NVIDIA
NVIDIA
Polyp Detection RT DETR Model
Model
NVIDIA
NVIDIA
Polyp Detection RT DETR Model

Polyp Detection RT-DETR is a RT-DETR model that is designed to detect polyps in colonoscopy images. This model is ready for commercial use.

This model is backed by NVIDIA's Plus Plus (++) Promise
to learn more about the quality of the datasets used to train this model.
FieldResponse
Intended Domain:Polyp Detection in Colonoscopy Images
Model Type:Object Detection
Intended Users:Intended for users that use models for polyp detection tasks.
Output:A dictionary which contains two keys: "pred_logits" and "pred_boxes".
Describe how the model works:The model processes a colonoscopy image to output predicted boxes around detected polyps and their associated logits.
Name the adversely impacted groups this has been tested to deliver comparable outcomes regardless of:Not Applicable
Technical Limitations & Mitigation:Limitation: This model's detection performance may not be 100% accurate, particularly in instances with variable lighting conditions, motion blur from camera movement and different video resolutions. Mitigation: The model can be finetuned to adapt to local imaging conditions for better performance.
Verified to have met prescribed NVIDIA quality standards:Yes
Performance Metrics:Mean Average Precision
Potential Known Risks:The model may result in missed polyps (false negatives) or incorrect detections (false positive) in clinical scenarios. This model is only intended to support diagnostic purposes and should serve only as a supplementary screening aid to assist healthcare professionals in identifying areas of interest during colonoscopy procedures.
Licensing:GOVERNING TERMS: Use of this model is governed by the NVIDIA Open Model License. Additional Information: Apache License Version 2.0.