Description

This is an audio-only podcast episode.
Listen to VETchat on Spotify here.
Listen to VETchat on Apple Music here.
Joining Anthony for this episode of VETchat by The Webinar Vet is Neil Shaw, founder of SignalPET. SignalPET’s advanced technology utilizes machine learning and artificial intelligence to assess radiographs in real-time for normals and abnormals on 50+ radiographic tests. You take x-rays. Your images instantly get reviewed by SignalPET's AI. And you get instant test results in addition to a custom client-facing radiology report. Help your veterinary team provide faster intervention, less expensive care, and enhanced patient outcomes during radiographic procedures. In this episode, Anthony and Neil discuss the benefits of Artificial Intelligence and the common challenges in a veterinary practice that AI can help with, such as; time management, lack of confidence and tunnel vision. Neil explains how SignalPET works, how they developed the system and what is in the near future too. They also discuss how the system continues to learn itself and how we can improve as vets as a result of its help.
Learn more about SignalPET here  
Thank you to SignalPET for sponsoring this episode.

Transcription

Hello, it's Anthony Chadwick from the webinar vet welcoming you to another episode of the top UK's veterinary podcast Vet chat. I'm really pleased and honoured today to have Neil Shaw in the room with us, Doctor Neil Shaw, who is a University of Florida graduate, then went on to do his internist qualification also at Florida as well. Like myself, founder of the Blue Pearl Veterinary group, which grew as a, as a specialist referral group of hospitals until it was sold to Mars in 2015.
He was then the chief medical officer there. I think it started like a lot of founders and creators do, having itchy feet, wanted to do something new, discover something and has has been working very hard over the last few years to develop signal pets, and Signal Pets uses very advanced machine learning and and artificial intelligence to actually assess radiographs in real time. For normals and abnormals, and has obviously become a real help to veterinary practise, you know, we, we've just gone through and are still going through the pandemic where travel is difficult, it's difficult to get to specialists.
So the fact that people can have radiographs, vets can have radiographs interpreted using AI is fabulous. One of our value words at Webinar vet is innovation. I love innovation.
It's a real honour to have you on, Neil, and yeah, thanks for taking time out of your busy schedule to speak on vet chat. Thank you. Thanks for having me.
It's, it's great to be here. Hi from sunny Dallas where It's running about 41 °C consistently this week. So it's a little bit cooler.
It's warm. It's warm. We're having a bit of a heat wave here when we get into the mid-20s.
So yeah, 4040 is hot. It's it can safely be classified as hot, I think. Yeah, it's it's hot.
I said, things are otherwise wonderful here in Texas. And fantastic business that you're developing there, Neil. You know, one of my German locums used to say to me, if there are two vets in a room and they both agree with each other, one of them is not a vet.
And I think the problem with, with anything that we do is that it is standardisation that we all agree with each other and actually it doesn't happen very often, so. I suppose one of the great things with with AI is that it can really help with that standardisation and. The fact with all of this AI AI training obviously used in human medicine as well, it can work so quickly compared with an ordinary ophthalmologist, radiologist, or whatever, can't it?
Yeah, no, it's, it's, it's actually amazing, and I remember when we, we spent several years developing the system and we ended up creating a strong system to, to learn the algorithms and and patent in that system and it was a comfortably. It's comfortably at the state of the art in veterinary and human medicine. The initial challenge that we had in developing the system was that no one veterinarian.
He looks at the radiograph similar in similar fashion to another veterinarian, whether it be a clinician, whether it be a radiologist, whether it be a whoever, and, and that was actually a challenge in developing the system is because the system, the algorithms, artificial intelligence likes consistency. Yeah, so we needed to create that consistency for the system to learn. And we, we managed to do so and are able to reproduce those results literally millions of times over.
And it's, it's been a, it's been a fantastic experience. It's really been, it's, it's been absolutely amazing, yeah. Neil, what made you go to radiology, because as I said at the beginning, you know, this could be used in pathology, it could be used for backs of eyes, it could be used in any number of areas.
So what was the reason for, for looking at radiology? 33 reasons. One from a from a veterinary perspective and the other 2 from a, from a very pragmatic perspective of why we chose radiology first.
From a from a veterinary perspective and coming from the specialty medicine background. Having films interpreted by radiologists is significantly limited by a lack of availability. So it, it's simply the, the radiologist to read the films that are produced, there's a mismatch there there are not enough radiologists and so they're working those that are are reading are working very hard, but simply there's not enough of them.
