Is this the Varian/Elekta killer?

Started by Beerus
Physicians and veterinarians: Turn your expertise into action! Explore CDC’s Epidemiology Elective Program and Epidemic Intelligence Service today.
This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.
Beerus said:

uRT-linac 506c (Integrated CT-linac) | United-Imaging

Discover uRT-linac 506c, which enables precise treatment and supports all advanced radiotherapy applications. Large FOV and scan length.
global.united-imaging.com global.united-imaging.com

uRT-linac 506c offer diagnostic-quality CT-based IGRT and online ART.
Click to expand...
Could be if it’s a lot cheaper. The new economics of how little we get paid for radiation now (as low as $120 a fraction in many cases) will dictate who sells most linacs in future. Linacs need to come down to half million price range.
 
Advertisement - Members don't see this ad
A lot of obstacles to this gaining traction in the US:

1. Corporate friendliness between ASTRO, academic RO, and big hospital RO with Varian/Elekta. There is also embedded OIS and TPS which has taken decades to root.

2. Varian/Elekta lobbyists will absolutely pounce on this to avoid approval of this in the US. Given this current administration's position of "America first" and weaponization of federal bureaucracy, they will find a sympathetic ear.

3. Consider Chinese EVs (especially the Xiaomi SU7 sedan) which reviewers have judged as at least as good but way cheaper than US equivalents. It is already rolling out in the EU but is heavily restricted in the US due to #2.
 
WildRivers said:
Shameless self promotion, as I’ve written about United Imaging’s linac.

open.substack.com

Varian vs. The World: Can Anyone Break Through?: Part 1

A deep dive into the potential of United Imaging's uRT-linac 506c and its challenge to Varian’s reign.
open.substack.com open.substack.com

open.substack.com

Automating Radiotherapy: The Promise and Challenges of AI: Part 2

A closer look at a multistep integrated AI-powered workflow and its potential to disrupt the status quo and implications for radiation oncology.
open.substack.com open.substack.com
Click to expand...
It's very interesting. "Software" is the solution not the hardware per se. I have been saying for a few years that if Apple or Google made linacs we would already have fully automated everything (contouring, planning, fusion, adaptation, even dose recipes)... we'd be sort of very professional rad onc customer service people for face-to-face and maybe tweaking a few parameters here and there. Maybe doing some custom bolus for skin in the sim suite etc. Managing side effects. On site for RMA and SRS. But this is a solvable problem. Don't need the Chinese to do it. Eventually Varian and Elekta etc will have it. I predict we will get something like this first vs robots doing great surgeries autonomously first, but I do predict the latter WILL happen.

The whole "med school" concept is looking more and more suss. Lots of things in society going to need reworking. But if you don't realize that radiology and rad onc have MANY A.I.-doable things, then you aren't being smart.
 
This might be an unpopular opinion but I don’t care how good or cheap this is, it should never be allowed in US.

I’m a big supporter of “America first”. Especially when it comes to China, we have to be very careful.
 
Varian needs a competitor badly. Elekta tries and it’s fine but we need another player in the market badly.

Just don’t put the new linac in a clinic familiar with a Varian. When they run side by side you see how much smoother the Varian runs….but you’ll pay for it.
 
I also worry about Chinese tech. But, I think ship has sailed. A huge number of Americans have iPhones, play on TikTok and share gobs of data and utilize any number of Chinese apps/hardware that are potentially stealing our data or polluting our digital ecosytem. Our government is allowing the selling of NVIDIA chips to China. We half heartedly are America First, unless it can enrich us or make our lives easier in some way, while giving away the keys to the house. Uncertain how far United Imaging's linac will go, but I presume people know that their imaging machines are used all over the United States. They are a bit cheaper than comparable tech and use volume pricing to get into the market. My friend that runs an imaging company says, "They are ok, but people like them because they are cheap." That tells you how much people value data security and protecting US interests. FYI, they are applying for approval already, I think before the time I posted that article. Let's see what happens.
 
Advertisement - Members don't see this ad
DrProtonX said:
This might be an unpopular opinion but I don’t care how good or cheap this is, it should never be allowed in US.

I’m a big supporter of “America first”. Especially when it comes to China, we have to be very careful.
Click to expand...

Varian, while originally initiated in Palo Alto is now owned by Siemens, a German company
Eleketa is a Sweidhs company

So when you mean 'America first', you actually mean 'Caucasians first'?

