Science Sandwich podcast

What if we could observe the brain in real time? In this episode, Leonard Klar and Bahar Mostafiz discuss cutting-edge research on neurological diseases and brain monitoring technologies.

Their conversation bridges neuroscience and engineering, covering everything from multiple sclerosis and brain imaging to implantable sensors that could one day revolutionize how we diagnose and understand brain disorders.

Listen to the episode

Episode transcript

I: - podcast. This time doctoral researcher Leonard Klar and doctoral researcher Bahar Mostafiz talk about seeing and sensing the brain.

[music]

Leonard Klar: I am Leonard Klar and I am a Doctoral Researcher in the Clinical Research programme, specifically Clinical Neuroscience. It’s a pleasure to be here. 

Bahar Mostafiz: Hello! Hi! My name is Bahar, Bahar Mostafiz. I’m a Doctoral Researcher in the Material Engineering Department. And I’m a chemist by education, but I am now studying and researching implantable neurological sensors, that I will explain later, what that means. So, now that you have told us what the area of your research is, why don’t you tell us what it is that you’re actually researching?

Leonard Klar: Alright. To put it simply, the kind of research I do looks at (-) [0:01:08.8] and neuro-inflammation in multiple sclerosis. Multiple sclerosis, as some people might know, is an inflammatory disease that attacks the central nervous system.

Bahar Mostafiz: It’s MS, right?

Leonard Klar: Exactly! MS. And in MS, the immune system basically attacks the brain and the spinal cord. And what we find in MS is that there aren’t just the relapses that maybe most people know about, so at times people get worse, then they get better. But there’s also a progressive component that is a gradual worsening of symptoms and the neurological damage. And my research group basically tries to understand processes that are related to this progressive component.

Bahar Mostafiz: Can you tell me more about these processes?

Leonard Klar: Yes. So, the two processes are related to iron accumulation. So, we find that in lesions, so small areas in the brain where tissue is damaged, we see iron around some of these lesions, and we also see, in analysis of brains of dead MS patients, that there are some immune cells around these lesions that indicate that there is more neuro-inflammation. Especially-

Bahar Mostafiz: So, you study those?

Leonard Klar: Yes.

Bahar Mostafiz: Like le- lesions?

Leonard Klar: Lesions, exactly.

Bahar Mostafiz: Yeah, okay!

Leonard Klar: This is kind of the area that I focus on in my first paper, and we tried to see how the neuro-inflammation and the iron interact, and what they can tell us about the patient, basically. 

Bahar Mostafiz: Cool!

Leonard Klar: So, what do you do?

Bahar Mostafiz: Uh, I want to tell a story and then kind of expand upon what I do first. I am a history fan, so I really like history, I like philosophy. And years before, when I was reading about the history of how we came to know our bodies, there was this passage 2000 years ago that a very big philosopher (-) [0:03:01.5], they thought that the uterus, or the womb, is not attached inside a woman’s body and it kind of goes around and it detaches and it goes around your brain, it goes around your heart, and it can impact your personality, it can impact how you feel that day. And like the only way to keep the uterus in its place is to be pregnant, or like have babies. And at that point, the uterus is kind of fixed. And when it’s not, it goes (-) the brain and it makes people do weird stuff. All the time that I was reading that, I was just thinking “But why didn’t they look? Why didn’t they just see the human body every time the uterus, or the womb, what they called it, would be in the same place?”. And then it made me think that how can I make us look inside a human body and how it works, so we wouldn’t assume that like there is a uterus around my brain making me do weird stuff. And I came to making sensors, sensors that can be placed inside somebody’s body and what does that sensor tell us? So, bringing it to what I do in my PhD, I am really interested in understanding how is the nervous system of a person who has Alzheimer’s, Parkinson’s, work? What are the patterns they have in the neurotransmitters like dopamine and glutamine are getting released? And what does that translate into the abilities of the person, or like would they remember something? Can they move? How they can control their muscles. How they forget stuff. So, these patterns come from the patterns of those releases, and this is what I’m doing. 