So that that created a backlog. So the need, absolutely the need was there, the profession. Is asking for that need to be filled.
So that's, that's a veterinary perspective. From a technology perspective, there, there are two reasons why we chose first to go into radiology. One is A picture, specifically a radiograph, is made up of pixels.
Pixels are numbers and and the computers today are very good at assessing numbers and looking at numbers from all different angles. And, and so it's a natural adaptation for the systems today to, to look at digital images. The second is a very pragmatic reason most machines now are digital and most machines are connected to the internet.
So whilst if there were a cytology system, an image acquisition machine would needed to be added to the veterinary practise. What when we start with radiology, the image acquisition machine is already there, and so the hardware that's needed, we could, we were able to do this completely from a software basis, a software as a service connected to the internet. No hardware was needed to be added to the veterinary clinic there.
Be ready there. Unless of course you're still er developing your X-rays by hand by sticking them in the tanks. Do you remember doing that, Neil, or are you too young to remember that?
The the youngsters listening won't remember, but you used to have to go into a darkened room and stick them in different tanks to get the. The picture, didn't you? My father, my father was a general practitioner and I grew up being sent into the dark room in the dip tanks.
And, and not only that, more than once I left the, I forgot what it was called, the safe open and then opened the box and exposed the film to light and And I was advised on how much those films I just. And yes, it was I remember those days only too well. Yes.
So the history goes back a long way and of course you probably used to lose a month's pocket money if you did something like that, didn't you? Without a doubt I need the front lawn that much more. One of the points you made was this is the first because of course with that technology and with it being pixels and numbers actually.
You can do cytology. You could do backs of eyes. It just needs the piece of equipment when and if you decide to do that.
Correct. I would see I would say expanding and I and I intentionally say that radiology was the first area that we tackled and it's worked out very well. It's it's worked out extremely well.
I would anticipate moving into other areas such as cytology. And then, and then the other is actually through the the, the critical assessment of, of data of laboratory data, physical exam data of radiology data, and the assessment, and there's no reasons not to have machines augment and make and make us smarter. The way I look at it is a kind of like a movie with Iron Man is this never is a veterinarian.
This never replaces an individual. But it allows augments them, augments and put on a power suit more efficient, more powerful, yeah. It's that whole internet of things, isn't it, because now wearables are becoming very popular so we can actually take data from the dogs, know what their diabetes curves are doing, do they need a bit more, a bit less.
There's so much stuff that, that, you know, is possible, isn't it, in the next 5 to 10 years and is already probably happening in the, the medical field. It's already happening in the medical field. It's it's now happening in the veterinary field and it's, it's, it's absolutely fantastic and .
It what it ultimately does is it improves care. It promotes the quality of care, and frankly, it lowers the cost of care. It actually lowers the cost of care, which means it makes the care available to more pets when you, when you lower that when you lower cost as a barrier, yeah.
This was very much when we set up webinar that Neil, it was digitising information, it was digitising knowledge, and of course once you digitise it, you democratise it, it becomes, it gets out onto the internet, it demonetizes it to a degree. It can go to all countries of the world, you know, this system that you have now is you're in Dallas, but it's not just limited to Dallas or America. It's very much something that can be embraced by the world and will improve that as they use the system, won't it?
Very comfortably, very comfortably, and to think that, that we can take a, a perhaps a body of knowledge and a body of experience, perhaps what we're doing here in the States and export that to another country instantly in a scalable fashion. Is a is a strong asset for the profession, a strong, a strong benefit to the profession. Yeah, one of our vision is to have the world's most confidents within our platform and obviously when you have a radiograph and you perhaps look at it and think this is what it is, but then somebody at AI checks it for you and actually tells you what you thought was right.
That also helps to improve your confidence, doesn't it? Right, because the, the goal is to point is for veterinarians to be more confident, to have more, to have more, to have more skills that are available to them and the technology, the technology enables it. And, and the veterinarians, the veterinarians become more efficient.
And one of the, one of the things, one of the challenges that we're, we're managing through in the states. Is a is a lack of veterinarians to meet the needs and a lack of support staff to meet the needs and technology is actually the obvious solution to help with those shortages because it doesn't work. It absolutely it makes it more efficient.
It enables, for example, it enables technicians to actually do much more under the veterinarian supervision. And again, everybody wins, and there's a a backstop of quality control as well. We have a very fine diagnostic imaging specialist in the UK called Mike Herage, he was dean of Cambridge Vet School and he will do radio radiographic reading sessions for us on webinar vets, and he will spot things that I don't spot and I think what you can have as a GP sometimes with radiography is a bit of a tunnel vision.