Jennifer Lawrence Thumbs Up GIF
 
I don't understand why they insist on bolting this software to a LINAC. That's great that you can treat the patient 20 minutes later, but I don't really see that as a solution to real clinical problem - and short circuits our usual mechanism for peer review.
 
Beerus said:

uRT-linac 506c (Integrated CT-linac) | United-Imaging

Discover uRT-linac 506c, which enables precise treatment and supports all advanced radiotherapy applications. Large FOV and scan length.
global.united-imaging.com global.united-imaging.com

uRT-linac 506c offer diagnostic-quality CT-based IGRT and online ART.
Click to expand...

Ethos with Hypersight does that already. Downsides of Ethos - sucks at 3D. Doesn't have high MV beams. If it can fix those two things they have a chance.

Of course they'll have to integrate into currently existing planning software at an institution or you'll hear a giant collective *ughhhhhhhhhhhhh* from a department's Radiation Oncology staff having to learn yet ANOTHER planning software.
 
evilbooyaa said:
Ethos with Hypersight does that already. Downsides of Ethos - sucks at 3D. Doesn't have high MV beams. If it can fix those two things they have a chance.

Of course they'll have to integrate into currently existing planning software at an institution or you'll hear a giant collective *ughhhhhhhhhhhhh* from a department's Radiation Oncology staff having to learn yet ANOTHER planning software.
Click to expand...
After this many years, you hardly see any publications from Ethos doing this for more complex cases. Upon my review I saw that it was mostly prostate. Are the Varian machines really able to sim/plan/QA/IGRT in 20 minutes for HNC? If so, why are they not advertising this or demo-ing it? I think what I found most interesting with United was that they picked a site (nasopharynx) considered one of the more challenging to both contour and to plan. And then months out, did a blind QA of contours and it was so accurate.

If Varian or Elekta can do this, this is not a hard study to put together. There are high volume centers across the country. Publish it and say, "we can do it, too". But, research that is meaningful is quite hard to come by in the United States.

I did a little digging/investigative type of work and there is of course concerns (probably true) that United steals technology. And, can we be sure their research is legit? Red Journal is the one that published this. The peer review is not amazing at that journal, so who knows the reality? But, this is one of those things that should get our people motivated to show we can do it and more importantly - do it better. American innovation, ingenuity and competitiveness appear to be at an all time low, particularly with regards to radiation oncology. With this new coding situation and future case based, we are continuing to incentivize lower rates of innovation, rather than higher. Time will tell, but I am not optimistic.
 
WildRivers said:
After this many years, you hardly see any publications from Ethos doing this for more complex cases. Upon my review I saw that it was mostly prostate. Are the Varian machines really able to sim/plan/QA/IGRT in 20 minutes for HNC? If so, why are they not advertising this or demo-ing it? I think what I found most interesting with United was that they picked a site (nasopharynx) considered one of the more challenging to both contour and to plan. And then months out, did a blind QA of contours and it was so accurate.

If Varian or Elekta can do this, this is not a hard study to put together. There are high volume centers across the country. Publish it and say, "we can do it, too". But, research that is meaningful is quite hard to come by in the United States.

I did a little digging/investigative type of work and there is of course concerns (probably true) that United steals technology. And, can we be sure their research is legit? Red Journal is the one that published this. The peer review is not amazing at that journal, so who knows the reality? But, this is one of those things that should get our people motivated to show we can do it and more importantly - do it better. American innovation, ingenuity and competitiveness appear to be at an all time low, particularly with regards to radiation oncology. With this new coding situation and future case based, we are continuing to incentivize lower rates of innovation, rather than higher. Time will tell, but I am not optimistic.
Click to expand...

They feel no need to because they don't feel like they have anything to prove.

Can you link the IJROBP paper in this space?

All AI contours are 'good enough' if you accept whatever the AI spits out at you. That being said, Ethos not having AI influencers above the diaphragm, meaning that say HNC is a deformable registration of the CT sim contours which is edited PRN by therapists prior to physician review. 20 minutes of physician time at a facility where therapy/dosi may not be unreaonable, but I'd be surprised if even UTSW (who is doing a metric ton of adaptive) is having a H&N pt on table for 20 minutes for an adaptive treatment.
 
evilbooyaa said:
Varian, while originally initiated in Palo Alto is now owned by Siemens, a German company
Eleketa is a Sweidhs company

So when you mean 'America first', you actually mean 'Caucasians first'?