Leonard Klar: So, do I understand correctly, you work on sensors that you could, in the end, implant in a human brain to read out neurochemical signals?

Bahar Mostafiz: Yes, yes, that’s the goal. It sounds like sci-fi. And it’s a long way to go. So, we are trying to make it better and better every day, making it more robust to a lot of different nuances. But, yeah, that’s the goal.

Leonard Klar: Okay. That is something that would be extremely useful in our research, because we use PET-imaging and MRI-imaging, which always just give us snapshots of what we have going on in the patient. And these sensors could really transform how we read our data of the patient.

Bahar Mostafiz: Yeah! I think our research kind of goes hand in hand. Because it’s based on the research of you guys that we kind of know what a person that doesn’t have like neurodegenerative diseases like Parkinson’s and Alzheimer’s and such looks like, and what is the chemistry of it. And then we can kind of calibrate our own data that we are getting, saying that this is the pattern in which- that the PET-scans have shown. But we are making it like ten steps ahead, that it’s not just one (year) [0:06:11.8], one image. It’s real-time, it goes on and on and on. And it gives a lot of data that then we can even transfer to you guys and say like “Okay. You had just one snapshot, now let’s give you, I don’t know, data of like weeks and weeks ahead.”

Leonard Klar: Yeah, it’s very interesting. I can only speak from the experience that I have with MS, but a lot of the groundwork is from the brain of diseased patients, so it is always just sort of the picture that you get at the end of the process. Then, the kind of research that looks at living patients, where we get MRI or PET-data, is, like I said, a snapshot. And then the closest thing that we get to what you guys are working on is that we take blood samples, samples from patients, and these blood samples then can be taken more frequently and give us some sort of indication where in the process of the disease they are.

Bahar Mostafiz: Yeah.

Leonard Klar: So, it’s very interesting that this is not something that is just going on in the field of MS, but that you are also interested in many other neurodegenerative diseases.

Bahar Mostafiz: Yeah! It introduces such a big amount of data to some organ that is the most complex in our body. And then- it’s kind of like going, again, in tandem, because if it’s not for the research that is done by the neuroscientists, we wouldn’t even know what is the right- not the right way, maybe the baseline. So, yeah, I think it will make treating and diagnosing much easier, let’s say. 

Leonard Klar: It does. And I’m always fascinated by how far technology has come, because to me, the simple technology of an MRI machine is so difficult to understand and it’s so hard to think how an engineer could be like “We could use a machine with a giant magnet to then read out a signal the way an MRI machine does”, and to think that you are now working on an individual sensor that could pick up the signals of one nerve cell is mind-boggling to me.

Bahar Mostafiz: Thank you so much! But I want to emphasise, this is years ahead. But you’re actually taking PET-images and MRI-images, so I kind of want to go back to your research and focus on that. What are the most interesting things you have seen or discovered? Or like even been exposed to? Because it’s a big science community, everybody is doing all they can, so what is it that was really mind-blowing to you?

Leonard Klar: I think one of the things that is extremely interesting to me is just the modality of PET-imaging. The fact that we can give people a radioactive substance and then it binds to specific places in their body, and that tells us something about what’s going on inside the body. Especially because Turku has a PET-scanner, and not that many research facilities have that privilege. So, I think- and a funny story was I was at a conference, my first conference, I wasn’t even enrolled in a doctoral programme, and the owner of a clinic asked me if I think he should get a PET-imaging machine, just because he wasn’t familiar with it. And I sort of was the expert on that topic for him. So, it just shows me what a privilege I have to work here with such high-tech equipment.

Bahar Mostafiz: Okay, yeah. But like it’s not that- you know, it’s not something you go get from the supermarket.

Leonard Klar: Yeah, exactly!

