Paediatric Post Mortem Imaging with Dr Susan Shelmerdine

Hello and welcome to another episode of Radcast. Today we're joined by Dr. Susie Shalmadeen, who's a consultant pediatric radiologist at Great Almond Street Hospital. She has interests in academia, forensics and AI, and she's also co-founder of a company based on a niche subspecialty interest. And has recently completed a radiology outreach trip to Nepal with RADAid and Hexarad. So we're going to get into all of that and find out how she manages to keep all those plates spit.

Yeah, absolutely. Um quite the impressive resume, so lots to get into. Um so Susie, can you um just start by introducing yourself to the list? Yeah, hi everyone, my name is Susie Shelmerdeen, I'm a consultant pediatric radiology working in Great Ormond Street Children's Hospital in London in the UK. Um, as you very kindly said, I do have quite a lot of interests, um, mostly in pediatric imaging, body imaging, to be more specific, but also post mortem and AI.

I've been working there since twenty sixteen, um, when I was a research fellow and then a PhD s research fellow and then an academic research fellow and then finally a consultant since twenty twenty one, so quite a some time. Wow. So uh you so you're part of the furniture there now. And then before that, um you did what a fellowship in um Toronto, right?

That's right. So I did my specialty training in radiology within London in the UK and then after my ST five year I went to Toronto to the hospital for sick children in Canada and um spent a year doing body imaging there. It was a great experience, um, but a huge culture shift and obviously extremely cold. Another big name on the C V though. Yeah, absolutely. And then when you came back, um you you didn't come back straight into a consultant job, you then did some fellow positions.

That's right. And I think that's a really interesting thing to talk about. Um so when I was out in Canada, I hadn't really secured a job to come back to in the UK and lots of people had said, Oh If you could secure an NHS job to come back to, that will give you some security when you go away.

But I didn't want to go away knowing that I would come back. I wanted to go away thinking the world's my oyster, I can go, be, do and have anything I want, you know, and not feel like restricted to any particular life path, even though I did end up coming. But the intention when I left was to have all doors open and all options open for me. But then whilst I was out there, um

I had a lot of different opportunities. I could have stayed in Toronto. There was a job going there. There were also jobs going in Singapore, in Hong Kong. But I think even after having done all my training and a fellowship, I still felt there was more to learn and more I wanted to experience. And I felt that at that time if I settled for a consultant job

Not that the learning would stop, but I felt like I was then responsible for teaching others rather than having more learning given to me. And I really felt that I wanted to expand my learning in academics. radiology and learn about a niche area that I could call my own. Hence the reason I came back to Gosh because there was a research fellowship with a um very eminent um radiologist, Professor Owen Arthur.

who I had gotten on well with during my time at Great Woman Street before, who was very willing to supervise me through a PhD. And hence I chose that option. So I guess that's something that um radiologists out there who are training and thinking about the next step in their career might want to contemplate. Like what learning do you want? What experience do you want in life and what's going to give you that as your next step? Yeah.

So like what attracted you to pediatric radiology in the first place? That's a really good question. I think In initially it was because I couldn't settle on a subspecialty, which sounds really bad. But the the truth is that pediatrics is so general So it wasn't like I had to focus on any one particular body part, which I found very attractive at the time as a trainee where you were very confused, you don't know what you should be doing or not doing.

So I thought, well, at least with pediatrics I still have all options open. I can still do a bit of chest, a bit of MSK, a bit of abdomen. And also children are so much fun and they're so engaging and they're so like positive. And on top of that it was also a multimodality specialty. So I wasn't cutting off, you know, any particular s um imaging type. So I could still do fluoroscopy, ultrasound, C T M R I. So that was what attracted me to it, the variety and also the positivity and fun of it.

And then the more I did, the more I enjoyed it, the more I felt really good about it. And also I know this sounds a little bit selfish, but a lot of people are scared of children and scared of imaging little ones. And so when you become quite good at doing pediatrics, people then

I don't know, almost revere you, like, oh my god, you've got this special skill that no one else wants to do. So you feel really proud and really like unique. So in a way it also made me feel quite special. So it's partly to do with the imaging, partly the patients and partly in my own self. you know, for centeredness, I suppose. All very good reasons. Indeed. Yeah, Jamie Jamie's allergic of children. Allergic of children.

Um so presumably working at a place like Gosh, everyone is super specialized. Uh so what are your sub specialties in My subspecialty interests are um musculoskeletal in the beginning, but that has now evolved into forensic and post mortem imaging and they're now evolved even further into AI for musculoskeletal pathologies.

Um, I think you're right. When you work somewhere as specialist as gosh, it can be really hard to find your place and your niche in that area when everybody else is so good and so advanced. So it's almost like you have to carve something out for yourself. Um So this is what I've ha started carving out for myself in the few years I've been there.