Jennifer Lawrence Thumbs Up GIF
Click to expand...
I’m less worried about Germany and Sweden (our allies) than China. Also how many items in your household/office have “made in Germany/Sweden” on them vs. “made in china”?
 
This post was meant to discuss the unique features this company is offering that might disrupt the duopoly Varian and Elekta have built. I did not expect this post to enrage white nationalists, and ironically neither Varian nor Elekta is US-owned.
This Chinese company is already selling all kinds of products in the EU and the US, I think it is only a matter of time until they get FDA approval and start selling their LINACs as well.
I also highly doubt that the LINAC market is big enough to trigger another US–China turf war. The LINAC market is tiny compared to mobile phones and EVs, so I doubt it will even be noticed at that level.
Regarding security and data, I highly doubt the Chinese will care about your GTV and CTV that you made for your 80-year-old grandpa. Even if it for patient, like “Biden”, why would China care whether you met the constraints or not?
 
WildRivers said:
After this many years, you hardly see any publications from Ethos doing this for more complex cases. Upon my review I saw that it was mostly prostate. Are the Varian machines really able to sim/plan/QA/IGRT in 20 minutes for HNC? If so, why are they not advertising this or demo-ing it? I think what I found most interesting with United was that they picked a site (nasopharynx) considered one of the more challenging to both contour and to plan. And then months out, did a blind QA of contours and it was so accurate.

If Varian or Elekta can do this, this is not a hard study to put together. There are high volume centers across the country. Publish it and say, "we can do it, too". But, research that is meaningful is quite hard to come by in the United States.

I did a little digging/investigative type of work and there is of course concerns (probably true) that United steals technology. And, can we be sure their research is legit? Red Journal is the one that published this. The peer review is not amazing at that journal, so who knows the reality? But, this is one of those things that should get our people motivated to show we can do it and more importantly - do it better. American innovation, ingenuity and competitiveness appear to be at an all time low, particularly with regards to radiation oncology. With this new coding situation and future case based, we are continuing to incentivize lower rates of innovation, rather than higher. Time will tell, but I am not optimistic.
Click to expand...
Varian/Siemens initiated something like 7 adaptive trials across multiple disease sites on the Ethos platform. Imagine they'll read out over the next few years. Most are single arm, so how much impact they will have is questionable. But Varian really pulling back from funding trials at this point.

Can't agree more with IJRBOP peer review. After another unfortunate experience last year, told the editors would decline all further invitations.
 
evilbooyaa said:
more of my household items are made in china than Germany/Sweden my dude/dudette.
Click to expand...
Exactly my point. That’s why I care more that fewer “made in china” items come to US. Germany, Sweden etc… I don’t really care.

I know LINAC market is not big enough but “America first” starts with each and every single one of us.
 
Beerus said:
This post was meant to discuss the unique features this company is offering that might disrupt the duopoly Varian and Elekta have built. I did not expect this post to enrage white nationalists, and ironically neither Varian nor Elekta is US-owned.
This Chinese company is already selling all kinds of products in the EU and the US, I think it is only a matter of time until they get FDA approval and start selling their LINACs as well.
I also highly doubt that the LINAC market is big enough to trigger another US–China turf war. The LINAC market is tiny compared to mobile phones and EVs, so I doubt it will even be noticed at that level.
Regarding security and data, I highly doubt the Chinese will care about your GTV and CTV that you made for your 80-year-old grandpa. Even if it for patient, like “Biden”, why would China care whether you met the constraints or not?
Click to expand...
It definitely took a turn, no question about that. Holy moly.

I don’t want to feed some of the energy here, but potential security risks are very real with any software. All of these programs and planning systems integrate with your broader network and can offer an access point to send or receive data. As you pointed out, none of these are American companies and we are notoriously good at ripping ourselves off. Not sure I see this as substantially riskier with a Chinese company myself.
 
Advertisement - Members don't see this ad
WildRivers said:
while giving away the keys to the house
Click to expand...
Does anyone who has been in a hard sciences, math or comp sci department in the USA within the last 30 years really think that China can't and won't build their own house?

Cultural exceptionalism is transient...always. Personally, while it would probably not meaningfully or positively impact my workplace for what is likely to be the remainder of my career, meaningful market competition for Varian/Elekta is probably a good thing. It sucks being a captive consumer when you are a smaller site.

