Dataset Composition and Annotation Statistics for Blood Smear Analysis

Neha Margret Issac published annotation statistics for a peripheral blood smear image dataset on 2026-05-12. The dataset supports the development of Elastic YOLO, a morphology-aware object detection model for red blood cells, white blood cells, and platelets. The model achieved a [email protected] of 94.7% and inference speeds up to 78 FPS on an NVIDIA RTX 4090 GPU.

Use Cases

• Training object detection models for automated blood cell classification based on the described cell types.
• Benchmarking detection architectures for medical imaging based on the reported mAP scores.
• Supporting computer-assisted hematology workflows based on the described automated smear analysis.
• Enabling rapid preliminary analysis for telemedicine-based triage scenarios based on the described remote screening use case.

Strengths

• Dataset annotations were performed by clinicians, suggesting high-quality ground truth.
• The associated Elastic YOLO model achieved a high [email protected] of 94.7%.
• The dataset is openly available under a CC-BY-4.0 license.

Limitations

• The dataset is very small at 5.5 KB, indicating it contains only summary statistics, not the underlying images or detailed records.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for training models.

Provenance

Source

figshare

Collection Method

Annotations performed by clinicians for a research project.

Freshness

Last updated 2026-05-12 17:44:35; freshness should be verified.