This article was initially published in French on infobref.com
The technique of so-called minimally invasive surgery has brought many benefits to patients and health services. But it comes with one requirement: a tiny camera inserted into the patient’s body must give the surgeon a clear view of what’s going on inside while he or she operates. However, the camera’s lens is often obstructed, which handicaps the surgeon’s work. The Quebec-based start-up Vope Médical has found a way to remedy the problem, thanks to artificial intelligence.
The problem the company is tackling is well known to surgeons who perform minimally invasive surgery, that is, with long instruments inserted into the patient’s body, usually after a small incision.
The surgeons wield their instruments while watching, on a monitor screen, the image of the inside of the body sent to them by a small camera.
“In more than a third of cases, the camera lens becomes contaminated during the procedure,” says Amy Lorincz, co-founder and CEO of Vope Medical. This contamination can be due to blood, grease, or simply fogging. The result is that the lens is at least partially obstructed and the surgeon no longer has a good view of the surgery.”
When the lens is too clogged, the surgeon can usually clean it in a mechanical way, by commanding fluid to be delivered to the lens through an irrigation device.
But sometimes the camera has to be removed from the patient, the lens cleaned, and then the camera reinserted.
“This lengthens the time of the surgery,” notes Amy Lorincz. Also, it puts stress on the surgeon who has no visibility while the camera is being cleaned, which can increase the risk of errors.”
The solution imagined by Vope Medical is to use computer vision, a branch of artificial intelligence.
“The AI analyzes the image collected by the camera in real time,” explains Amy Lorincz. It detects what type of contaminant is partially obstructing the lens. The AI software then compensates by recomposing missing parts or details in the image. This allows the surgeon to see a more complete and clearer image on his or her monitor.”
If the lens is too obstructed for the image to be adequately corrected, then Vope’s system commands the irrigation device to clean the lens.
On the left, a fogged image from a camera; on the right, the image corrected by Vope’s software [source: Voce Medical]
The business model is based on the sale of licenses for the software created by Vope Medical. Customers will be medical equipment suppliers who want to incorporate Vope’s technology into their products. “We won’t be making our own camera,” explains Amy Lorincz. Instead, we will integrate our solution with devices and systems that surgeons already use.”
Currently, Vope Medical has developed an initial “minimum viable product” version of its algorithm. It corrects for fogging and detects physical contaminants that obstruct the lens.
The company employs three people, but plans to hire in the coming months.
Several surgeons from McGill University are collaborating with the company to help it finalize its product.
Vope Médical is one of 20 start-ups selected this year in the Bourse+ program of Startup Montréal.
Next steps for Vope Medical
The company wants to apply its algorithms to real-time imaging, then integrate them into existing medical cameras to develop a “proof of concept” version by next spring. This will allow the company to take its tests with surgeons to the next level. It also plans to conduct a pre-seed financing round at that time.
In the immediate future, she is looking for AI and computer vision experts interested in solving technical challenges, and she plans to partner with hospitals to test her product on a larger scale.