Accuracy, and how we define it.

[user name]

Ive been having encounters with the local iTero rep lately, and she is frustrated that I dont play well with others. She insists that iTero is the way to go. I tell her that any time a Dr needs work for themselves, family or if its a critical anterior case, its done with PVS.
My reasoning is...PVS captures all the surface details in enamel thats missing with an iTero scan. I want to see every crack and scratch. With a scan, I believe the scan density doesnt have enough points to capture all that, and the 'skin' we see applied to the scan mesh is just an illusion.

I may be blurring some terms, so help me out.
Scan accuracy. How well it relates one point to another. 5 micron, 10 micron? that applies not only to margins, but the over all case, bridge or bar. And its 5 or 10...+ or -. So worst case scenario, with a 10 micron accuracy, you could have deviations of up to 20 microns.

Scan density...how many points are recorded. Could still be very accurate, one point to another, but some detail wont be available because its a lower resolution. When an animated like 'skin' is applied, we become content because our lying eyes accept what we see.

Thats my point of contention with the rep. Sure, your iTero might be accurate, as in not having distortions. But the scan density is not enough to do quality anterior work. I need detail.

Lets go

 

[sidesh0wb0b]

Ive been having encounters with the local iTero rep lately, and she is frustrated that I dont play well with others. She insists that iTero is the way to go. I tell her that any time a Dr needs work for themselves, family or if its a critical anterior case, its done with PVS.
My reasoning is...PVS captures all the surface details in enamel thats missing with an iTero scan. I want to see every crack and scratch. With a scan, I believe the scan density doesnt have enough points to capture all that, and the 'skin' we see applied to the scan mesh is just an illusion.

I may be blurring some terms, so help me out.
Scan accuracy. How well it relates one point to another. 5 micron, 10 micron? that applies not only to margins, but the over all case, bridge or bar. And its 5 or 10...+ or -. So worst case scenario, with a 10 micron accuracy, you could have deviations of up to 20 microns.

Scan density...how many points are recorded. Could still be very accurate, one point to another, but some detail wont be available because its a lower resolution. When an animated like 'skin' is applied, we become content because our lying eyes accept what we see.

Thats my point of contention with the rep. Sure, your iTero might be accurate, as in not having distortions. But the scan density is not enough to do quality anterior work. I need detail.

Lets go

so far ive been getting mainly posterior itero scans. some cases with PVS in addition to the scan.
ive made crowns from the itero models, just the scan data, and from the PVS model we made. generally, they all fit and functioned nearly the same on each with very minor margin discrepancies. nothing that i felt was going to be clinically significant once proper cementation protocols were followed.
so where am i going with this? well.....even at a 20micron accuracy, it wasnt enough to see anything significantly wrong/incorrect. at least not as of yet.

granted these are again mostly posterior cases, and singles at that. bridges and anteriors....well, i will let you know when we get some in for comparison!

sadly, my rep doesnt call me haha. hes in with the big lab locally so he only calls and stops by there. im still a nobody!

 

[user name]

I should add...I have no idea of the iTero scan density or how thats even described. Im hoping someone will educate me.

 

[CoolHandLuke]

I could spend a day talking about this.

Itero was innovative, back in the day. It was an exceptional leap forward in technology because it was a cerec alternative that cost 1/10th the price, and let you control who made your restos without the hassle of doing it yourself. Very much like what rkm's lab setup is. You just scan and send and wait and pay. Since then it has ceased to be innovative. This philosophy is apparent in almost every ios player now.

Scientifically speaking it relies on photogrammetry to stitch pics together to create topography. This has not been innovative since the early 80's. But the hardware became cheaper and software became smarter as the years went on. Now that same technology is 1/100th its price and 1000x more accurate. Drone cams now are able to take a series of photos and create a digital map accurate to 1mm from 1km away.

So why.does it need us to stop accuracy down to 2 decimal places of the camera and stitching is so good? Well while the tech can conceivably infer precision to 4 decimal points, the data files are 300x larger by necessity. Nobody cam work on the fly cad with files that big except nvidia themselves on supercomputers woth ridiculous cards.

Until our cad products are developed woth 4th decimal place accuracy, it is by and large useless.

Then we face the dilemma of actually turning that design into cam. While individual motors, guides and rails, lr ball screws can acheive single micron precision it is in controlled environments. Dustless, loadless. You will look at lasers to be the machining solution of the future. Laser ablation technology will get us the 3rd and 4th decimal place, but not for a while because it only works up to a certain depth, and only on certain materials.

As to whose data is most dense/accurate at the moment is a topic you should probably talk to mr delorean about. I think he had a thing about testing scanners output some time ago.

But yeah could talk about this for hours.

 

[Contraluz]

so far ive been getting mainly posterior itero scans. some cases with PVS in addition to the scan.
ive made crowns from the itero models, just the scan data, and from the PVS model we made. generally, they all fit and functioned nearly the same on each with very minor margin discrepancies.