So where did the um forensic radiology interest arise?'Cause I mean I we we don't get any exposures to that really in training and I imagine it's like not the most appealing or popular area even in pediatric radiology. It's not. However, I feel that in pediatric radiology compared to other subspecies in radiology

you do get a little bit more exposure because we do have to learn to report skeletal surveys for suspected physical abuse. So even as a fellow you are a little bit aware of the forensic aspects and some medical legal aspects of

uh imaging. Um however, moving on to post mortem imaging, I then saw a lot more of the forensic sides. For example, in the post mortem CTs that we do for children who have been found, you know, Through sudden unexplained infant deaths or sudden unexplained children's deaths or, you know, unfortunately some children do commit suicide, we are having to do a lot of post mortem CTs as evidence for the histopathologist.

And so that is where that sort of started to arise. There is a need for it and there's clearly not a lot of people who can do it. So you soon become quite expert in this area without trying too hard. So how much of your day is actually spent doing that kind of work? Well, at the moment as an academic radiologist, I have one day of clinical work and four days of academic research work.

So really the one day that I have clinical is the day that I spend doing live children as well as post mortem imaging. And most of the time we don't get a huge amount of post mortem cases, which is good. We don't want children to die, but um but I would say we get about Three post mortem CTs every week, roughly, maybe sometimes fewer, sometimes more. Um and we do quite a lot of micro CT and post mortem ultrasound at Great Ormon Street for stillbirths and miscarriages.

But the post mortem CT is more for older children who may have died unexpectedly or where there's a forensic suspicious death um implicated in that.

So, I mean, I can imagine how this sort of work can be very sort of challenging psychologically, particularly like postmortem ultrasound. How do you deal with that? Yeah, thanks for asking. I think it's something that not many people think about when they first do post mortem imaging. Um, but you're right. I was very lucky that I had a very supportive team when I first started doing this work, who said to me, you know, if at any point you feel you can't do this or

things change in your life, um, then please say and please let us know. And I think having that good support network and knowing that you don't have to push through and do something you're uncomfortable with really helps as well. I think also knowing that there's a family at the end of this who just want some closure really helps me as well um push through and know that what I'm doing has a purpose.

At the end of the day, a loss has happened and anything you can do to kind of make a bad situation a little bit more bearable for families.

um is really, really important. I think um it's not for everyone and I think if people are thinking of getting involved in post mortem imaging, then it's really important that you do have a bit of a yw'n yw'n yw'n yw'n yw'n yw'n yw'n yw'n yw'n yw'n yw'n yw'n yw'n Personally for me when it's dealing with fetal losses and pregnancy losses, I find that a little bit more easy to cope with because

I almost feel like these children never had a life and but I find some of the older childhood deaths a little bit harder because they obviously had memories, they had a family, they had you know, a lot um riding on them and friends and, you know, all of that. And so that that can be quite hard. And and sometimes it's also the stories, you know, you may not find the whole

dealing with a dead body that's unpalatable or you might be very used to it, but then knowing the circumstances of the death puts some emotion and puts it into some sort of framework for you which can be very emotive, you know? So, um It's a bit like, you know, when you have to do cadaver dissection in medical school in some places.

maybe just seeing a cadaver isn't necessarily a motive, but then knowing about that person, their family, their likes, their hobbies really then brings them to life and makes it hard. So that can be hard sometimes. And you mentioned that you do um post mortem ultrasound and and your PhD was in it's called novel applications of ultrasound techniques in pronatal and infant there. So are you using um like techniques from that in your day to day work?

Yeah, we still um do post mortem ultrasound for some of our fetal uh losses in Great Woman Street. So the reason we started pioneering doing ultrasound for these um baby losses was because MRI is the best standard of imaging for this, but not everybody has access to an MRI, and even if they do,

They're reserved for live cases. There's a huge backlog, as I'm sure you're aware, of MRI um slots. And so we wanted to find an alternative technique which would be almost as good or an alternative to MRI, which is why we started doing ultrasound. And actually um we found that ultrasound was comparable to MRI in terms of diagnostic accuracy if the images were of diagnostic quality.

The problem we have with ultrasound is that a lot of the time we're not necessarily getting the most best diagnostic quality images because of decomposition, because of swelling, because of gas. But um it is a viable alternative if people are trying to start a post mortem imaging service for fetuses and um stillborn children but maybe don't have access to an MRI.

It's never going to be as good as MRI because that is the best imaging we can do, but it's it's something at least. So what sort of causes of death can you diagnose on post mortem ultra cells A lot of the things that we're looking for are congenital abnormalities that may not have been spotted during the antenatal imaging. Or maybe they were missed. So for example it might be abnormal renal um systems or perhaps it might be just a lot of ascetes from fetal hydrops.

Um, we can also look at the brain. Um, you m you're probably very familiar with the fact that we can do cranial ultrasounds and neonates and we can do that very easily in um stillborn children as well. So we can look for any sort of, you know, absent uh corpus callosum or any sort of malformation, maybe Dandy Walker malformations. It as I say, it will never be as good as an MRI in terms of looking at the anatomy, but it gives you some indication as to what's going on and Can be um

some tools of helping you decide whether or not further imaging is required or whether or not, you know, you can say that fits in with the antenatal imaging, or maybe or not we need to go to an autopsy. But um those are the sorts of things we look for. It's very hard for us to tell whether or not someone's died through infection with an ultrasound. Um so there are still things we can't tell, but mostly it's structural abnormalities we're looking for. Big structural things, interesting.

