27,558 segmented cell images categorized into parasitized and uninfected classes for malaria detection. The collection includes images derived from Giemsa-stained thin blood smear slides of 150 infected and 50 healthy patients.
Use Cases
- Train binary classification models to distinguish between 'Parasitized' and 'Uninfected' cell labels
- Benchmark deep learning architectures on medical image classification tasks using the provided PNG image files
- Implement data augmentation techniques to improve model generalization across varying staining intensities in blood smear images
Strengths
- 27,558 total images across two distinct labels: Parasitized and Uninfected
- Sourced from Giemsa-stained thin blood smear slides
- Balanced distribution with 13,779 images per class