This has been my experience, with 3shape though, as well. Although, I don't have 'hundreds' of cases of experience, yet... But when asked about it, I tell them, 'it works'. Due to my limited experience, I can't say how well, or bad though

 

[Wainwright]

A 3D IOS like traditional impression materials are just tools that are only as good as the user. They both are one of the core layers of a onion, meaning from the patients mouth to the final product there are many steps in the process that stack up, good or bad. The really difficult part of the equation is everyone has different expectations on what good really is, technicians and dentists.

I think its a mistake to get into the technical details of one part of the overall process, like a 3D IOS for example. The further you dig the less clear the answer or even the question is. To top it off, it can just be downright confusing.

Maybe the iTero and its technology isn't great for veneers, what about the prep style, or tissue management.

I think I'd handle this the same way I would a traditional impression that wasn't good enough for a case, we made the restoration based on the information/impression provided and this is the result. If it does not meet expectations what can we do when the patient is in the chair next time to assure the best result. Maybe take a second impression using a different material as a confirmation of fit or take special care to make sure all critical areas of the prepared teeth are as clear as possible.

Communication and working toward a shared goal is a universal guiding principle in analog or digital workflows.

Don't dig too deep, you might hit water.

 

[JMN]

I think that a lot of the issue is with an impression yu can definitively say "I did my job, it fit the model" where a ios has at least two layers of abstraction before ypu even discuss model creation by a slicing engine..

Plus, I'm thristy.

 

[zero_zero]

Drone cams now are able to take a series of photos and create a digital map accurate to 1mm from 1km away.

Really ? I don't buy it

 

[rkm rdt]

Really ? I don't buy it

you're busted!

 

[JMN]

you're busted!

View attachment 32923

(edit: that was just mean and unneeded)
That ain't millimeter resolution, that's not even centimeter resolution.

 

[rkm rdt]

(edit: that was just mean and unneeded)
That ain't millimeter resolution, that's not even centimeter resolution.

Maybe she's looking for the controller?

 

[zero_zero]

you're busted!

View attachment 32923

You need a bigger boat...

 

[CoolHandLuke]

DARPA shows off 1.8-gigapixel surveillance drone, can spot a terrorist from 20,000 feet - ExtremeTech

DARPA and the US Army have taken the wraps off ARGUS-IS, a 1.8-gigapixel video surveillance platform that can resolve details as small as six inches from an altitude of 20,000 feet (6km). ARGUS is by far the highest-resolution surveillance platform in the world, and probably the...

www.extremetech.com

article now 6 years old.

 

[JMN]

DARPA shows off 1.8-gigapixel surveillance drone, can spot a terrorist from 20,000 feet - ExtremeTech

DARPA and the US Army have taken the wraps off ARGUS-IS, a 1.8-gigapixel video surveillance platform that can resolve details as small as six inches from an altitude of 20,000 feet (6km). ARGUS is by far the highest-resolution surveillance platform in the world, and probably the...

www.extremetech.com

article now 6 years old.

That tech is at least a decade old.
I remember that article as it was a massively integrated and very cheap, relatively, initiative.
Still, six inches is bigger than 0.04mm by a few orders of magnitude.

From a hot air ballon, yeah, maybe. Newspaper headline for sure, but newspaper article, not this year. Next year maybe.

 

[CoolHandLuke]

even if you dont believe its out there, there are tons of private sector companies offering 1mm per pixel uav mapping services, most of which do so in the 40-120ft unregulated airspace. You can bet the military can and does do it in the 1-5km zone for light aircraft. If darpa had 6" per pix half a decade ago, theyve made 3-5 moores law improvements by now already.

Like nasa putting 3d printers on the ISS. They would have two kinds. One, the kind that doesnt depend on powders or liquids that take postprocessing. Two, the kind that are more advanced than tbe ones you find in Staples, but generally built along the same principles. Filament printers. Spools. Years ahead of the consumer market, printing lightweight plastic shore harness near metal in spools.

 

[zero_zero]

even if you dont believe its out there, there are tons of private sector companies offering 1mm per pixel uav mapping services, most of which do so in the 40-120ft unregulated airspace. You can bet the military can and does do it in the 1-5km zone for light aircraft. If darpa had 6" per pix half a decade ago, theyve made 3-5 moores law improvements by now already.

More like six inches per pixel for private companies... to about a foot accurate if is done right, I had it done for one of my properties last year, paid and arm and a leg for a model which was only accurate to +/- 5 feet compared to survey markers... BTW vegetation and water doesn't read very well either.
I am sure military can do way better, but still not to 1mm

 

[JMN]

Land surveying can be accutate to a few mm with the right, and very expensive, gear. Brother in law had a cave entrance and surroundings mapped last year.