How can future technology help my hormones?

Could 3D printed pancreases be the cure for type 1 diabetes? Will AI take over in the IVF clinic? How and why have researchers created a remote controlled adrenal gland? Simon Hanassab, Dr Vicky Salem and Professor Polina Anikeeva are separating science fact from science fiction and seeing how future technology could solve today’s hormone problems.

Play podcast

Hormones: The Inside Story

Episode 18 – How can future technology help my hormones?

Sally: Hello and welcome to Hormones: The Inside Story, the podcast from the Society for Endocrinology. I'm Dr Sally Le Page, an evolutionary biologist and science presenter, and I'll be chatting with a whole bunch of hormones scientists - or endocrinologists - to bring you surprising stories and cutting edge research from the world inside your body.

Sally: It's the final episode of this series, but instead of reminiscing in the past, we're looking firmly into the future, and asking how futuristic technologies might be able to help my hormones.

Sally: How can AI help doctors make better decisions during IVF treatment?

Simon: If we personalise the dosage of drugs maybe we see that actually, I can give a lot less and we can save maybe 10-20% of costs.

Sally: Could 3D printed stem cells finally offer a cure for type 1 diabetes?

Vicky: It's something that when I left medical school 20 or so years ago felt like science fiction, but is now really tantalisingly close.

Sally: And why have scientists created a remote control adrenal gland?

Polina: How can we kind of hijack this whole circuit because if we could do that then we will be able to control them releasing their hormones.

Sally: Modern humans - and our hormones - have been around for about 200,000 years, more or less. And for most of that time the field of medicine was limited, to say the least. Doctors would test for diabetes by tasting their patients' urine; people would rely on herbal remedies to treat their hormonal issues to varying degrees of success; and many people with untreatable conditions sadly just didn't make it.

Sally: Today we’re able to diagnose, treat and even prevent diseases and health conditions that our ancient ancestors would have thought were incurable with tests, therapies and technologies they probably would have viewed as magic. In this final episode we're taking a leap into the future, to see which technologies that seem almost magical to us today may become commonplace in healthcare, starting with artificial intelligence.

Sally: We are living through an AI revolution. From voice assistants to ChatGPT, artificial intelligence technology has progressed at an astonishing rate in recent years, and now it's transforming healthcare. AI tools can already spot anomalies on X-ray scans to help radiologists diagnose cancer. Others enable researchers to design more efficient clinical trials, while machine learning can sift through enormous quantities of data in the search for new drugs. Scientists have even looked at ways that AI could help us make babies… and before you get any ideas, I mean in the IVF clinic.

Sally: But before we delve any further, let’s make sure we've got our words straight. When we say “AI” or “machine learning”, what do we actually mean?

Simon: Machine learning is really just looking at loads of data and looking at the patterns that we can spot in that data and then draw inference. So draw conclusions or draw new insights about how that data affects something that we're interested in.

Sally: That's Simon Hanassab, a PhD student in the AI for Healthcare Centre at Imperial College London.

Simon: And I purposely said the word machine learning and not AI because that's really a subset of that concept of AI, and AI encompasses a lot of things. I think in its simplest form it's just looking at data in the past, things that have happened, and then saying 'is there some sort of pattern in that data' and then apply that for different purposes.

Sally: IVF, or in vitro fertilisation, is in and of itself one of the most remarkable scientific advances of the last century. 'Test tube babies' used to be the topic of science fiction, but now over 8 million children have been born thanks to this technology. But that's not to say it's easy; it's still a complicated, lengthy and often expensive process. And with the chances of having a successful pregnancy hovering around 30% or less, it can be physically and emotionally demanding.

Sally: This is why researchers like Simon are looking to new technological advances such as machine learning to improve the IVF process.

Simon: This field is only around 40 years old, pretty new if you think about it in the context of medicine and you've got all these different steps at different time periods and it's quite a long time period. We're basically here to try and understand, what things work, and what things don't work, and how can we personalise that.