So, aside from your clinical work, you've actually taken this subspeciality interest to the next level and founded a company called Anubix. Can you tell us the story behind this? Yeah, absolutely. So um My co founder, Natasha Davendralingham, did a masters in adult post mortem imaging up in Leicester when it was still running. Unfortunately that degree has now subsequently um

been discontinued. However, she has this skill in postmortem imaging in adults and I have this PhD in pediatric imaging. And there was a growing need for postmortem imaging now, partly driven by the fact that a lot of families

find autopsy unpalatable, especially a lot of religious groups, but not just religious groups, you know, it's very emotive to think of your loved one being cut open after death, even though they're dead. And secondly, um, something that I think not many people are aware of is that We do have a shortage of all doctors and shortage of pathologists.

But in pathology training now it's not compulsory for them to know how to do autopsies. So many pathologists are now choosing not to do autopsies or not to learn how to do them anymore.

So imagine if we had radiologists who could choose never to report X-rays or never to report MRIs. We would then have a huge backlog of very few people who can then report those specific things. And that's the same in pathology. And so there's this drive to see how we can replace or help some of the pathologists who don't do autopsies or who do it but only in selective cases with imaging as an adjunct or even as a replacement.

And so um Natasha and I really wanted to see how we could expand those services and teach more people and drive service um change within the NHS. However, we were finding it was very, very hard to do that within our um job. But it seemed much more

I don't know, streamlined, more effective to try to do that as a company, as an external provider. And so that's how the company was founded. We initially wanted to make it a charity, but we were advised that as a charity we couldn't be sustainable and so we needed to find a way to make this financially viable if this was a service um that we wanted to provide an education we wanted to give to other radiologists.

Hence the company was founded, and it's only been around since May of this year, but already we've had so much interest. So many people wanting to join to learn how to rapport post mortem imaging and so many local authorities and coroners who have expressed interests in how they can set up their own local services and how we can consult for them and also provide reporters for them.

Oh wow. That's uh yeah, really really cool. Um So uh y you originally wanted to set up as a charity and that so w what what were the issues with that? Well, we felt very bad about asking people for money in order to help them with their bereavement and we really felt that it was alm I think within the NHS and as doctors you almost feel guilty asking for money.

And so that's why we really wanted it to be a charity. We wanted it to be externally funded through donations. But the problem with that is that you are then very much reliant on donations and Um that that may not be a sustainable long term solution to providing a service of care because the best the the vision for the company is to make sure that anybody anywhere within the country who needs a post mortem imaging scan can get one and get one reasonably easily and to have that sustainable and

if we were trying to do it as a charity, it may be that we can only do it in certain places or for a certain amount of time or only in certain select cases and we wanted it to be more open to everyone. So it seemed like the viable solution was to make it

a business. So that that's why we had that um discussion. If it's successful you can always sort of reinvest the profits in charitable um areas that to sort of appease that side of your brain that wants to be um that doesn't want to make money off of a sad situation. Absolutely. We've already identified quite a few different charities that we would like to sponsor once our business becomes profitable. Things to do with bereavement counseling, as well as helplines to help with suicide support.

Um, things like that and also miscarriage and stillbirth charities as well. So we've identified a few things where we would like to develop and grow and give back to the community as well. So we hope that we can make this something that's beneficial for everyone. And the n the name is really cool. Um, where does that come from? Thank you so much. Um yes, so we didn't want anything to do with death in our business title.

Because that sounds super emotive and we really, really want the message not to be about bereavement or sadness, but to be about empowerment and serving families and people and the dead. and giving a voice to the deceased, finding out why they died. So Anubis, uh spelt with an S at the end, is um an Egypti uh Egyptian god of the dead and one of the roles of Anubis is to help empower um the dead into the afterlife and help them

have a peaceful afterlife in some ways. So um we felt like having a God who was there to serve the dead would be a very empowering sort of title to have without necessarily mentioning death, bereavement, sadness, loss in um a company title. And we didn't want it to be very clinical either. You know, we didn't want it to be medical. We wanted it to be more emotional and helpful and, as I said, empowering. And then I guess the X at the end was just um to give it a stronger name.

I think having an X in the title most of most companies does sound a bit more powerful. Um and also, you know, gives it the kind of flavor of being c innovative in technology, you know, like SpaceX and X for Twitter units. So y you went to the Elon Musk School of Marketing then? Well it sounded it sounded more like, you know, cool and forward thinking. So You're probably the only person who prefers the name X to Twitter, right? You're like, Oh yeah, it's a master stroke.

Well, I didn't say I preferred it but it was it's it's more futuristic, I suppose. So that's And the title. Um yeah, and obviously all of us in um the company are big, big animal lovers and so you know, um Anubis, the god, is depicted as a black hound. Um so, you know, it it fitted in with our love of animals too. So is the is the reporting the main the main activity that you do, or do you think it's kind of primarily education rather than reporting?

