GI Genius vs Standard Colonoscopy: Is it Better?
Rosario Ligresti, MD, FASGE Speaks About The GI Genius Intelligent Endoscopy Module (colon cancer)
For many years, colonoscopy has been considered the gold standard for colon and rectum cancer screening. And we know, based on many years worth of data suggesting how accurate colonoscopy is. But when you actually dive down somewhat more deeply between the lines, you find out that colonoscopy, as we practice it currently, may not be as accurate as intended, with potentially missed colorectal polyps.
You consider some of the more recent reviews of colonoscopy and where the accuracy kind of lies. Just garden variety screening colonoscopy is about 89% sensitive, which would sound fantastic if we thought there was some ability not to do better.
What Type of Data Do You Find?
So when you look at some of the more nitty-gritty parts. The data, you find that if you do, for example, a colonoscopy in the morning, you’re more likely to be more accurate than doing a colonoscopy in the afternoon. And this is again what gets to the heart of the matter of why artificial intelligence has been thought to be a very good thing to bring to colonoscopy and colon cancer screening, generally basically gastroenterologists, so the number of missed colorectal polyps could be avoided.
Good, but not quite that good. And no matter how good of a colonoscopist the gastroenterologist thinks they are, we can certainly do better. When you look at, what artificial intelligence brings, it brings basically an expert into the room who doesn’t get tired and who always is looking at the screen behind you and commenting in real-time.
On the screen, what it thinks is a clinically significant poly, and it’s doing this based on a database of about 13 million images stored over a period of three years of the development of the technology in this box that goes between the patients and the screen. And it’s watching the colonoscopy in real-time (for colon cancer).
And alerting the colonoscopist when it detects a colorectal polyp detection rate, and you would think again, how accurate could that be? As it turns out, artificial intelligence brings an increased level of accuracy of probably between 11 and 15% above our normal ability to detect colon polyps (possible colon cancer). So again, if you look at what our most important metrics are for colonoscopy as an introduction to this talk today, we look at something called the adenoma detection rate.
It’s essentially how many adenomas are being detected in patients undergoing a screening colonoscopy and what that percentage is compared to a negative screening colonoscopy, and a good colonoscopist should be in the thirties, 30 to 33% adenoma detection rate a very good colonoscopy certainly could be much higher than that.
What Happens When You Add GI Genius (intelligent endoscopy (artificial intelligence)) to a Gold Standard Colonoscopy?
But if you add artificial intelligence to that mix, all of a sudden, we’re finding colorectal polyps between 10 and 15% higher adenoma detection rates. So we’re clearly improving the test substantially at the really minimal additional effort. So what does artificial intelligence bring to a standard colonoscopy? It brings a much better colonoscopy for the patients screening for colon cancer.
So to dive into a little bit more about artificial intelligence and what it brings to computer colorectal polyp detection rate, if you look at compared to a standard white light colonoscopy, adding computer-rated detection rate or artificial intelligence to the mix results (outcomes) in, as we said, a 14% increase in adenoma detection rate a 30% relative increase in adenoma detection rate, but more importantly, a 50% greater likelihood to detect multiple colorectal polyps during a standard colonoscopy. And this data was shown quite nicely in Mike Wallace’s recent paper that was published this year and presented at Digestive Disease Week in San Diego.
So really, you’re, we are at minimal additional effort making a colonoscopy that much better. The actual procedure itself is about 99% sensitive to detecting colorectal polyps, and it detects the colorectal polyps probably 82% faster than a gastroenterologist would. And again is doing this in real-time as you’re doing the colonoscopy, basically on your screen, the normal colonoscopic image, and then there’s a little green box (to highlight regions) that follows you along the way.
And anything that the computer detects polyps as being a clinically significant colorectal polyp detection rate, it puts a flashing green box around that area, drawing your attention to it. So, in reality, it’s fairly unobtrusive or unobtrusive and easy to use.
What is The Technology Behind GI Genius and It’s Polyp Detection Rate?
GI Genius (tool), the technology behind GI Genius (intelligent endoscopy module (artificial intelligence)). I should say. So computer-rated poly detection has been a process and evolution over the last 5 years in development. It was predominantly developed in Europe.
More importantly, developed for the most part in Italy, the FDA approved the technology behind GI Genius (intelligent endoscopy module (artificial intelligence)) in April of 2021 through a Denovo pre-market review pathway. What they looked at is a multi-center prospective, randomized controlled study in Italy done over a period of a year and a half with 700 subjects who were between the age of 40 and 80.