Sally: The IVF process involves a lot of decision-making. Which eggs should clinicians choose to fertilise for the highest chance of success? Which embryo looks like it has the greatest chance of developing into a healthy baby? How many embryos should they implant? And those are just some of the decisions in the early lab stages.

Sally: These choices are currently made by embryologists; human beings making subjective decisions to the best of their ability. Scientists are now developing AI-powered tools to help support this difficult decision-making process, and these are already showing great promise. We already use this technology to look at photos of the embryos to see which ones look healthiest for implantation.

Sally: Simon wants to expand the use of AI in the clinic even further, by looking at ways to personalise treatments for the patient.

Simon: Obviously in the clinic, there's also very complicated decision points going on. Things around how much of a certain hormone we should give in terms of dosage, at what time, and for what duration and what type. So there's several different aspects that we can think about, how we can personalise those decision points.

Sally: It's estimated that over 1.3 million IVF cycles have been performed in the UK alone. That's a lot of data that can be used to train machine learning tools. Simon is hoping to use this wealth of information to help optimise specific hormone injections that occur early on in IVF.

Simon: That whole process involves a lot of different hormones that are being administered. So something called FSH, follicle stimulating hormone. There's other administrations that are going on as well. And it's a sort of balancing act, right? So you want to not overdo it. You want to make sure it's personalised so that person in a specific way, depending on their condition or depending on the reason that they are infertile at that point of time.

Simon: And bringing that all together is often very tricky. And there's a lot of things to look at as a clinician at the same time. So what we're trying to capture is what are the things they should be focusing on? What are the guidelines that maybe we can introduce, that are based on now thousands and thousands of data points that we have access to? And how can we make things better?

Sally: If you're a regular listener of this show, you'll know by now that the world of hormones is complex and variable. Hormone levels vary between individuals for a whole host of different reasons - it's far from a one size fits all.

Simon: No two people are the same, right?

Simon: And we've got different ages. We've got different weights. We've got different reproductive hormone baseline levels. We've got different ethnic communities.

Simon: We've got different diagnoses of infertility and all these kind of aspects come together to decide what dose we should use.

Sally: There's a lot at stake when choosing the correct hormone dosage. Get it wrong, and not only might it impact the overall success of the procedure, it can also have huge cost implications. And if there is one thing I know about IVF treatment, it's that it can end up being very expensive.

Simon: In the UK, we offer one or two cycles free, uh, depending on certain criteria. But, it's not obviously free in actuality, it's thousands of pounds, around 10 to 15,000 pounds of cost per IVF cycle.

Simon: And what's interesting is that the drugs can be part of around 35% of that cost. It can be up to around 35%. And that's what we're seeing in the US and the UK.

Sally: If AI can help prevent us from using more drugs than absolutely necessary, it could reduce the cost of each IVF cycle, reducing the financial burden for patients already going through a stressful time, and making IVF more accessible to people who aren't as well off.

Sally: But it's not all about the money. If you don’t give people the right dose of hormones, you can end up causing some pretty major health complications:

Simon: Often, for example, we have women that have a condition called polycystic ovary syndrome, PCOS. And that usually means that you'll end up being what's called a hyper-responder or high responder.

Simon: And therefore, if you give too much FSH and too much hormone, you'd overdo it, let's say. And that may cause complications down the line.

Sally: So how can a clinician decide how to personalise these hormone injections to best fit each individual patient? Well, this is where machine learning comes into play.

Simon: As a human, we may look at two or three different factors, let's say age and weight, for example, and see how they relate to a dosing of a drug, or they relate to any other sort of healthcare aspect we care about.

Simon: What machine learning can do is instead of looking at two factors, it can look at hundreds of factors and look at 100 different things and how they relate to an outcome that we care about and see if maybe, in fact, you are right as a human. Only two of them matter, but maybe actually 12 of them matter, and they matter in different ways. So using machine learning, we can do this and we can do this at a much larger scale so we can look at many, many previous data points and try to come up with smarter, cleaner and more personalised answers to the questions we care about.