The reporting is the thing we want to be known for. We definitely want to be known for our consulting services and helping advise people how to set up a service that's um workable for their particular local authority and for their population and also the reporting because at the end of the day it is empowering the families and that that's what we want to do. And so through the reporting we give that answer to the dead and to their families.

Um but the education is a big, big part of it as well. It's probably a secondary part of the business, but a crucial component because teach them and at the moment there really aren't that many courses that allow people to learn how to do post mortem imaging. I mentioned Natasha went on a master's degree course in Leicester, but unfortunately that no longer is running. Yeah.

There are only a couple of two or three day courses that are run around the country but Honestly, a lot of the feedback I've been getting from radiologists who have attended those courses is it's great, but that doesn't give them a lot of confidence to then go out and do their own thing, you know? Once you do a two day course you probably still feel like you need a bit more support or mentorship or guidance or practical ability and

Post mortem imaging, although it's on the rise, doesn't actually happen that often. It's not like your busy trauma CT list where you have loads and loads of cases all the time. You may get only a few a week. And so keeping up that abil yes, thankfully, but keeping up that ability and that skill then means that you go on a two day course, you may not

report any for like months and then suddenly one comes along and someone goes, Oh, you've been on a course, you report this and you may not feel super happy about that. So um what we want to do is is not only provide people that information but some sort of mentorship, guidance and also some sort of double reporting with them until they feel confident in taking that lead on themselves for their local trust or, you know, working with us to do that for other parts of the country.

How have you found the process of founding a company? It's been very stressful and a lot more work than I thought it would be. And I know that sounds very silly to say that, because obviously founding anything should be a lot of work. But there are a lot of things that you're not necessarily trained to do as a doctor when you come to founding a company, such as accounting business setup.

um VAT exemptions and um HR, how do you hire people, how do you contract them? Um also, you know, working with other providers, other businesses, for example, setting up a reporting platform. setting up encrypted email solutions so that when people want to ask you about a case you're not doing this in a sort of unsafe way and you think about cybersecurity risk. And also, you know, just generally working in a team and team management. A lot of ti a lot of the time you start a company with

a group of people and you think you're all on the same page and you find out actually you're not. Not that you're on different sides of things, but maybe the way you envisaged how this would be carried out is slightly different to how someone else thinks it should be. carried out or you want to do certain things and they don't or different people have different strengths and abilities and

you find that out about yourself through this process. So it's been um a learning curve, a very steep learning curve, but a great one. I think it's um Yeah, something that um has given me new skills that I didn't think I could develop any other way. Yeah. As I'm sure you probably have experienced yourself. Yeah, that's very true.

Yeah, like but on a on a much smaller scale. I think yeah, setting up a completely new business that you're definitely gonna develop lots of transferable skills which will be useful in all facets of your life. Um And balancing balancing this with your day job is is that being the challenge?

Very much so. It's almost like I have several jobs, you know, because I've got the clinical side, the academic side, and then this business side, which um has to happen out of hours and on weekends. So having spare time to um have hobbies and To hang out with friends and family it does take a bit of a back seat when you're trying to start something new. But um hopefully with

More time and more experience, this won't take up as much time. Uh, but you know, it's exciting as well and it's something new and definitely needed. There's literally just aren't enough hours in the day, are there? Sometimes I just think like twenty six, twenty seven hours, that'd be perfect, but twenty four

Absolutely. And things always take so much longer than you think they're going to take. Something that sounds super simple. Then you you know, you come up with lots of snafos and you're like, Oh no, I didn't realise this would be a problem. Yeah. Um so on on top of that, you're also an academic radiologist. Um so how much of your time is spent on research um in a week?

Four days out of five are spent on research activities and then one day clinical. So um Oh wow. Yeah, so quite a lot of academic time. And it's definitely needed because there's so much admin.

networking, collaboration, conference attending, presenting that goes on that you you can't fit into half a day a week, you know, that, you know, some people think is enough for research. It's really very involved. Um It was scary going from a full time job to then a part time clinical job to then only one day a week clinical as I can have

you know, ascending this academic career ladder. But and and I do always w worry that my clinical skills aren't going to be as good as many of my peers. But I think you then develop other skills in other areas and realise that that's important too and, you know, you can't be good at everything. But I think that's something that um does worry people going into academia. So I always, always

felt, you know, I'm first a doctor and secondly an academic. And now I I still feel that way, but I do appreciate the skills I've learnt through academia a lot more than I used to. Because, you know, as a junior person you you don't really know what academics do. You don't really understand. You think it's just about writing manuscripts all day. Um but then as you get more entrenched into it, it's about teaching, education, grant writing.

policy writing, fun you know, lobbying for things on a national level, setting up services. And you know, you need someone who's got that experience and knows the evidence to be able to affect change on a higher level and

Who else can do that unless someone who's, you know, entrenched in the academia and knows the literature inside out, right? So it's really important. So have you have you always been interested in research? Like did you do academic foundation program and all that sort of thing? No, absolutely not. And I think that makes me quite different from a lot of academics. I actually shunned academia quite a lot. I never did a foundation academic programme. I never was an ACS.