You think you know what you want to see, so you're looking at that area, but of course you're not looking at, Right. The thing that actually is significant because it's not within your head to think about it. So you can really have that tunnel vision when reading radiographs, can't you, if you're not careful.
Right, it's, it's it's, it's easy for that to occur. For example, a patient may come in hit by a car and there may be changes to the lungs, maybe some air in the chest cavity, a pneumothorax. And we're concerned about the pneumothorax.
We're concerned about the lungs, 100% appropriate. We're concerned about the breathing. At the same time, we may forget to scan the film quickly and notice that the scapular is fractured or the party is fractured, which plays into the management of that case significantly in addition to managing the thorax.
Other examples that we've seen, and these are actual case examples that keep coming back to us every week are. A patient comes in to a GDV and the system of course confirms that there's a GDV. But there's so much excitement about the GDV.
There's the fact that there's a nodule in the lungs may get overlooked, but the system doesn't get excited about the GDV. It causes it and at the same time scans the lungs for nodules or masses so that it's, it's complete. Yeah.
And then of course the management of the case. May be different because we may decide not to operate on that dog but actually euthanize it, for example, mightn't we? Right, right, it'll change the course because we could, we could operate and and successfully treat and manage a life threatening disease, which is the gastric dilatation ovulus at the same time that patient only has several months to live due to what's going on thorax.
And so that's, that's an informed decision where the client. Can make a decision whether to, to do a life saving measure or to stop the process. And it, it, it's, it's important to have that information before the surgery, not after.
Yeah. Neil, again, you know, loving what you're doing here and, and obviously I've read around the subject as well, it's something that interests me being involved in the internet with the training that we do. I, I think the great thing also about AI is.
You've probably found that the machine actually begins to teach itself so it improves the more plates that you stick through. Have you seen that with your own system as well? No, it's it's a great, it's a great point.
I remember we were in the development phase and we spent several years developing the system and the initial goal was to create, create a system that That could meet what we did, we could read as clinicians. So can a system meet the current level of expertise as clinicians? And then we came to a realisation that the system was becoming critical of us, which is a very humbling thought.
And that our review of radiographs was being evaluated by the system of whether or not we were consistent. And, and that's, that's 180 degrees difference. So now we went, and this is in the development phase, we went from saying, can the system meet our current needs to Oh gosh, I hope I'm getting this correct to demonstrate that I'm consistent on this and, and that's a, it was a, it was one of those moments that I remember exactly where I was.
I remember exactly the time of day. I remember, and I'll never forget it. And, and so, which is, which is the power of technology and we read about it, but only when we apply it to, for myself, only when I could apply it to things that I knew, which was veterinary medicine.
Did the magic appear and see the magic of technology. So again, I would stress I don't see the system replacing us as humans, but I see it significantly augmenting our abilities and what we can do and and and that will progress and these systems will become better and better over time, no doubt, you know. All of this digital technology, I think leverages us, doesn't it, because For us to look at a plate properly, you know, with the several views probably takes 1015 minutes.
I remember reading about this, I have glaucoma and and studied, you know, ophthalmology being used for the same sort of thing that you're doing. And. The speed at which you can actually do that is, is truly amazing and presumably continues, as you say to quicken up as it understands more and actually can then be critical of, of humans, but what it can't do is then pick up the medicines and actually treat the dog or the cat with the relevant medicines.
So as you say, it's, I, I, I, I agree with you. I think there are some jobs now that will be lost to automation and to robotics and so on, but. Jobs like ours, I believe, as you've said, we will become better by using computers and robots to assist us.
Absolutely, we, we will, we will become better. We will be able to provide greater value to pet owners. We will be able to provide greater care and what that is going to do, it's going to raise the expectation of what's the minimum standard of care.
That should be provided at at most general practises. And, and we're going to raise that standard of care. Veterinarians are going to be enabled by technology.
Pet families, pet owners are going to come to expect more and we'll be set up to provide it and at, at the, at the level of the primary care practise. And there's, there's, that's, that's, that's called a win-win. Everybody wins in that scenario.
And I would actually if we look to the future, I would look at I look at technology expanding our abilities beyond radiology and you mentioned in cytology and in critical thinking, technology will aid us to provide better care and that better care will become an expectation of pet. Well, as you say, retinal scans and ultrasonography and everything. So today, if you can take a, take a few, say 3 views or 2 views of films and almost instantly, the, the interpretation is available to look at along with the films.