Sally: The plan is for this AI tool to work alongside human clinicians. You can think of it a bit like Clippy; "I see you're trying to calculate the dose of the hormone FSH, would you like some help with that?"

Sally: But might we one day see the entire role of the clinician being automated? Is there a future where patients only ever speak to a computer and get treatment from a machine?

Simon: I don't think so. I think we're very, very far from that being a reality. I really imagine that the best we'll see in our lifetime is something around decision support. So aids clinicians in that when they're making decisions along the way, they're very much backed up by data, and they're very much data driven, and they can make those decisions in a more informed manner, and they have more information available to them to make those decisions.

Sally: If we can use machine learning to personalise a patient's treatment plan, the hope is that we can make IVF cheaper, safer and more effective for everyone.

Simon: We want to maximise the success and make things more efficient and more efficacious, but then we also want to make things safer. And they both have a value and they both have utility to people.

Simon: So I think using machine learning we can reduce costs in a way because you'd see that if we personalise, let's say, the dosage of drugs we need to give: maybe we see that actually, I don't need to give that much. I can give a lot less and we can save maybe 10-20% of costs, but obviously we have all these intermediate success points along the way and all those are kind of intermediate success goals that we can think about.

Simon: I think the most obvious success point would be having a healthy baby at the end, right, we're trying to maximise the likelihood that we actually end up having a live birth.

Sally: Earlier in this series, we talked about type 2 diabetes and the genetic, ethnic and environmental factors that may stop your body from responding to the insulin you produce. Well today we're going to switch our focus to type 1 diabetes, a disease where you don't make enough of the hormone in the first place.

Sally: Before the discovery of insulin, being diagnosed with type 1 diabetes was a death sentence. Then a hundred years ago, doctors found that injecting an extract made from a dog's ground up pancreas could treat the disease, and ever since, insulin injections have been the stalwart of diabetes treatment, although thankfully technology has moved beyond getting insulin from a pup's pancreas.

Sally: But even though researchers have been working on improving treatments for over a century, living with type 1 diabetes still requires a lot of management, as Imperial diabetes researcher and clinician Dr Vicky Salem knows all too well from her patients.

Vicky: If you're diagnosed with type one diabetes, it means that you have to take insulin for the rest of your life. And usually that means four, five, often more injections of insulin a day. And you have to know how to dose that insulin by responding to what your blood sugar is doing. So you also have to be able to measure your blood sugar. It's a really complicated thing to do: day in, day out for the rest of your life, worrying about what your metabolism is doing, what your blood sugar is doing and dosing your insulin accordingly is a huge mental burden for all of our patients.

Sally: Since the 1960s, people with diabetes have been monitoring their blood sugar levels multiple times a day by pricking their finger with a needle and squeezing out a drop of blood onto a testing device. It's about as fun as it sounds.

Vicky: It's painful and it does, over time, become very uncomfortable for a lot of people. It's just not nice.

Sally: Then in the early 2000s, a new technology arrived - the continuous glucose monitor, or CGM.

Vicky: They're about a 50p size sticker that sticks on your arm. And underneath there, there's a tiny little needle that sits under the skin and it's actually sampling not your blood sugar but the sugar in the fluid underneath the skin, which is a pretty close match to your blood sugar, every five minutes.

Sally: Today, continuous glucose monitors can send this data directly to your phone so you can get information about your blood sugar levels almost in real time, meaning you can inject yourself with insulin as and when you need it.

Sally: And for people who aren't keen on injecting themselves on a daily basis, there's a technological solution for that too - in the form of wearable Bluetooth insulin pumps that seamlessly secrete insulin directly into the abdomen in response to the blood sugar data.

Sally: Although these technologies have been transformative for people living with type 1 diabetes, making it much easier for them to manage their blood sugar and keep it in a healthy range for longer, they aren't a permanent solution.

Vicky: Technologies have revolutionised people's experience of self-management of type one diabetes, but it doesn't represent a cure in the minds of my patients and people living with type one diabetes. And when they think of a cure, most commonly that is defined as some sort of transplant of cells back into their bodies to replace the beta cells that they lost, such that they never again need to have something wearable on their skin that requires them to know what their blood sugar is doing and respond by dosing insulin accordingly.