Yeah, literally the first time I did an well, okay, maybe not the first time, but um I did an intercalated BSC during my med school, but I don't know if that counts as being an academic as such. A lot of people do intercalated BSCs. Um But the first time I did proper academic work was after I came back from Toronto and took up this research fellowship post that um

I explained earlier, yeah. So I'd never really thought about research. I always felt it was something that was going to get in the way of me being a good radiologist and being a good doctor, being a good radiologist was the most important thing that I had to be. Um, so that that was my thinking back then. And I don't think I had many role models in academia who could explain to me the importance of it, which is why I guess I shunned it from a

early stage. In fact it almost seemed like something that was just going to get in the way of other things and yeah. Interesting. And and the pathway to being an academic is not easy. People don't make it easy to do this, which is why if you're following the path of least resistance, then why would you go out of your way to do something, first of all, that's quite challenging to do? Second of all

doesn't necessarily make it easier for you to get a job. There are plenty of radiologist jobs out there. And thirdly, isn't necessarily like encouraged. So I do think there is a problem in how we encourage academic work in radiology. It's interesting you talk about um not having many like role models. I've uh when we were discussing before this pod, you were talking about um how particularly in AI, which is one of the areas that you are um researching, there's not a lot of female representation.

Um so yeah, and you were saying that was quite a big problem. Yeah, it is. And um I don't necessarily mean female role models. I just meant any role models at all who spoke to me. So within pediatric radiology there isn't a huge amount of academic work that happens compared to say oncology or cardiac or neuroimaging. Um and and one of the very few role models I had was Professor Arthur's who ended up being my PhD supervisor.

and close colleague now. But I think at the time it was very hard to see how I could fit academia into a pediatric radiology job. And also it didn't seem like there was a need for it. There there is obviously a need for it, but to me it seemed like, well, I can get a job anywhere as a pediatric radiologist. And people aren't saying, I'm going to hire you based on your academic interests. People are just saying, Come and work for us, we'll Yeah, exactly.

have you anyway, so it almost felt like you're giving yourself an extra burden when you didn't need to. But the it's so interesting and doing academia has opened up so many opportunities for me for travel, for collaboration, for attending conferences, for you know, working on cool projects that I otherwise wouldn't be able to do and and in the business world, I wouldn't be doing this business had I not done a PhD in this and that came from academia. So what are your current research interests?

My current research interests are artificial intelligence. specifically applied to pediatric imaging but also how we implement that in a clinical setting. So at the moment, the main project I'm involved in is how we can develop an AI algorithm to detect fractures for children. There are commercial products that can do that for adults. But very few AI tools are actually regulated and approved for use in children and they are needed to be used as off label, kind of like how drugs

are used off label for children. But I really do feel that unless we have the evidence to show that they're effective, they're accurate,

they can be used in the correct way, that we'll never know if we should be using it or not. And if we don't have enough academics who are able to do that study and to prove that, then that means children get left behind and when we have new technology, I don't think that's fair that a subset of our population don't benefit from it or don't have the knowledge to be no to know if that's beneficial to them or not.

So right now I'm working with we have a multicentric data set from various trauma centres across the UK. And we're using this data set of pediatric limb x rays. to both develop our own AI algorithm that can detect fractures in children, but also using that as an external validation set by which to test commercial offerings from AI companies that have been regulated for adults to see whether or not we can use the same tools but for children.

In um interviews we've done in the past, um, a lot of our guests have spoken about the importance of like good quality data sets and how that it's quite difficult to get that in adults. So I'd imagine it's maybe even harder in children, um, just to get sort of good standardized imaging. And then also there's probably more legal and ethical frameworks you have to go through in order to get access to children's Beta, is that a fair assumption?

It is hard to get good quality data sets and you never really know how representative it is of a wider population because you can only access what you can access and you try your best to make it as good as possible. Um but it would be impossible to get the full population of all ages of all fracture types. Mm, exactly. Yeah. So much variation.

Exactly. And in children there's a h greater variation I think because different ages bones look different and there's different growth plates and ossification centres that occur at different ages. which adds to the confusion and also different fracture types, like we have Salter Harris fractures and also cornometaphyseal fractures, torus fractures, green stick fractures. So the type of fractures that happen in children is different to adults as well.

Um in terms of whether we have enough data, I d I don't know. We try to get as much as we can and I've got a few re radiology research fellows who are going through the data set for me and annotating the images and checking that we've got a wide variety of Psalter Harris. type fractures and torus type fractures. So we're not just testing these adult algorithms on, you know, the same kinds of adult fractures but in children, because that wouldn't be fair, you know. So um

So there's that consideration. And the other big consideration we have, which is really tough to account for, is um there's a huge number of x-rays that are done for skeletal surveys in the under two population for possible abuse. And um a lot of these children um may or may not have fractures, but a lot of algorithms are not tested or trained on very, very young children because the number of x rays they can get hold of in the very, very young children are limited.

