92

Blood Cell Detection

Development of an automated system for the detection and characterization of cancerous cells in blood samples using image processing and machine learning techniques.

GitHub

The project was developed using data of blood cells affected by acute lymphoblastic leukemia (ALL) from Kaggle. The main objective of this project is to develop a system capable of automatically detecting cancerous cells in blood samples to facilitate the early diagnosis of blood cancer.

Key Features

  • Image Preprocessing: Cleaning and preparing blood images to improve detection accuracy.
  • Segmentation: Identifying and isolating individual cells in blood samples.
  • Feature Extraction: Analyzing cell properties to differentiate cancerous cells from healthy cells.
  • Classification: Using machine learning algorithms such as KNN, SVM, Decision Tree, and Random Forest to classify cells.
  • Evaluation: Measuring the precision, recall, and specificity of the models used.

Technologies Used

  • Programming Languages: Python
  • Libraries: OpenCV, scikit-learn, NumPy, Pandas, Matplotlib
  • Tools: Jupyter Notebook, Kaggle

Results and Performance

  • Models Tested: KNN, SVM, Decision Tree, Random Forest
  • Best Performance:
ModelPrecision
Random Forest92.9%
KNN92.6%
SVM93.1%
Decision Tree88.9%

Web Application

  • Frontend: HTML, CSS, JavaScript
  • Backend: Flask

Future Perspectives

  • Integrating deep learning techniques to improve detection accuracy.
  • Validating the model with additional blood samples to enhance the system's robustness and reliability.
  • Use dynamic blood cell images to improve the system's adaptability to different scenarios.

Contact

If you have any questions or inquiries, feel free to contact me:

I look forward to hearing your feedback and suggestions!