Vicky: So really the holy grail for type one diabetes is what we call a cell-based cure. And we think that that is likely to be in the form of a stem cell based treatment.

Sally: Before we go any further, it's time for a quick refresher on where insulin comes from, starting with that all-important source, the pancreas. The pancreas is a leaf-shaped organ about as long as your hand that lies deep in your body behind your stomach.

Sally: Not all of the pancreas can produce insulin, though. That job is reserved for highly specialised cells known as beta cells, which are housed in little structures called islets.

Vicky: So the pancreas has interspersed throughout it millions of these tiny little micro-organs, which are about 0.1 millimetres in size and highly optimised to produce insulin in response to changes in blood sugar levels.

Sally: And it's these beta cells that are missing or destroyed in people with type 1 diabetes, leaving them unable to make their own supply of insulin. So, if we want to find a permanent solution to type 1 diabetes, why not just get donated beta cell islets and transplant them into the patient's pancreas? Unfortunately, it's not that simple.

Vicky: The first issue is that the supply of eyelets that come from somebody who has died and donated their organs - the supply of those eyelets is nowhere near enough the number of patients that require them. So there's a fundamental issue with regard to supply and demand.

Sally: And there's another problem. As with all organ transplants, recipients need to take immunosuppressing drugs so their immune system doesn't recognise the transplant as foreign and reject it.

Vicky: And immunosuppression is not something to be taken lightly. It results in lifelong dependence on very, very strong drugs that suppress the immune system and themselves are toxic and can cause lots of side effects, including an increased risk of other cancers and infections.

Sally: With over 400,000 people currently living in the UK with type 1 diabetes, including nearly 30,000 kids, organ transplantation just doesn't add up. But there's another way that we can get hold of beta cells without relying on organ donors, and that's by using stem cells.

Vicky: In essence, a stem cell is a cell that has the capability to turn into any cell in the body. Scientists have now cracked the cocktail of factors and additives and chemicals that need to be added to a stem cell over the course of 30 to 40 days to turn that stem cell into an insulin producing beta cell.

Sally: You might have heard about embryonic stem cells, which are taken from early stage human embryos and grown in the lab, but that's not the only source available. You can also find stem cells in adults and reprogramme them back into beta cells in the lab.

Sally: And if you can take adult stem cells from a person with type 1 diabetes, turn them into beta cells, and then put them back into their own pancreas, then the body's not going to reject the transplant, avoiding the need to take immunosuppressive drugs. But there's still a problem...

Vicky: That patient is still at risk of losing those beta cells through autoimmune destruction, i.e. the same process that caused their type one diabetes in the first place.

Vicky: Their immune system was triggered to destroy their islets. So why wouldn't that same process happen again in the transplanted islets?

Vicky: So, how do we get these cells into a patient in such a way that they don't re-destroy those cells.

Sally: If there's one thing we know about scientists by now, it's that they love solving tricky problems. So, what's the solution?

Vicky: The answer to that is going to be twofold. I think, just as we are learning to genetically edit our stem cells to make them even better at secreting insulin, the latest generation of these stem cells will also be genetically engineered to evade the immune system; to make them invisible, potentially, to the immune system.

Vicky: And then the other way to do that is to work with biomaterials, and this is really my area of interest, but to work with materials and hydrogels that can actually also act as a physical coating and a barrier from immune destruction. And that can then form a printed novel organ-like structure that can be transplanted directly into the patient, either under the skin or inside the tummy.

Sally: Vicky and her team are working on printing pancreases in the lab, so they can be transplanted into patients. And, as you might have guessed, this isn't your typical office printer.

Vicky: So in our bioprinters we have different nozzles and each of those different nozzles contain different materials and some of them contain gel-like substances that have got suspensions of cells in them. More and more we're also moving into 3D printing whereby we are building up a matrix of cells of different types. So for example, in our case, we're interested in building up a matrix of blood vessels within our stem cell derived islets, and all of that is mixed in with hydrogels to provide the right degree of structure and function to help those cells do what they need to do.