All of those patients were undergoing colorectal cancers screenings. They based the data for computer detection on a subpopulation of 263 patients who were being screened or surveilled for a period of every 3 years. And they compared one group that had a standard white light colonoscopy done by an expert gastroenterologist versus a standard white light colonoscopy with the addition of computer-aided detection for colon cancer.
What Did the Study Find?
And with that study, they concluded that the technology behind GI Genius (intelligent endoscopy module (artificial intelligence)) was able to detect (polyps) lab-confirmed or pathologist-confirmed adenomas 55% of the real-time compared to standard colonoscopy without GI Genius (tool) involved, only 42% of the real-time. So there was an observed and clinically significant difference of 13% between colonoscopies done without computer-aided detection for colon cancer versus with so based on that data, which was fairly astounding, showing how much better colonoscopy (images) could be made if the FDA approved this technology.
And it was marketed here in the United States under the brand named GI Genius (intelligent endoscopy module (artificial intelligence)) by a company called Medtronic.
What Is Needed To Get the GI Genius Technology?
Again what was found and the reason why not every gastroenterologist in the world is switching over to artificial intelligence or computer aid detection. Again comes down to probably cost as the biggest issue. These devices, again, they’re new. They are, again there’s a hardware and a software component.
There’s a continuous upgrade. That’s required. They’re very expensive pieces of technology to implement. And if you just simply left it to market forces, more than likely, this technology would not be distributed to every market where it’s actually needed the most. At the beginning of this year, Medtronic, together with the American Society of Gastrointestinal Endoscopy and Amazon Web Services, put together a program called Health Equity Assistance Program.
And in this program we. Found that you were able to bring the newest technologies to communities that were otherwise very unlikely to get these technologies either anytime soon or ever. If we look at just stepping back for second colon cancer in underserved populations, just thinking about this broadly, only one in three adults meets color colorectal cancer screening guidelines or criteria.
I’ve not been screened at all, and that’s across the board in the United States, but if you start breaking that data down even further, you find that Black Americans are about 20% more likely to get colorectal cancer screening, but 40% more likely to die from it, therefore would help in detecting colorectal cancer. And again, this comes down to, a deficiency in screening. When you look at Asian Americans, another group that’s traditionally underserved, less than 50% of Asian Americans are up to date with colorectal cancer screening. And there’s tremendous variability in Asian subgroups. So again, you’re looking at large segments of populations that are not able to get colorectal cancer screening.
How Could We Prevent Colorectal Cancer?
And we know that colon cancer is one of those cancers that can be imminently prevented. Probably more than 65% of colon cancers could be prevented if patients could be properly screened. So with this grant process that started at the beginning of the year, we were looking at underserved communities and access to technologies such as GI Genius (intelligent endoscopy module (artificial intelligence)). And as it turns out, Hackensack is in a community or serves a community in Northern New Jersey that is dramatically underserved. And we have colorectal cancer screening rates that are substantially below the national average because of that.
How Can You Get the GI Genius Intelligent Endoscopy Module?
And because of what’s involved, we said, why not apply for this grant? At the beginning of January, so we did, and thankfully we were awarded 3 GI Genius (intelligent endoscopy module) units for distribution in our center to bring this technology to our unserved community. So we just started implementing the use of this technology in May.
We were notified that we were grant awardees at the beginning of May. They were installed in the middle of May and started using this technology in the very last week of May into June. And it’s been wildly successful.
Thank goodness. The addition of computerized (early) detection doesn’t affect the billing of colonoscopy at all. So if patients are otherwise appropriate for screening colonoscopy, they can certainly be appropriate for screening colonoscopy with GI Genius (intelligent endoscopy module) attached, and it doesn’t add additional costs to the colonoscopy.
If anything, the overall cost probably will go down as more colon cancers are detected at an earlier stage. The initial worry with this whole technology in worldwide use was that it was gonna increase the cost of a colonoscopy because all of a sudden, the gastroenterologist is going to find more colorectal polyps to remove.
But ultimately, as it turns out, when cost analysis has been done over the last 12 months, it turns out that overall cost has decreased. So certainly, this is covered by insurance because, again, it’s part of a screening colonoscopy. It adds only a few minutes to each colonoscopic examination. It doesn’t add any additional risk to the patient, and if anything, it’s mostly adding benefit.
