1000Lungs: Synthetic Virtual Bronchoscopy Dataset for SLAM Training

A synthetic virtual bronchoscopy dataset built using SynAirG, a customizable model for generating synthetic worlds for endoluminal interventions. The dataset contains models built from MAISI-generated chest CT volumes, each paired with airway masks, meshes, centerlines, and rendered camera paths. It was created by chrisvoncsefalvay and last updated on 2026-06-18.

Use Cases

• Training bronchoscopy SLAM models based on synthetic airway meshes and camera paths.
• Developing computer vision algorithms for endoluminal navigation based on rendered procedural imagery.
• Benchmarking medical image analysis tools based on synthetic CT volumes and paired airway masks.

Strengths

• Contains 1000 synthetic lung models, providing a substantial base for model training.
• Includes multiple paired data modalities per case, such as airway masks, meshes, and centerlines.
• Built using a customizable synthetic generation model (SynAirG) designed for bronchoscopy.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

huggingface

Collection Method

Synthetically generated using SynAirG from MAISI-generated chest CT volumes.

Freshness

Last updated 2026-06-18 12:07:32; freshness should be verified.