[laughter]

Bahar Mostafiz: There should be lots of planning. Because, well, one of the issues that we had- so, it kind of motivated us for this research, was that you see long lines of patients waiting to get their image in MRI or PET-scans, because they have to go the hospital, there has to be very trained personnel there, it costs a lot, and it’s- like the public health sector expenses. All of those things. And it’s not very accessible, let’s say. So, the idea of a personalised device that can be in every household if need be was something that was kind of in our minds, is it marketable or a desirable trait of our research? That you don’t even have to wait that long, you don’t have to rely on one snapshot, you don’t have to rely on someone who has like a million other patients waiting on the other side of the door to just look at your results. So, yeah, that’s what I was- like when you were talking about like getting a PET-scan.

Leonard Klar: Yeah, with your experience, would you say that- I know about these glucose sensors that you can basically just put on your arm and then it gives you more or less real-time information.

Bahar Mostafiz: Yeah.

Leonard Klar: Do you see, in the near future, anything like that happening with the sort of sensors you’re working on?

Bahar Mostafiz: Well, like to be honest, it’s not just glucose anymore. Like cortisol is making a very huge noise in this part of like research and the (-) [0:11:21.3] and stuff, so you can actually have personalised cortisol sensors, and they are directly indicative of the stress level of someone. So, it’s kind of in the neurological environment, let’s say. But the difference between glucose and cortisol are that, first of all, their concentration in the body is really high. So, we are talking like millimoles, and they are almost stable, so they don’t fade away. It’s more accessible to be monitoring glucose. But when you’re talking about neurotransmitters, because it’s kind of between neurons and it’s just like milliseconds, like less than a millisecond, a fraction of it, that they get released, they get absorbed, and they convey the message. Then the challenge becomes that like it’s just a little bit in a very small space that is very sensitive and then it just vanishes. So, the challenges that neuro-detection brings with itself is much more complicated than the materials that are in the blood just roaming around. So, yeah, that’s one thing that is making it a bit less realistic to be having it just one year from now, because you have to take all of these into account. 

Leonard Klar: Absolutely. One question I had when I prepared for this podcast was: I know you research the geometry of these sensors, can you tell us a bit about what exactly that has to do with the sensors?

Bahar Mostafiz: Yeah, sure! The thing is that like the material that should be used for the sensor is a very big factor, because they shouldn’t, first of all, be toxic. That’s like base level one. So, if it’s placed inside your body and like even in your brain, let’s say, it should be something that the cells can live next to and not die and not get toxified, whatever. And then, what physical properties that material should have should be taken into account. And also, how we can make sure that it remains functional in a long period of time inside the body, is another thing. We kind of know what type of material we are working with. We kind of know what the shape of the sensor should be like. But then, the issue of remaining inside the body is a big factor. And the option of just adding everything on top of it is not available to us. So, we have to change the geometry to see if the response becomes better, because adding more stuff just to make it more sensitive or durable is not an option anymore. 

Leonard Klar: Alright.

Bahar Mostafiz: Yeah. You don’t place much materials in the body, you are looking at visuals. So, can you let me know what the things are that you are most- like, you know, you look for in those visuals, that tells you something? Because we are looking at graphs. We are just looking at like electric waves and stuff, but you are actually looking at regions of the brain and it’s like very complicated, has a lot of details, so how do you make sure that you are actually looking at what you assume it- what is it that you’re looking for?

Leonard Klar: Yes. It’s a very interesting question, because the kind of analysis we do is sometimes a mix of visual and computer-guided. So, when it comes to the lesions that I talked about earlier, we use an algorithm to predict where they are, but then we always have a human check if they actually identified the correct region. But then when it comes to larger regions of the brain, there is actually a lot of software that lets you do that quite accurately. So, we do use a lot of quantitative analysis, it’s not all visual, but there is some subjective aspect to this visual identification of the lesions, at least. And this is actually something- makes us always wonder how much we can actually trust this data, because some of it is based on visual observations, even though we know there is also some quantitative part, which scientists love. They love the numbers. It is sometimes difficult to be 100% sure if you can trust all of the numbers that you get in the end.