Sally: For now, 3D printed pancreas transplants and other stem cell technologies for curing type 1 diabetes are still some way off, although they're fast coming closer to science fact than science fiction.

Vicky: The dream is to find an inexhaustible supply of beta cells to use as transplants, but in a system that protects those cells from immune rejection, whilst at the same time keeps them healthy, long lasting, and strong.

Sally: So if we can get this to work, what does the future of type 1 diabetes treatment look like?

Vicky: What I want to see is a world whereby, when that diagnosis is made, that patient can come in to a hospital and have a procedure that is relatively simple, maybe a small, minor operation, that means that when they go home, that's it, it's done. They've got their replacement beta cells and they last forever in a healthy, safe way. And that patient no longer has any of the burden of managing their diabetes, but equally doesn't carry any of the risks associated with imperfect blood sugar control. So that's the dream and we share that with all of our patients who have spent so much of their lives worrying about a condition, which I think, is, actually really often misunderstood.

Vicky: That patient that has had stem cells injected into them and is now insulin free, has for the first time in decades, not needed to ever measure their own blood sugars or take insulin injections. I mean, you can imagine the sense of freedom that goes with that is immense.

Sally: Now, if you thought putting a 3D printed, lab-grown organ into someone sounded futuristic, how about switching an organ on or off with a remote control? Well, that possibility may be closer than you think - at least, if you're a rat...

Sally: As we heard in our pets episode, although we might think of stress hormones as 'bad', they're actually essential for our normal physiological responses to the trials and tribulations of life. The most famous of these stressy substances are adrenaline - the fast-acting 'fight or flight' hormone - and its close cousin cortisol, which keeps your body amped up and alert in the longer term. Both of these are produced by the adrenal glands, which sit just above your kidneys.

Sally: But while these hormones are vital for our health and survival, in the wrong amounts, they can cause serious health problems. So naturally, researchers are interested in understanding how these stress hormones and responses are regulated, and how we might be able to manipulate them. But unfortunately, studying the impact of stress on the body isn't a particularly precise endeavour.

Polina: Typically you can - of course, in humans - you can show somebody scary movies, but that's not necessarily a very specific manipulation for an organ.

Sally: That's Polina Anikeeva, a Professor of Material Science & Engineering and of Brain & Cognitive Sciences at MIT. And you heard right, one of the common ways of getting participants suitably worked up for your stress hormone studies is by making them watch a horror film.

Polina: Alternatively, of course, you can give people drugs, but often you can get some side effects that you actually did not anticipate. And also you're doing it all at once through this direct injection into blood, rather than having cells naturally produce it and release it at their own pace.

Polina: So, the goal for us was to develop technology that would allow us to manipulate the natural function of cells in the adrenal gland - and only the part of it that we are interested in.

Sally: Let's take a step back for a minute. If we want to control how the adrenal gland releases stress hormones, first we should understand how it normally works. So settle in for some adrenal gland 101.

Sally: When you watch a scary movie - or something actually scary happens - your body needs to prepare to react to the threat. So your brain sends a signal to the adrenal glands to release the stress hormones adrenaline and cortisol that we all know so well by now. But how does that work, exactly?

Sally: When the signal from the brain reaches the cells in the adrenal glands that produce stress hormones, it opens up microscopic channels that allow lots of calcium to go into the cells, triggering the release of adrenaline and cortisol into the bloodstream.

Sally: It's a bit like a smoke detector that sets off a sprinkler system: detect a smoke signal, pump in a bunch of water, deal with the fire. Or in this case, detect a brain signal, pump in a bunch of calcium, and deal with the stress by releasing hormones. But if you're trying to study a system like this, you probably want to find a way to trigger the smoke detectors without having to burn down your entire house...

