CADDIE

What is CADDIE?

CADDIE is Odin Vision’s cutting-edge Artificial Intelligence system for a new era of endoscopy. It supports endoscopists to detect and characterise colorectal polyps in real-time during colonoscopy procedures.

In the UK, there are over 42,000 new cases of colorectal cancer (CRC) and 16,000 deaths per year, making it the 2nd leading cause of cancer deaths. The number of CRC related deaths is predicted to increase by 51%[1] over the next 15 years, with increased prevalence in young people[2].

Challenges in Colonoscopy

• Polyp Detection: Up to 25% of adenomas are missed during
  colonoscopy [3], with each 1.0% increase in the adenoma detection rate correlating to a 3.0% decrease in the risk of interval cancer[4].
• Polyp Diagnosis: Only 3% of NHS histopathology departments have enough staff to meet clinical demand. As demand increases, this will lead to longer waiting times for results[5].
• Colonoscopy Quality: 70% of post colonoscopy colorectal cancers have avoidable factors including poor bowel preparation and shortcomings in caecal intubation[6].

Comprehensive AI for Colonoscopy

AI for Polyp Characterisation

CADDIE supports trained endoscopists by displaying information relating to the visual characteristics of colonic polyps, enabling the user to characterise the highlighted tissue as adenoma or nonadenoma.

Endoscopists who are suitably trained, can use optical diagnosis to decide the histology of a polyp and follow clinical guidelines for resection and histopatholgy. Optical Diagnosis means patients can benefit from same day results, reducing unnecessary waiting time.

AI for Polyp Detection

A clear green rectangle is placed around the potential polyp in realtime. The endoscopist is also drawn to the area of interest by an audible alert.

AI for Bowel Cleanliness

Incomplete bowel preparation can result in hidden colorectal mucosa and hinder polyp detection. CADDIE provides an additional aid to the endoscopist – an easy to view and intuitive on-screen live display, showing the percentage of visible mucosa in the endoscope’s field of view.

AI for Caecum

20% of interval cancersare located at the caecum. A key quality performance indicator of the colonoscopy procedure is to identify caecle anatomy such as the Appendiceal Orifice (AO). CADDIE’s AI can identify images containing the AO and remind users to switch on polyp detection for the withdrawal phase.

Comprehensive AI for Colonoscopy

Easy to Use

CADDIE has a simple and user friendly interface that integrates into the current clinical workflow. Visual information and audio alerts are used to direct the user’s attention to regions of interest.
Due to the intuitive nature of CADDIE, minimal training is required. Hands free interaction during the procedure. Detection and characterisation controlled by foot pedal.

Set up and Integration

CADDIE can be configured for single monitor use or used with a second monitor. The system can integrate with existing clinical workflows and endoscopy equipment. It is not tied to one endoscopy equipment manufacturer. It can be used as a stand alone computer installed on the endoscopy stack or via the cloud

CADDIE in the Cloud

CADDIE’s Advanced Cloud Technology

Our secure cloud solution uses a small computer connected to the endoscopy stack to stream the video
from the hospital to the cloud. The system analyses the images in the cloud in real-time and sends the results back to the endoscopy monitor for the endoscopist to see.

Real-Time

CADDIE uses bespoke data streaming technology to securely send data to the cloud and back in real-time.

Flexible

On demand service and available everywhere. Install into all endoscopy rooms and customise your usage to suit your demand. Per procedure cost effectiveness.

Future Proof

Instant and automatic updates. Get the latest AI without engineers having to visit the hospital to install
software/hardware updates. There is no waiting and no down time with new features accessed at the click of a button.