Bahar Mostafiz: Mhm. Now that we are there, let’s talk about the challenges, because you were just talking about these parts that you are a bit unsure of. So, I guess then we can find a common challenge, because I have some reservations about- okay, I am doing my research in a lab, I’m getting some results, but what if I actually take it to the living situation. If we have cells, if we have like animals. And then, there are like a ton of things that are unaccounted for, still. The research is going there, but for example, there are a lot of different molecules in the body, there are a lot of, you know, species just running inside our veins. Then pinpointing the pattern that we are seeing is because of that neurotransmitter, that becomes a challenge. So, I guess for you, pinpointing what you are seeing, that is an indication of the MS or the relapsing, then it’s like still challenging, so you cannot say 100% sure “This is what we are seeing.”

Leonard Klar: Yes.

Bahar Mostafiz: It’s just probability.

Leonard Klar: Exactly. It’s very difficult to know exactly what your signal tells you, because these methods are not perfect, but we can put many different methods together to form a sort of understanding of what’s going on. So, there is for example this type of analysis that lets you know exactly what messages a cell sends out. Exactly what RNA-transcripts they do. And this is very good to get an understanding of what was happening in that cell right before you took this tissue out of the body. But then, as we talked about before, what our group tries to look at is what actually happens when the body is still alive. And even though we have both of these information and both of these sources, to put them together and get one method that gives you all is, so far, not possible. And I think that’s why it would be very interesting to combine our research. But then, as you said before, maybe we have to wait a few years.

Bahar Mostafiz: Yeah! And, you know, the thing is that the body kind of surprises us. You know, every time you make one challenge disappear by finding your way out of it, then you are met with something that is just a total surprise. This is not very recent, there has been a lot of studies on this, but one of the challenges of the sensors being put inside the body is that we are full of proteins. Proteins are in our brains, in our- basically every fluid that we have in our body is kind of loaded with these big molecules called proteins, and they are charged. And they are really heavy. And they’re really big, again, because that’s like a big part of the problem. And when you place a sensor inside the body they come and like because of their weight, because of their charge, they kind of block the surface altogether. And then you are not sure if you are seeing basically anything. Because you are not sure that this is the signal of what it is that you’re looking for, it’s just noise. So, yeah, the body is really surprising. And I guess- we talked about this before, I asked “What if somebody comes for an imaging, one of your patients come for imaging, and he’s drunk, or he just had a birthday, or he just lost someone?”. I think like the patterns in which- that the image shows is very different, so how do you make the judgement call that like this is, in comparison to the last time, it has made these changes. So, I think this unpredictability of the body is a bit of a challenge for both of us.

Leonard Klar: Yeah, I think that sums it up quite well. I think there are so many things going on at the same time within a living body, that it’s impossible to take everything into account. 

Bahar Mostafiz: Yeah, I agree. So, what now, then? We know what we are doing, we have our goal, we even know our challenges and limitations, but listening to this, do you think that we can find a common ground, like for research?

Leonard Klar: I think if the technology develops a bit more. Where I see a specific use case for these sensors within the brain in MS, would be if we would be able to detect the cellular signals called cytokines. Because they are very important for informing about the kind of inflammation that is going on, and to have that sort of information in real-time would be invaluable to a clinician.

Bahar Mostafiz: Mm, yeah. Well, we talked about you knowing the chemistry of the brain. Like you kind of know what materials are present inside the brain, but one of the issues was the resolution, so you- I think you said that it’s like one millimetre.

Leonard Klar: For MRI it’s one millimetre. For PET it’s even a bit more.