Is There Anywhere Else the GI Genius (Intelligent Endoscopy module) is Available?
Very slowly. In New Jersey, we were the first to be awarded this grant to have 3 full installations of the devices done in our outpatient endoscopy unit.
In the rest of the country, there are, there were 62 other grant awardees. So this technology is slowly being rolled out to the rest of the country. And like I said, the main issue with why it’s not being adopted by everybody is cost. Each individual box is about a hundred thousand dollars.
So most large hospital systems and certainly most practices, and most endoscopy centers are gonna have a little bit of a tough time considering implementing this technology routinely at a hundred thousand dollars per installation. Thank goodness. Part of, as part of the grant, the installation for Hackensack was cost-free, absorbed entirely through the grant. And it again makes us the first center of New Jersey to have this at cost and certainly in high volume capability.
As I said earlier, I think it’s a game changer. And why do I think it’s a game changer? Every 1% increase in adenoma detection rate results (outcomes) in a 3% decrease in ultimately diagnosed colon cancer. So no question, if you’re, again, the bottom line is if you can find if you can make colonoscopy better and you can detect more adenomas in patients.
How Is This Affecting The Detection of Colorectal Polyps?
You’re clearly affecting overall rates of colon cancer, and that ultimately saves lives. In the end of the day, we know that colon cancer is anywhere between the top 2 or top 3 cancers in the country. We know that a large number of patients are getting diagnosed at a younger age.
What is the Effectiveness of GI Genius (Intelligent Endoscopy) to Help Identify Polyps?
We know that this is cancer that can be beaten if they identify it early enough, and we know that large segments of the population aren’t availing themselves of the technology period. My feeling would be if it makes the colonoscopy better, more accurate and more effective, and potentially a more impressive test, patients again are more likely to come to get it again.
It doesn’t make the whole experience of a colonoscopy better, but it certainly makes the test a much more valuable test for the patient. So as far as I’m concerned, game changer because, again, it makes a colonoscopy that much better.
What Are The Highlights About The Gi Genius (Intelligent Endoscopy)?
At the end of the day, what is this show? It shows that gastroenterologists performing routine colonoscopies are okay, but being okay is just not okay. These days, I really think that if, at the end of the day, we can add minimal effort, bring a much better test for patient colon cancer screening.
Then the answer should be that we should. And more importantly, if you can do this for every colonoscopy you do. By all means, you should. And if you can avail yourself of that technology as a patient and as a provider, again, you should. Where does the future point for artificial intelligence?
For the time being, and that’s something that the FDA readily approved, the addition of GI Genius (intelligent endoscopy module in real-time) to colonoscopies makes the colonoscopy better at detecting polyps.
Colon polyp detection rate will be in the near future will be, and hopefully, down the road, the far future will also be that ultimately you can do away with physical pathology and physical biopsy interpretation, and the computer itself will be able to tell you in real-time what the colorectal polyp (detection rate) is pathologically.
Why is That Important for Patients with Colon and Rectum Cancer?
Because again, it’s gonna save on biopsy results. It’s gonna save on further colorectal polyp (detection rate) up characterization a few days after the colonoscopy, but in real-time, it’ll tell you whether a polyp is necessary to be removed or not. That’s where the future is gonna be. So polyp characterization, right?
At this moment, it’s all about colorectal polyp detection, and clearly, adding artificial intelligence to the mix of colonoscopy polyp detection makes that much better. So I take home the message, why wouldn’t you have a colonoscopy with GI Genius (intelligent endoscopy module) attached to it?
Dr. Rosario Ligresti, MD, FASGE – About The Author, Credentials, and Affiliations
He is the chief of the gastroenterology division at Hackensack University Medical Center, which was recently named by U.S. News & World Report as the finest hospital in New Jersey for GI treatment. Dr. Ligresti is dedicated to using cutting-edge, evidence-based medicine to improve the lives of his patients and has access to cutting-edge technologies to deliver high-quality care.
Dr. Ligresti specializes in the diagnosis and treatment of intestinal malignancies utilizing modern interventional endoscopic procedures. He also does pancreas and esophageal radiofrequency ablation. He has been a member of the American Society of Gastrointestinal Endoscopy for many years and has served the Bergen County community of Hackensack for many years.