Polina: So what we thought is: how can we hijack this whole circuit and bypass those natural commands into the adrenal gland, and instead create a method that will allow us to regulate this calcium flux inside those adrenal cells? Because we knew that if we could do that then we will be able to control them releasing their hormones.

Sally: Control the calcium channels and you control the release of hormones. Sounds simple enough in principle, but how do you put it into practice? The answer comes from somewhere you might not have expected: chilli peppers!

Polina: We found that cells in the adrenal gland actually have a really interesting protein in them, and it's called capsaicin receptor.

Polina: It is the receptor that is present in the cells in your tongue, and when you eat chilli peppers, you get the sensation of burning pain - and interestingly, it also senses heat.

Sally: Capsaicin is the chemical that gives chilli peppers their spicy kick when it activates the capsaicin receptors in your mouth. But it's not just the chemicals in chillies that triggers these capsaicin receptors. They can also be set off by an increase in temperature, which is why your tongue feels like it's on fire even when you eat a pepper at room temperature.

Sally: Crucially, capsaicin receptors aren't just found in your mouth. They are found in various places all over your body, including, you guessed it...

Polina: So we found that this capsaicin receptor is also present in the adrenal gland. And it so happens that it is a calcium channel.

Sally: In other words, if you heat up these capsaicin receptors, they will open up and let calcium through. So in principle, if you heat up the capsaicin receptors on the adrenal gland, calcium will rush in and stress hormones would get released on cue.

Sally: To come back to our smoke alarm analogy, we've now got both smoke detectors and heat detectors that can set off our calcium sprinklers. But we're still left with the same problem: how do you set off the heat detectors without burning down the whole house?

Polina: So that was our hypothesis. So now the question is, how do you actually do it? How can we deliver a little bit of heat? How can we make parts of the adrenal gland warm?

Sally: The answer, as Polina and her team discovered, was to use magnets. Teeny tiny magnets.

Polina: Our group has engineered small magnetic nanoparticles. Those are approximately one five thousandth of your hair in diameter. So because they're so small, you can dissolve them in a liquid. So it looks like a little droplet of coffee except it's full of those tiny particles.

Polina: We engineered these particles so they have a little magnetic moment.

Sally: A magnetic moment? No, that's not what happens when you meet an attractive physicist!

Sally: Instead - remember those red and blue bar magnets from your physics lessons at school? Each of Polina's magnetic nanoparticles is like a tiny version of those. If you put them near a bigger magnet, they'll want to line up so they're all pointing the same way. And if you switch the direction of that big magnet around, all the little nanoparticles will switch around too.

Polina: So you can think of them as like a little arrow that wants to point with the direction of the magnetic field. So this red blue cartoon signifies the arrow from north to south for our magnetic particles. In our cases, they're super tiny, and that is the thing that actually switches. The particle itself stays in place, but its orientation of that blue to red, or north to south, is switching.

Sally: Every time Polina changes the direction of the magnetic field, the little magnetic particles have to reorient themselves, and that creates heat. And the faster they switch, the hotter they get.

Polina: If we use a magnetic field and we quickly switch it up and down, up and down, for example, then, at every cycle of that switching, we will be making a little bit of heat from every particle.

Polina: And now, if you have a droplet of those particles, and you're switching it very, very quickly, let's say 150,000 times per second, then you will get enough heat to excite those chilli pepper receptors.

Polina: Now we've got our magnetic nanoparticles, and we can make them heat up on cue by quickly changing the magnetic field around them. The only thing that's left to do is to get them to the right place to heat up the capsaicin receptors in the rats' adrenal glands, turn on the calcium sprinklers and flood the body with stress hormones.

Polina: We inject the tiny droplet into a location of an adrenal gland. So it's just like a simple injection, we then close everything up and now we can apply a magnetic field to an entire subject, in our case a rat, and in response it's going to make parts of the adrenal gland warm.

Polina: That heat will excite the capsaicin receptors, our chilli pepper heat sensitive channels. Those channels will open up and pump a whole bunch of calcium into the cells.