Bahar Mostafiz: Yeah, so like cells are like super small, so in comparison to that one millimetre. So, if we actually have a more accurate resolution, then we kind of know the chemistry of that region much better, and then we can anticipate what our sensors will pick up, and make sure that like okay, this noise and this peak that we are seeing is not what we are interested in, because we know what it is. So, I guess like advances in imaging resolution, let’s say, would help us a lot as well.

Leonard Klar: Yes, yes, it would help both of us. So, is there any one thing that you could tell the listeners to remember from your research?

Bahar Mostafiz: I think like the whole concept of my research is mostly- if there is nothing there, like if it’s not visible, it doesn’t mean that it’s not there. So, you just need a better tool to find it, and understand it. Don’t assume the lack of indication as the absence of the phenomenon. And make sure that you are still curious enough to go and find a way to see it. So, that’s like maybe not very research-oriented, it’s maybe a life lesson, but I think all of the research that we are doing has a hard science part and then like maybe a higher goal and I’m really interested in that higher goal, because once you have it, you will find a way to get there. Yeah.

Leonard Klar: I think that ties in quite well to what I learned from my research, that it’s- we had this theory going into the project that I’m working on right now, and this understanding also in the literature, which was not a 100% truth, but everybody accepted it the way it is. And then, by looking at it from another perspective, we found that maybe it’s not as simple as what was previously thought. So, with different methods we get a different perspective, which opens up new questions, but also shows us that some of the things we knew before were maybe not as we thought them to be.

Bahar Mostafiz: Yeah.

Leonard Klar: And maybe to sum up where we have common ground with our research, I would say that we are both trying to look at something that we know is more complex than we currently can look at, and that we both strive to understand a process and develop new methods to understand something that we know we can’t see with the correct resolution yet.

Bahar Mostafiz: Yeah, I agree. All the things, or the points that you just mentioned. And I guess we had a good talk. I want to keep this conversation going, and –

Leonard Klar: Yeah, I hope that people from different fields come together to discuss their research and maybe come to appreciate other topics and their own topics from a different perspective.

Bahar Mostafiz: Yeah! You’re always welcome to the lab!

Leonard Klar: Thank you very much, Bahar!

Bahar Mostafiz: Same to you!

[music]

In this episode, doctoral researcher Sara Simonen and project researcher Usman Azmat explore how innovative health technologies can improve emergency care. From AI-powered wearable devices used in ambulances to advanced laboratory tests for heart attack diagnostics, they discuss how faster and more accurate patient assessment can help healthcare professionals make life-saving decisions when every minute counts.

Listen to the episode

Transcript

I: The Science (-) [0:00:03.1] podcast. Today, we have Doctoral Researcher Sara Simonen and Project Researcher Usman Azmat discussing how wearable devices and laboratory tests improve patient care and can save the lives of many people.

Sara Simonen: Hello! I’m Sara Simonen, a doctoral researcher in the Cardiac Markers group at the Department of Life Technologies here at the University of Turku. I’m a biochemist by training and I have specialised in Molecular Biotechnology and Diagnostics. And today I’m happy to be in this studio with Usman Azmat. He’s a project researcher from the Health Technologies group at the Department of Computing. His background is in Mechatronic Engineering and AI for healthcare purposes. So, Usman, I was wondering if you could start by telling us what you are actually doing in your research?

