SKDrepaData: A Multisite West African Blood Smear Dataset for Automated Sickle Cell Disease Detection and Classification

Domaine:

healthcare

Type de record:

dataset

Créateur:

Éditeur:

Hôte:

SKDrepaData is a curated microscopy dataset of peripheral blood smears from sickle cell disease (SCD) patients, collected in situ in West Africa and annotated in collaboration with expert hematologists. The dataset comprises 1,861 full-field microscopy images acquired from more than 100 patients across three clinical sites, using a standardized May-Grunwald-Giemsa (MGG) staining protocol and three distinct optical microscopes at x40 and x100 (oil-immersion) magnifications. It contains 33,832 annotated red blood cells, nearly evenly distributed across three morphological classes: normal (11,106), sickle (11,473), and other abnormalities (11,253). Annotations were cross-validated by five independent experts, reaching a mean Intersection over Union (IoU) of 0.85 for detection and a Fleiss' kappa of 0.90 for classification, both indicating near-perfect agreement. The dataset is distributed as three complementary subsets:- SKDrepaData-v1: full-field images with YOLO-format bounding-box annotations (detection)- SKDrepaData-v2: 224x224 cell patches organized by class (classification)- SKDrepaData-v3: masks and morphological descriptors (quantitative analysis) Unlike existing public resources, SKDrepaData uniquely combines a substantial multi-site volume, standardized clinical-grade staining, expert-validated annotations with quantified reliability, and joint support for detection and classification, while being explicitly grounded in a West African clinical context, where the SCD burden is highest. See README.md for full documentation, dataset structure, and usage examples. Ethics: Data collection was approved by the relevant institutional ethics committee, with informed consent obtained from all participants or their legal guardians. All images are fully anonymized.

Visit

doi.org

Tasks

computer vision image classification

Licenses

Similaires

SKDrepaData-v2: Microscopic images of red blood cells for sickle cell disease research.Optimal Machine Learning Based Automated Malaria Parasite Detection and Classification Model Using Blood Smear ImagesRed Blood Cell Morphology Dataset for Image ClassificationDeveloping a prediction model for early detection and classification of kidney dysfunction in children with sickle cell diseaseMorphology-based classification of sickle cell disease and β-thalassemia using a low-cost automated microscope and machine learningImage Segmentation and Classification for Sickle Cell Disease using Deformable U-Net

SKDrepaData-v2: Microscopic images of red blood cells for sickle cell disease research.

SKDrepaData-v2 is a publicly available dataset of 33,333 Giemsa-stained microscopic blood smear imag

Optimal Machine Learning Based Automated Malaria Parasite Detection and Classification Model Using Blood Smear Images

Red Blood Cell Morphology Dataset for Image Classification

The dataset was compiled from multiple sources: Chula RBC-12-Dataset (Naruenatthanaset et al., 20

Developing a prediction model for early detection and classification of kidney dysfunction in children with sickle cell disease

Introduction Children with sickle cell disease are at an elevated risk of developing chronic kidney

Morphology-based classification of sickle cell disease and β-thalassemia using a low-cost automated microscope and machine learning

Abstract Sickle cell disease (SCD) and β-thalassemia are the most common monogenic

Image Segmentation and Classification for Sickle Cell Disease using Deformable U-Net

Reliable cell segmentation and classification from biomedical images is a crucial step for both scie