Apart from if they happen to be for suspected abuse. And we really want to try to develop something that can at least highlight potential areas for review on these skeletal surveys as well in the future so that we can safeguard children who are um at risk.

So it's actually a coincidence, and before... we uh decided to approach you for this interview, which was mainly gonna be about your outreach work, which we're gonna talk talk about shortly. Um I saw you wrote an article in the R CR Autumn newsletter um which was about artificial intelligence training for radiology trainees. So like how did that article come about and what are the take home points

So that article came about for two reasons. Number one, the newsletter was focused on training issues. And then number two, um I am part of the RCR AI Committee and also the lead of a subcommittee within that bigger committee trying to develop a course on AI for radiologists.

not specifically trainees, but for all radiologists. I feel that it's a new topic for even consultants who are very experienced. So it was more to introduce the topic to everyone. Um So that article was trying to highlight the points that yes, we do need more awareness of AI, but not just from a tech techy kind of computer science y point of view, but also an awareness of how that sits within our clinical practice.

whether or not we should or shouldn't use AI, whether AI is even the answer to our clinical issues, and then how to evaluate the right AI to use within our own particular subspecialty field. So the course that we've developed through the RCI Learning Platform tries to introduce the topic of what is AI, what's the jargon that's being used so that when you read an AI paper or some text, you understand what's going on.

And then also to get you thinking about how an AI is trained and developed and what are the limitations for what AI can do. So when it fails, people don't just think, oh, AI's rubbish. Actually think, actually it failed because this reason or that reason or the data set or the way it was trained and not to just say it's good or it's bad because that's very binary and

you know, but to think a bit more about it, about it's good for this particular use or in this particular setting or for this particular problem and not good because of this particular reason rather than just going, I like it or I don't like it, you know, which is kind of how I felt it was going because a lot of people were thinking it's a hype, it's not gonna be here or

I've tried it once and it didn't work and therefore I'm never going to give it a second chance again. So um that was really the the thinking behind it and trying to tackle that sort of mindset because I think now we realise AI is here to stay in some shape or form. We don't know exactly how.

or what impact it will have, but it will be around in some way. And um trying to moderate how we use it in the safest way. And I think everyone has a role in that. So we all need to have a bit of an understanding. So that module is on the Learning Hub now. Yeah, so if you um go through the RCR um website, there is a course now that's a blended course.

online learning modules combined with an in person or online workshop, depending on what you can attend, to try to um first impart the knowledge in a sort of didactic, self paced way, but also to try to harness the um

benefits of networking and talking through things and trying to clarify any misunderstandings that might have come up from your reading or understanding of the literature. And I also think because it's quite I say quite new, I know people will say, Oh, it's been around for ages but I think The trickling down of AI into clinical practice is Quite new.

Um you do need um a support network of people who have tried to implement things or who have spoken to industry partners and gotten certain bits of information to share that information more widely. And I think that's where having the workshop and networking with like minded individuals really help.

So the last thing that we wanted to talk about is some of the radiology outreach work that you've been doing abroad. We know that you've recently returned from a trip to Nepal. What were you doing over there? Thank you so much for raising that. Yes, so I was in Nepal in August for a couple of weeks working on behalf of the Rad Aid charity to try to teach pediatric radiology to the radiologists there.

I had two um outcomes there. One was to work with the main teaching hospital and the trainees, radiology trainees in that hospital, to teach them about interesting um radiology cases that we see in the UK and interesting and challenging pediatric radiology pathology that we encounter and help them learn a little bit more about that. but also a second part was working in a very remote clinic in um A small little village outside the pool called Benepa.

which was an orthopedic centre that had no radiologists whatsoever and one ultrasound machine that they had been donated but didn't really know how to use. So I spent half my time in Banapa teaching the orthopedic surgeons there, as well as the nurses, the physiotherapists and the um house officers how to do ultrasound for um dysplastic hips in babies.

Mae'n ymwneud â nhw'n ymwneud â nhw'n ei wneud â nhw'n ei wneud â nhw'n ei wneud â nhw also doing very, very basic chest and abdo ultrasound scans for any sort of post op complications so that they would know if a patient was recovering from a hip or a knee or an ankle operation, whether or not it was important for them to be sent over to the main teaching hospital or whether it was just nonspecific abdo pain.

So we we did quite a lot of that teaching, hands on practical teaching in the outreach centre, but then more sort of didactic, interesting cases teaching in the main teaching hospital. I mean it's terrifying to think of the level of radiology services that that people have to make deal with in these countries. Um don't know how they survive. I imagine the clinicians are much better at making decisions without radiologists unlike

Um yeah, it is scary, isn't it? Because we're so reliant on imaging in our modern day healthcare systems. I think they I don't know if I would say they were better at making clinical decisions. Um they they are very good that because they've had to cope without the expertise of imaging, but

they do really see the benefits of imaging and how that can enhance their clinical decision making. So I wouldn't say that they feel like their clinical decision making is superior to ours and they don't need imaging. They are very, very welcoming of anybody who can teach them imaging.