Usman Azmat: Yeah, thank you, Sara. It’s really nice to be here with you. Our group is working on EU funded project (--) [0:01:11.8], and it aims to develop point-of-care devices for hospital (-). So, we are working on a cardio-respiratory device that is a multimodal device, and it detects the heart and respiratory systems’ conditions in the patients when they are travelling to the hospital in an ambulance. Other than this device, this (--) [0:01:36.8, the EU funded project] also has three more devices, and they work for similar kinds of emergency situations like strokes and some blood tests. And if I want to explain something in this regard, I would like to present it in a scenario. So, let’s think that a person is in an emergency situation and he had some attack, and he called an ambulance and an ambulance is there to take him to the hospital. So, the person is now travelling from his home to the hospital and the paramedic staff in the ambulance, they can hook up our device on his body. It is only one single wire, so it is really quick to apply on the body. So, when they apply the wire on the body, this device (-) [0:02:33.9] the ECG, electrocardiogram, and also the chest movement and the body temperature of the patient. From the data that the device collects, like ECG and respiration rate and also the body temperature, the vital signs are calculated from the device, and the vital signs are respiration rate, heart rate, heart rate variability. And other than the vital signs, this data is used to predict a disease that the person might be suffering from. So, all of this data is then transferred to an intermediate device that collects data from our device and (other three devices) [0:03:19.0] as well. And then as a whole, they send the data to the hospital. And the doctors can prepare even before the patient arrives. And they can give him the treatment as soon as possible. And additionally, they can also, in a case that there are multiple ambulances heading to the hospital, they can (-) [0:03:42.8], that’s the primary goal, as I said in the intro. So, they can assess the criticality of the patient, and they can see which patient is more critical. So, they will attend him first and the less critical one second. So, that’s what we are trying to make. And it also, in a way, makes the life easier for the ambulance personnel, like usually they are in a lot of hurry, so they need something really quick. So, this wire, our device, is really quick to apply and it handles everything on its own. Then the medical personnel don’t have to worry anything about it, and it doesn’t need any internet connection, and it can work standalone. 

Sara Simonen: Okay, wow! That definitely sounds very useful and very important. And that’s also something that I’m not very familiar with. You mentioned that all the measurements are sent to the hospital from the ambulance, but if the device doesn’t have an internet connection, how is it actually sending the data?

Usman Azmat: Yeah, that’s the interesting thing. We- as I mentioned, that this device is part of a big EU project, and we have three other devices as well. So, all of the devices send the data to an intermediate device that is connected to the internet and it then sends all of the data together to the hospital. And we cannot send the individual data, because it is sent as a whole package to the hospital and they have every information that they need. Otherwise, there is another benefit of- like our device can work without WIFI, and it enables it to work in remote areas where there are connectivity issues. And once it can collect data, then doctors can still preview it later on. So, this is a good benefit.

Sara Simonen: Yeah, that’s actually exactly what I was thinking. Like I was wondering “Okay, why is it not having an internet connection?” But that explains, that if you are in an area where it’s not possible, then it’s good to just collect the data. Okay. Then I’m also curious to ask about the AI model that you use. Like how is it trained? And how does it work?

Usman Azmat: Yeah, that’s really an AI- we are taking most of the AI model, like AI models are usually really large and they can- like running on the (-) [0:06:32.6] devices is a bit difficult for them, but we compressed the AI model, we trained it on real ECG data that was collected from the real patients travelling in the ambulance. And after training the model, we compressed it using some specialised compression techniques that compress the model as well as still keeping it as efficient as possible, so that we don’t lose the credibility of the model and we get reliable predictions from it. So, that’s the thing. We train the model on real ECG data and then we compress it and deploy it on the (-) [0:07:15.9] device. And then it is ready to make its predictions, and we can then send them to the hospital.

Sara Simonen: Yeah, this sounds like you’re already well advanced with your project. Like this is already in use, actually. 

Usman Azmat: Yes, and indeed, this device is already made and is with the clinical partners now. They are collecting some data, and they are testing it on some healthy volunteers, at this time. So, in the future we are planning to make it available for public use and patients can benefit from it. Well, all of this happens before reaching the hospital, but Sara, you are doing something that can improve the patient outcomes after the patient arrives at the hospital. Would you like to tell us more about your work?