Polina: Cells in response to that calcium will release their hormones and we get a response that is the same as the natural response that you will see when the command comes from the brain, for example.

Sally: It's really clever! But how hot are we talking here? Are we cooking the rats?

Polina: No, we're not cooking the rats. We are raising the temperature from 37, which is our normal body temperature, to about 41 or 42, so we go approximately 4 to 5 degrees Celsius. And we get to that temperature and immediately turn off the magnetic field so that we don't continue raising the temperature and everything cools back down to a natural temperature and then we can do cycles of this raising of the temperature.

Polina: So, if we decided to hold the temperature at 42 degrees for a long time, that of course would be not very good because that would be commensurate with fever, but we only get there for about one second.

Sally: By controlling the magnetic field surrounding the rats, Polina and her team can control the exact parts of the adrenal gland where they injected the magnetic nanoparticles, and trigger the release of hormones down to the millisecond.

Sally: This level of control makes it so much easier to study the intricate details of the stress response under different circumstances, such as how turning up the stress affects the rats' memory.

Polina: They were trying to remember an event that they have been training to remember. And it turns out that they end up doing a little bit better.

Polina: So a rat that is slightly stressed ends up remembering an event that they were trying to learn a little bit better. So that's the bigger effect that we have seen.

Sally: That's useful if you're a rat trying to remember events, but what about humans?

Polina: There are a lot of people struggling with things like post traumatic stress. And there is a lot of research that shows that it is inappropriate production of stress hormones that often results in people developing post traumatic stress.

Polina: And there's some literature that shows that raising people's levels of stress hormones right after the traumatic events can actually help them to not develop post traumatic stress or potentially even recover from post traumatic stress faster and better. So, that was the motivation for our research.

Sally: We’ve already come a long way with developing technologies to help us understand and control our hormones. And every new technological advancement opens up even more ways we can improve our health. Soon it won't just be using AI and machine learning as a medical tool in IVF labs, but it'll be in our homes too.

Simon: We're already seeing it already. If you think about apps that track your menstrual cycle, it's like having those kind of apps. You would provide it information; it would suggest certain treatment options. You would know more about what treatment you want to undertake and what treatment makes most sense. Can we validate these systems in the meantime, in the next 50 years, and can we validate them in a way that helps everyone?

Sally: And the same nanomagnetic breakthroughs that allow us to understand stress and PTSD could also help us understand the brain and dementia.

Polina: Right now, what I'm excited about is using this technology to not only understand adrenal function, but also functions of other organs that are similarly difficult to control. We have applied this method to stimulate particular regions in the brain, in the context of deep brain stimulation, which is used, for example, to treat Parkinson's disease. And we have shown that we can basically do just as well as the implanted electrical wires used to treat Parkinson's disease.

Sally: So while AI clinicians, 3D printed pancreases and remote controlled hormones might seem super futuristic to us right now, perhaps in 100 years they’ll seem as normal in the future as automated insulin injections are to us today.

Vicky: We're slowly but surely entering a new era whereby the concept of a cell-based cure for type one diabetes is a realistic proposition. And honestly, it's something that when I left medical school 20 or so years ago felt like science fiction, but is now really tantalisingly close.

Sally: That's all for now. Thanks to our guests, Simon Hanassab, Dr Vicky Salem and Professor Polina Anikeeva.

Sally: And that's also all for this series; we really hope you've enjoyed it. If you have, do consider leaving a rating or review wherever you listen to your podcasts as this helps more people discover the show.

Sally: Hormones: The Inside Story is a podcast from the Society for Endocrinology. Explore more about the world of hormones at and follow them on Twitter @your_hormones.

Sally: The show is a First Create the Media production. It was researched, written and produced by me, Sally Le Page and Emma Werner. Our executive producer is Kat Arney, with additional research and support across the series from the rest of the First Create The Media team including Fergus Powell and Holly McHugh. Special thanks to Heather Lampard, Sarah Don-Bramah, Katherine Single, the Society for Endocrinology’s Public Engagement Committee, to all of our guest experts for this season, and of course, thank you for listening.