If a technician takes it and says that the films are already with the, with the interpretation and the veterinarian can look at the films with the interpretation that speeds up the process, that speeds up the knowledge. And it delivers delivers better care. So it'll go to other areas, yeah.
And I think we probably should go over it for those people who perhaps haven't been thinking down this route yet. You know, obviously, as, as you said, you began to train the AI but actually over time, it's begun to sort of actually learn by itself and then actually teach you, Neil, with your, with your great knowledge in this area, so it, it, it improves all of us with this sort of strategy and you know, as you've said it. It it gets better than the specialists who've started the training of it.
It really, that that's the ability and it never forgets. And with more exposure and more knowledge, it sharpens its skills and sharpens its skills, and which is fantastic. And the other thing when you think about it is we're talking about a diagnostic technology, specifically, taking films, X-rays, radiology, taking films.
And we're looking at now at creating greater value out of the process of, of taking radiographs without having to do a CT, without having to do an ultrasound, with without without additional major expense, we're able as a profession to revisit a fundamental technology that exists in the vast majority of veterinary practises and gain more value out of it. And that they that's when. That's, that's, that's absolutely a win, yeah.
How are you finding Neil, because it's, it's great to have this AI technology, but if I don't place my patient correctly on the radiographic radiographic table, if I don't actually direct the radiograph at the right area. Then I will get the right results. So does, does the AI also comment on, on quality of of the actual image and how that's been created?
Great question, great question. Yes, yes and no. So when we, when we trained the system, we trained it on real life images.
In other words, All the images weren't exactly perfectly placed. There perhaps was a little bit of movement. Perhaps the patient wasn't exactly lateral, a little bit turned, and so the AI learned to look at it in a real life fashion.
The AI also looks at the images in a three dimensional, its own three dimensional view, not through human eyes. So it can it can adapt for, for minor alterations. And so the AI is actually Better at that and a good rule of thumb is if the image is, is for most clinicians, if the image works, it's diagnostic for most clinicians, it absolutely is diagnostic for AI comfortably brilliant.
We do the no part is we actually collect that information and we, we assess it ourselves on, on quality of images. We don't currently provide that back, but I anticipate we will be in the future. And where it'll help will be, say, for example, with the nurses or technicians who are taking the films, because then that's one additional step the veterinarian doesn't have to involve themselves in to quality control the films and approve them.
They, the nurses can do that directly with the system without having to pull the veterinarian out of a room simply just to approve films that they're adequate. Yeah. Obviously, you know, we're talking about this technology in veterinary medicine.
The great thing is, it's already here, isn't it? I know you've been going since before the pandemic actually offering the service and How many, how many sort of, radiographs have you interpreted over that two years? It's been a, a busy time because of course people couldn't get to specialist centres as as easily and so it was very much.
Get them to the vets, we couldn't really travel that the vets was close, you get the X-rays, you can then get an interpretation rather than having to go off to a specialist centre. Yeah, so we, we spent, we spent several, several years developing the system. And then, prior to releasing it, and we released it, we were at the meeting in Orlando, the big, the big North American meeting in Orlando, announced it and then COVID hit, so there were no more meetings.
And, currently, that said, we've, we've, we've continued on a ramp up. So currently on an, on an annualised basis, we will, we will review comfortably more than 2 million films a year. So annualised, it's a continuous ramp up.
And so that that that comfortably puts us into one of one of if not the largest reviewer of films and, and that's that's, that's consistently growing, and so it's a Again, the, the benefit of technology is the scalability. And so the, the nickname, the nickname for the system that that some in-house have called it is Adam. So the system is called Adam.
Initially we said it should be called Eve, but it turned out to be called Adam and Eve, but it turned out to be called Adam. And they say Adam is very hungry. Adam wants to look at more and more images.
He doesn't get. Yeah. And so he's he's he's he's a hungry.
Well, I suppose this is where, you know, this is the future we can do this, but what's maybe the next future, what's in the Near future for Adam, hopefully you're going to keep apples away from him. Yeah, I think it's we'll keep the apples. I think that we will the next, if we look at, look at radiographs taken in in a primary care practise, in most primary care practise, most patients do not receive the benefit of films.
Specifically, the percentage of patients that based on the figures that we've determined, the percentage of patients that receive the benefit of diagnostic films is comfortably less than 5%. So less than 5 out of every 100 patients that present in a month or a year receive, receive radiographs and I would anticipate this number going up quite a bit. And Whether it's assessment of the sick patient, the patient that's limping and vomiting or coughing.
