How does red dot® work?:
![](https://www.behold.ai/wp-content/uploads/2022/03/riding-lightbulb.png)
Developed by consultant radiologists
Our AI algorithm has been developed by highly trained and experienced US Board Certified and NHS Consultant Radiologists.
![](https://www.behold.ai/wp-content/uploads/2022/03/time-back-grey.png)
Results in 60 seconds
Our AI algorithm provides results back in the PACS within 60 seconds on receipt to behold.ai systems.
![](https://www.behold.ai/wp-content/uploads/2022/03/superhero-icon-3.png)
Immediate access
The result is directly accessible by the user. No additional application is required.
Compared to outsourcing teleradiology, red dot® is...
![](https://www.behold.ai/wp-content/uploads/2022/03/cost-save-icon.png)
Less expensive
compared to outsourcing and training radiologists
![](https://www.behold.ai/wp-content/uploads/2022/03/faster-icon.png)
Faster
Delivers a diagnosis
in 60 seconds
![](https://www.behold.ai/wp-content/uploads/2022/03/40x-icon.png)
More accurate
than a human with 95% confidence
![](https://www.behold.ai/wp-content/uploads/2022/03/superhero-icon-lg.png)
How does red dot® perform?
AUROC
0.975
with over 90%+ confidence intervals [0.966 – 0.984]
Accuracy
90.20%
with over 90%+ confidence intervals [88.06 – 92.08]
Sensitivity
94.65%
with over 90%+ confidence intervals [91.46 – 96.91]
Specificity
87.95%
with over 90%+ confidence intervals [85.04 – 90.46]
Dataset was composed of 888 CXR images from 2 clinical sites from the United States (n=738) and 2 clinical sites from the United Kingdom (n=150). The final standalone dataset included pneumothorax positive (n=299) and negative cases (n=589). Both groups contained confounding imaging features.
Interested in talking to us about AI?
Get in touch to learn more about how red dot® can make a difference to your patients and your organisation.