And um the difference between over there and here is I guess over here we have more doctors and more people who are subspecialised over there in the middle of a clinic in the jungle where you don't have any radiologists. you are the radiologist, even though your specialty is orthopedic surgery. And so there's a huge willingness to learn because they know in the middle of the night they are it. They have to know about the imaging. They've got to be a jack of all traits, you know, so

So, um, there's definitely a a lot of respect for radiologists and what teaching you can offer them and they don't feel you're treading on their toes. They don't feel like, Oh, who are you to teach us? You know, they're very, very happy to learn from you. I went to Nepal for my uh medical school elective in twenty ten and we were based at County Children's Hospital, which is it was done in partnership with Alderhay and um I remember

The um the surgeon there on his list, there was just so much variety. You'd go from doing some sort of congenital malformation to an amputation for someone with a severe burn. um appendisectomy, it was just it was everything. He was the surgeon. He wasn't an orthopedic surgeon or a GI surgeon. He was just the surgeon and he had to do whatever was put in front of him.

Absolutely. The funny thing is I went there with some cases of T B that I thought were interesting in inverted commas and they were completely bored with that. They were like, Yep. See about ten of them a day. They actually found a lot of like the boring thing, well, things that we find boring, like appendicitis, hypertrophic.

pylorix no like, Oh wow, that's really cool. You know, Crohn's disease. Wow that's really cool as well. So it's funny, isn't it? The things that they find interesting and we find interesting. Yeah. But they don't see anything in its early stage. It's always like end stage and everything looks the same. Like pneumopritoneum and uh ischemic bowel.

Oh yes, absolutely. We saw a lot of cases of very chronic osteomyelitis there as well, which is sad, but Um, I had previously done another outreach trip a few years back before my PhD in Laos in a children's centre in a small little um town called Luang Prabang. Um and over there that was a small children's hospital. It wasn't specifically orthopedic, it was a general children's hospital. Center that was set up and

the range of cases you saw there. I mean, there were a lot of children who came with right iliac fossa pain and in my sort of UK brain I was like, Oh, appendicitis, appendicitis and the surgeons there were teaching me and going, This is typhoid, this is typhoid, this is really clearly typhoid.

Uh yeah, so you do have to get used to a different type of like pathology mix. Which is important for AI by the way, bringing it tying it back to a different topic. If you were to develop an AI, you know, in the UK based on UK population diseases and then apply it to a third world, it may not generalize.

Yeah. There's no such thing as like a generalizable AI for every community. You're gonna need to develop it specifically for the different data sets and obviously some communities are gonna be better represented in that data than others, so You can already see how the um differences in like the Absolutely. It does raise a lot of ethical questions as well. Um

So, you know, what's generalisable, what's not, what's good enough, what's better than the status quo, even though it's not specific for you, but it's better than what you have now, you know. It it's hard to say, isn't it? Even a bias data set. If it still performs better than doctors currently, it might still be better. Exactly. Um

when we were doing some research for this, um, was your vlog that you recorded where you documented your travels and it's on YouTube and it's very cool and really professionally edited. And we've had our own unsuccessful trials with video editing recently. um in a podcast which is um never going to be seen by the world. So we know how d how difficult it is. Um so yeah, I mean, uh have you do you have experience of video editing before?

No I don't and it's very kind of you to be so complimentary about my video. editing. I um I I don't and it's very much homemade on iMovies on my laptop and not anything fancy or I haven't sent it to any third party at all for that.

Um, I uh I still have so much material that I still want to make another three episodes on that, so it's still a work in progress, but um I'm very happy to um have anyone give me any feedback on it or if they want me to create something or speak about a particular topic, I'm very happy to do that. Uh I would say that I have been on a couple of video editing courses through the university, mostly because Now um in academia, having video and visual abstracts for your research and

having your research being more sort of publicly promoted is becoming a bigger thing, you know? People want more engagement with academic um pursuits. And so grant funders and also universities are very keen that academics do promote their work in a more accessible way to a greater audience. So I have gone to a few local courses and obviously YouTube's a great um resource for learning how to do things yourself.

So there was uh there was there was one bit where you were you were filming yourself in the airport during a stopover. Was it weird recording yourself in public? Or I suppose it's quite a common thing nowadays. They probably thought you're recording a TikTok. Um

Yes, it is weird recording yourself in public and you do feel like a bit of a mad woman with people going, Oh my god, what are you doing? But actually sometimes when you video record yourself, people just think you're on a video chat or FaceTime someone, so it doesn't feel too bad. And I think most people are quite, you know, familiar with people doing their own sort of I dude. YouTube channels, vlogs, TikToks

What have you. I'm not doing anything embarrassing like dancing or singing. So I'm self conscious doing that. I think um the only thing I have to be a little bit careful about is trying not to get too much too many people in the background because I know some people don't want to be filmed and so I'm always very mindful of trying not to, you know, have too many people in the background and and

