle aspirate (FNA) cytology samples from breast masses.
- The dataset consists of 569 instances and 32 attributes. The key attributes include ID number and Diagnosis: This indicates whether the tumor is benign (B) or malignant (M).
- The remaining 30 features are computed from the FNA images and describe various characteristics of the cell nuclei present in the images.
The app focuses on five specific features and translates them into their corresponding cytological terms based on the Yokohama System for Reporting Breast Cytopathology:
- Marked Nuclear Indentation (Worst Concave Points): Refers to the most significant indentations in the nuclear membrane, which is a typical feature in malignant cells.
- Irregular Nuclear Membrane (Worst Perimeter): Indicates the irregularity in the shape of the nuclear membrane, often associated with cancerous cells.
- Variability in Nuclear Membrane Smoothness (Smoothness Error): Represents the variation in the smoothness of the nuclear membrane, which can indicate abnormal cell growth.
- Increased Nuclear Area (Worst Area): Larger nuclear area is often seen in malignant cells as they tend to have larger nuclei.
- Nuclear Indentations (Mean Concave Points): Refers to the average number of indentations in the nuclear membrane, which can be a sign of malignancy.
To run the app locally, follow these steps:
- Clone the repository:
git clone https://github.com/isi-mube/breast-cancer-app.git cd breast-cancer-app