Sara Simonen: Yes, of course. So, me and my research group, we are currently working on to develop a better laboratory test for heart attack diagnostics. And I also want to give a practical example to explain this. So, imagine for example that someone arrives at the hospital with symptoms that suggest a possible heart attack, which is very severe. So, of course, the doctors would then perform an electrocardiogram and take a blood sample, and from that blood sample then, they would test if the protein called cardiac troponin is elevated. And cardiac troponin, it’s a protein that is normally found inside heart muscle cells, and when the heart muscle is damaged, the muscle cells release cardiac troponin into the bloodstream. So, that’s why elevated cardiac troponin levels, they can be used as a sign that something bad is going on with the heart, and they may indicate that a person is having a heart attack. Currently, there are many cardiac troponin tests that are routinely used in hospital laboratories, but they have problems. And, for example, I could say that troponin levels, they can also be elevated in old individuals, or in people that have certain chronic diseases. So, those are people that are not actually having a heart attack, and that’s a problem for the healthcare professionals. Then they have to do multiple tests and different procedures to find out what is going on with that patient. And then, also, another problem that comes to my mind is that there are many things in the blood sample that can cause the test result to be falsely high or falsely low. So, in my doctoral research, I’m focussing on these specific challenges. Our goal is to develop a more heart attack specific and interference-resistant cardiac troponin test. 

Usman Azmat: You talked about some factors that interfere with the test, would you like to explain a bit, what factors can interfere with these tests?

Sara Simonen: Yes. A blood sample is a very complex sample, so there are so many things in the blood sample. For example, fats, antibodies and haemoglobin, then some medications that people are using. So, always when the test is in the development stage, a wide range of different blood samples containing these factors should be tested to see whether the test is really working reliably. 

Usman Azmat: What kind of the- most difficult challenge do you face in your research?

Sara Simonen: Okay, that’s a difficult but very good question. I would say that, in my opinion, the biggest challenge is the cardiac troponin, the protein itself, because it’s a three-dimensional structure. It’s not even completely known, and actually, it’s composed of three different parts. So, understanding what the optimal way would be to measure it in different conditions, it’s very time consuming and challenging. There are many research groups worldwide that are doing valuable work in this area.

Usman Azmat: Mm. Last but not least, how does your work make something beneficial for the healthcare professionals and patients?

Sara Simonen: Hopefully this test could, of course, help doctors to make more accurate diagnoses for the patients. Like quicker diagnoses, so that the patients would not have to stay in the hospital emergency for too long a time. And overall, like also this relates to your research, I think that my research area and your research area, they both have potential to improve the healthcare a lot and make life easier for the healthcare professionals. And also, these technologies could reduce the mental stress of the patients. That’s very important as well. Overall, the patient outcomes will be improved.

Usman Azmat: Yeah. I also agree. And my work is like more related to the pre-hospital settings, like a person is travelling to hospital in an ambulance, and we can save time there to help doctors be aware of the situation at first. And be efficient when the patient arrives in the hospital. And when your- your work is more related to the scenario when the patient is already in the hospital. So, we can combine, really, in a way, that when the patient reaches the hospital, the doctor uses your test, and even our AI prediction models, it’s just suggesting what the disease can be, but the final decision is still to the doctor. So, a doctor can use these couple of information and be efficient in saving lives, and we can, as a whole, contribute to people and make the healthcare better.

Sara Simonen: Yes, exactly. That’s what I really hope, and I see a huge potential. But also, what is really important to remember is that these technologies are still under development and we need a lot of studies to validate the performance of these technologies, but we are on our way!

Usman Azmat: Yeah, sure. This is a sensitive thing; healthcare is always a sensitive thing. And we need to be really sure of what we are providing, and that’s why we are validating and making sure that everything is really nice with the devices and they can benefit the people. So… Well, this is a point where we can conclude this interesting podcast with comments that every minute matters. When it comes to saving lives of patients who are in critical condition, the big aim of the presented works is to improve the patient outcomes at the hospital.

Sara Simonen: Okay, we hope you remember these three things from this episode. So, first of all, with these technologies, we hope to have less burden on the healthcare personnel. Secondly, we aim to have improved patient assessment before and after the hospital arrival. And thirdly, quicker patient diagnosis.