That sometimes works. Um sometimes I have to edit quite a lot or cut people out or try to blur the background a little bit. That's a big thing in Dubai with uh with doing your own videos. Um privacy is

is very important over here and it's not uncommon like to see people walking around with what looks like a fairly professional setup. And um if they're wanting to go and film in a bar or restaurant or something then often security will say to them, you know, you're not you're not Mate, you're all influencers out there, like just a bunch of influencers gallivanting around. Yeah, yeah, that's the thing. So I guess people don't really bat too much of an eyelid, but I

You do feel really silly the first few times you do it. I but then the more you do it the more you know, happy and comfortable you feel about it. Just like doing podcasts, I'm sure the first few times you heard your voice back, you were like, Oh no, I sound so silly you get used.

No yeah, you do get used to and hopefully get better over time. It's been five years, so I think we've got a little bit better. But yeah. Um so like, um in terms of the outreach stuff, what are some good charities for people to reach out to or and if they want to get involved. I would say RAD it is probably one of the biggest charities that um I've been involved with. They do quite a lot of outreach programs across the world and for all different sub specialties, not just pediatrics, by the way.

So even if you're a breast radiologist or an oncologist or cardiac immature, there is something you can do as an outreach. I know a lot of people will say, oh, but I work in cardiac MRI, let's say for example, it's very niche. How can I do any outreach? These developing worlds won't have access to the fancy equipment I use. That's not necessarily true, you know. Some places do have MRI machines and

capabilities that you would be surprised about. They just don't have anyone there to teach them what the right way is to do it or someone who'd set up protocols for them. So um I would say if you're thinking about it, do reach out to RADAid and ask them what you can do and what skills you have. Other um charities I know of is um WFPI, which is the World Federation of Pediatric Imaging.

They also are very welcoming of people. I'm sorry that is pediatric specific, but welcoming of people who want to put together educational YouTube videos or lectures. You don't actually have to go and travel to these places if you don't like travelling or if for whatever reason you can't travel. Um and similarly there's another charity called Worldwide Radiology.

that also, you know, had do outreach work with other developing countries. I think they are probably a bit more Africa specific, but yeah, there there are many, many ways you can help.

There's also ways I think with some teleradiology companies where you can report for third world countries as well. So maybe you're not someone who wants to travel or teach, but you're happy reporting some chest x rays for them or giving an opinion on certain cases, then I you know, I know some teleradiology companies do try to advocate for that, so that might be something to think. And then what about if people want to follow in your footsteps as an academic paediatric radiologist?

thinking about building your C V um any sort of um courses to go on or do Gosh offer uh any fellowships, for example? That's a great question. I guess if you're a specific pediatric person wanting to do academia, the first thing is to just get very involved with some local projects wherever you're training. Maybe there's a small project you can do.

Um and then if you are interested in taking it further, I highly recommend the Royal College of Radiologists Research Certificate. It's currently undergoing a revamp to try to include webinars to help people learn a bit more about research skills and there's also a radiology research day where you can get help from various academics, not specifically Pete, but you know, there's a lot of academia that's very generic, like how do you apply for ethics?

How do you um speak to information governance? How do you put together a research proposal? How do you apply for grants? So you start doing small little things like that and applying for seed pump pyroing grants through the R CR and at Gosh we do have research fellowships. Um They're not super advertised as research fellowships, but if you do a clinical fellowship

And you want a research component, there are ways in which myself or Professor Arthur's can try to um work some research into that clinical fellowship with you. And obviously I went away to Toronto to do another post-CCT fellowship and

A lot of overseas fellowships in America and in Canada do build in one academic day a week within the fellowship as research component for you passing that fellowship. So For example, I couldn't get signed off for the for my Pediatric Radiology Fellowship in Toronto, unless I had produced a manuscript, applied for ethics and carried out and completed a pediatric radiology research project as part of that year away.

So there there are some that are a bit more academic and do require you to do that. So if that's something you're interested in, then you can go for the hat too. That sounds very intense, but I'm sure it would like motivate you to get it done. Um, that's that's great. So, um yeah, so I think that's been a nice whistle stop tour through your many sort of endeavors. Um, hopefully it maybe sparked some interest in some of in some of our listeners to get involved in

sort of radiology outreach work or an or um academic um radiology or PEED. So um yeah, um very interesting interview. Thank Thank you so much for having me. It's been a pleasure to be here and um thank you so much for researching my career so um intently and coming up with some very, very good questions that I've had to stop and think about myself. It'd be weird in any other context apart from uh podcast. Yeah. Well, that's why we're the best.

Yeah, you'll sort see me pop up on your LinkedIn now. Um using your profile. I'll probably have colleagues going, so and so is like asking questions about you the other day. Great. So uh thanks for listening, guys. We'll be back next month with another episode. And in the meantime, you can catch all of our previous episodes on the usual podcast platforms and at anchor.fm forward slash radcast.

If you want more updates, then check out our social media channels, which is at Radcast Academy on Twitter and Instagram. Bye. Goodbye.