Or not eating or it's becoming, becomes part of the standard evaluation for older patients. I would anticipate that that that percentage or utilisation of of radiology to go up because it's a benefit across the board. It's absolutely a benefit.
And so my anticipation of what's next is I would like to see our our profession move to a greater standard of care. From the, in the wellness, in the wellness arena, in addition to providing it when a patient is, is sick. And, and again, Neil, remembering the tanks that even as the vet I used to go in and, and, you know, push them through, which was using veterinary time, this is going to be a real opportunity to upskill veterinary technicians, veterinary nurses who can take over that process, some of which can be done with sedation.
Obviously if a general anaesthetic is needed, there'll be a bit more veterinary involvement, but if we can improve. The actual quality of the X-rays taken from a positioning perspective, that's gonna mean it could improve things more, but also increase the profile of the veterinary nurse, the veterinary technician. 100%, 100%.
What, what it does is it engages the nurses more, the technicians, we say technicians here in the states, nurses elsewhere, it engages, it engages them more. They, they become more involved and rather than just simple restraint or taking the films and submitting it for review, they're now empowered to, yes, that's a good quality study, let's move on. I'll, I'll approach the veterinarian once the study is complete.
They can also do, also, they're starting to learn to read the reports, the nurses. So it's very different because it's almost instant. So it's very different when that say there's an air in the in the chest cavity, a pneumothorax.
It's very different for a nurse to say a doctor that That patient that's in oxygen and not breathing well, it's X-rays are up. Can you please take a look when you get a chance? If the veterinarian rather hears, Doctor, the patient that's in oxygen and not breathing well seems to have a pneumothorax, I'm beginning to prepare to do a thoracocentesis to respond.
Can you please look at the films and report, confirm that it's pneumothorax, and then we can treat the patient. Yeah, it's a different conversation. Different conversation.
It's better for the veterinarian, it's better for the nurse, everybody is more engaged, the patient ultimately benefits. Absolutely. I think as a final thing, you know, Neil, first of all, you know, thanks again because I really enjoyed the chat.
I love talking to innovators. We in the UK have a shortage of vets. We have more pan pandemic puppies and kittens coming through, so the practise is so much busier.
And so we have less time, so these sort of time saving strategies. If we want to be a successful practise, we almost. Have to embrace them to, to stay ahead and and to be at the head of the curve as change comes.
Sometimes as vets, we're conservative, we don't like change, but actually change doesn't wait, does it? Change happens and those who embrace it usually do the best out of it. Without a doubt.
There's always the, the early adopters per se and they and and what we're seeing is we went through the early adopter phase and now as a profession, it's. Yeah, and we see we see that happening more in a whole scale fashion, not just the early adopters and it's, it's frankly, it's becoming, becoming the new standard and so folks will adopt it as they understand it, it's a new standard. It might as well be, there's no, there's no downside, so it might as well be part of it.
Yeah, yeah. One of the disadvantages, Neil, of running webinar that or being very involved in it is that I don't practise anymore, so I've not seen it out in the open, but I'm sure there are some people listening to this podcast, you know, who aren't aware of the company. How do they get in contact with you to, you know, get the technology and how does that connection happen?
So, you know, we've got a CR or a DR. We take a picture. How does that get sent over to your place and how do we get the results back?
Perhaps just tell us a little bit about the practicality. OK, thanks. Yes, happy, happy to, thanks.
There's no hardware, so folks can reach out through the website. It takes several minutes to onboard. And then what happens is on board is we, we connect with the, the, the in-house pack system, the in-house image handling system.
We do it behind the scenes online. And again, it just takes a few minutes. And so then we receive a copy of films as they're taken and then that and then the interpretation is available.
It takes about a 20, maybe a 30-minute orientation for the vets and the the staff. We do it online with a webinar and you're up and running and you go. Very, very straightforward, very simple.
We have a fantastic team in the states and actually also in South Africa that helps onboard folks and they, they're very good at it. And so just reach out and we'll get folks on board it and and really it takes no time at all. Yeah.
Fantastic Neil. Neil, thank you so much for your time. I know how busy you are, but I I've been fascinated to hear about the technology that read about for a number of years about the possibilities and it's great to now see that becoming, you know, actual.
Science fact rather than science fiction. Yes, no, it is all scientific, and the papers will progressively come out. It's very much a medicine and that's that's part of the beauty of it.
Yeah. Thanks Neil and thanks everyone for listening. Take care.
Bye bye.

Sponsored By

Reviews