Regarding the auto-contouring/auto-planning/auto-everything conundrum...I suspect there are multiple factors why this has not emerged as rapidly as some of us would have thought. But the biggest barrier, IMO, is the cultural emphasis on "perfect" contours, which for the difficult sites, don't exist. (H&N, anything post-op, anything with elective nodes). "Good enough" automation, defined as plans where any difference in clinical outcomes would be exceptionally difficult to demonstrate (non-inferior control and toxicity outcomes) has likely been feasible at a large scale for upwards of 15 years.

I'm glad it hasn't been implemented in this timeframe. Very glad. Efficiency is the enemy of the employed.
 
I doubt this is a Varian/Electa killer. In essence, this looks like a LINAC + a CT on rails + the software to adapt.
This is not a new principle. The only new thing is they put everything into one machine.
 
DrProtonX said:
Exactly my point. That’s why I care more that fewer “made in china” items come to US. Germany, Sweden etc… I don’t really care.

I know LINAC market is not big enough but “America first” starts with each and every single one of us.
Click to expand...

Again, America first means we should buy more things made in Germany or Sweden?
 
drowsy12 said:
oh I didnt even know it had 3D capability?
Click to expand...
It does. It's just suboptimal compared to a TB for it.

ramsesthenice said:
It definitely took a turn, no question about that. Holy moly.

I don’t want to feed some of the energy here, but potential security risks are very real with any software. All of these programs and planning systems integrate with your broader network and can offer an access point to send or receive data. As you pointed out, none of these are American companies and we are notoriously good at ripping ourselves off. Not sure I see this as substantially riskier with a Chinese company myself.
Click to expand...
Do we think TikTok is bad too? What about all the diagnostic imaging machines this company already has in the US and all the HIPAA information that will have? Is this the random 'line in the sand' we draw? How important do Rad Oncs really think we are? Really ****ing important, apparently, more so than all the Radiology-related patient information that is already in the hands of this company, lmao.
 
evilbooyaa said:
It does. It's just suboptimal compared to a TB for it.


Do we think TikTok is bad too? What about all the diagnostic imaging machines this company already has in the US and all the HIPAA information that will have? Is this the random 'line in the sand' we draw? How important do Rad Oncs really think we are? Really ****ing important, apparently, more so than all the Radiology-related patient information that is already in the hands of this company, lmao.
Click to expand...

I think we can all agree TikTok is bad
 
evilbooyaa said:
How important do Rad Oncs really think we are? Really ****ing important, apparently, more so than all the Radiology-related patient information that is already in the hands of this company, lmao.
Click to expand...
I know you know this, but almost no one cares about amassing health specific patient information. Hackers love smart refrigerators not because they care about what is in your refrigerator but because they care what you keep in your fridge, but because it is an access point to what they are really after which is your broader network. I don't know for sure what information people would be looking for, but I am confident it wouldn't have anything to do with treatment planning or volumes.
 
evilbooyaa said:
Again, America first means we should buy more things made in Germany or Sweden?
Click to expand...
Well, perhaps:
"Buy more things made in Germany or Sweden and less things made in China" is "America first", in the sense than not everything can be made in the USA at the same quality as elsewhere.

Being less dependent on a country which shares less values with the US and is less of a friend (some will call it a foe) to the US, is in a sense "American first".
 
ramsesthenice said:
I know you know this, but almost no one cares about amassing health specific patient information. Hackers love smart refrigerators not because they care about what is in your refrigerator but because they care what you keep in your fridge, but because it is an access point to what they are really after which is your broader network. I don't know for sure what information people would be looking for, but I am confident it wouldn't have anything to do with treatment planning or volumes.
Click to expand...
Fair enough, but why do Rad Oncs seem to think that ONLY Rad Onc systems would be susceptible to this, rather than the Radiology systems that are already in use nationwide?
 
You must log in or register to reply here.

Similar threads

D
Is this just the beginning?
Replies
14
Views
2K
WildRivers
W
F
Varian Executive Partnership "VEP"
Replies
20
Views
4K
ramsesthenice
R
TheWallnerus
Is Sameer Telling the Whole Truth and Nothing But the Truth?
2
Replies
79
Views
7K
TheWallnerus
TheWallnerus
Y
  • Poll Poll
AI is Evil
Replies
28
Views
3K
temujim
T
Palex80
Breast is the worstest x5?
2 3 4
Replies
153
Views
15K
A_DeMichele
A