From Microscopic Patterns to Diagnostic Insights: Exploring the New Frontier of Blood and Inflammation with Prof. Johannes Schmid

Our authors:

Benjamin Obexer

Lead content writer, life science professional, and simply a passionate person about technology in healthcare

Elisa Opriessnig

Content writer focused on the technological advancements in healthcare such as digital health literacy and telemedicine.

Get to know Johannes Schmid and Explore his Cutting-Edge Research Project

Prof. Schmid has a long list of international publications. We present just a few of them covering topics like:

Benjamin: Hello Johannes. It’s a pleasure to have you here. Thank you for taking the time for this interview. I would like to start right off with information about you and your role at the Institute of Vascular Biology and Thrombosis Research? What are your main research interests?

Johannes: Hello, everybody from my side. I’m the Head of the Institute of Vascular Biology and Thrombosis Research here at the Medical University of Vienna. I’m not a medical doctor, I have biotechnology training as a background and then specialized in biomedical research. My main research interest has been inflammation in all different versions. I have been studying inflammation on the molecular level for many years. At this point, I was looking at which molecules are involved in inflammation processes.

For the last few years, I have been focusing on how inflammation contributes to different diseases such as cancer or thrombosis, and in particular cardiovascular diseases. As an example, I was investigating atherosclerosis quite a lot, because that is a chronic inflammatory state of arteries, which contributes to many life-threatening diseases, such as myocardial infarction.

Benjamin: That sounds very interesting. What drives and motivates you to continue research in this area?

Johannes: I think an important point is that, in the last decades, medicine was very successful in extending our lifespan. We live much longer than former generations. But now it’s getting more and more important to lengthen the healthy lifespan.

Chronic inflammation is a state that increases with age and it is predisposing the human organism to diseases such as cancer or cardiovascular diseases, as I have mentioned before. Since cancer and cardiovascular diseases make up about 50 % of worldwide mortalities, it is really important to understand the underlying chronic inflammatory processes.

What drives me is that I want to take a step further from the traditional medical approach, which is to wait until a disease is diagnosed, followed by standard treatment. I think it is essential to get beyond that by building up health care, which prevents the development of diseases in a specific manner.

It’s more and more recognized now that personalized medicine has to be used to treat patients more individually. My vision is that not only personalized treatment is done in the future, but furthermore, personalized prevention, meaning that you do not wait until the disease comes up.

To achieve that, it will be necessary to develop individual precision diagnostics, to uncover pre-morbidity states. Then you try to give your personalized advice, on how the individual patient could stay healthy. So, in a nutshell, personalized precision health care is my vision for the future.

A Glimpse into the Project

Elisa: Thank you for sharing that with us. We would now like to talk about one of your recent projects, which is about multifactorial fluorescence microscopy and AI-assisted analysis of blood smears. Please, tell us about specific techniques or methodologies you use in this project.

Johannes: We mainly use microscopy. First, we use brightfield microscopy as the traditional approach in which blood smears are analyzed on a transmission light microscope, in combination with certain stainings. Then we try to depict the different leukocyte subtypes. At this point, we want to extend that to fluorescence microscopy because this will give us an enormous additional potential to look very specifically for certain markers of cells or diseases. We tried to complement this step with techniques and methods, such as flow cytometry, where you do not get an image of the cells, but a very good quantitative assessment of, for example, leukocytes, our white blood cells.

Elisa: Okay. Since you mentioned blood samples, what types will be analyzed and where do you get them from?

Johannes: We first started with blood from healthy volunteers. So that’s my blood, those of my colleagues, and the team members. We can say that we give our blood for science [laughs]. But we also plan to investigate the blood of people with chronic inflammatory diseases. Additionally, mouse models allow us to analyze chronic inflammatory states of blood vessels. This is a really simple process and all the ethical guidelines are followed.

Elisa: Impressive. How do you plan to ensure the quality and also consistency of these blood samples used in the project?

Johannes: We have standardized protocols for the blood sampling and the staining of the blood samples. We also analyze the blood of the same person several times, to see whether we get consistent robust data. And then of course, there are certain quality controls where we check whether the sample is okay on the microscope before we do a more detailed analysis with it.

Elisa: The analysis is going to be AI-assisted. So how does our software IKOSA AI support your study?

Johannes: IKOSA AI is very valuable and helpful because it allows us to automatize the detection of blood cells. We started by training the system to recognize the different subtypes of leukocytes such as monocytes, lymphocytes, neutrophils, eosinophils, and so on. That is something that normally a pathologist would have to do manually or use flow cytometry methods. But with the latter, you do not get the morphology or the shape information.

IKOSA AI helps us to do microscopy analysis in an automated manner and much faster than what a human could do. And so by that, we quantify the state of the cells and can get information about the health status of the patient among other things.

We assume that acute and chronic inflammation will create structural changes in the white blood cells, which would not be seen in flow cytometry. This can then be detected with the help of artificial intelligence and image analysis.

Delving into the Realm of Blood and Inflammation

Benjamin: Alright. I would like to go a bit more into detail now and talk about blood and inflammation. Could you maybe explain the role of leukocytes and the inflammatory process? And how are they transported to the site of inflammation?

Johannes: You can see leukocytes as the policemen or the army of the organisms. Different types of leukocytes have different defense tasks, meaning they have to defend not only against invading pathogens but also against foreign material. But the point is that in blood circulation, they are controlling, and they often do not know where they are needed.

For example, when you hurt yourself bacteria enter the wound. Then leukocytes, the white blood cells of the organism, have to know where the wound is to fight against the invading pathogens. This is where endothelial cells come into play. Endothelial cells are those cells that line the blood vessels on the inside, facing the blood circulation, and are activated by these inflammatory triggers. Usually, local immune cells are required that release molecules that give a signal to the endothelial cells.

Then endothelial cells express docking sites for the leukocytes, so-called adhesion molecules, and by that, the leukocytes are fished out of the blood circulation. First, they roll along the blood vessel, and then they attach firmly. This way they can migrate and squeeze through the endothelial layer into the tissue where the invading pathogens are. They actively crawl toward the site of the injury.

Benjamin: I see. Is there a specific sub-population of white blood cells you want to monitor or are you interested in the whole population?

Johannes: We are interested in the whole population because the different leukocyte subtypes have different functions. They help each other in the defense mechanism.

In blood circulation, we find different percentages of these subtypes. The predominant ones are neutrophil granulocytes serving as the first defense line. They can release reactive oxygen species and by that kill the bacteria – and they can also swallow them.

But they are not as good at reporting what type of attack that was and which bacteria there are that have to be fought by the organism as monocytes and macrophages, which phagocytize parts of this foreign material. These cells can migrate via lymph vessels into the lymph nodes, presenting pieces of bacteria or pathogens to other cells of the so-called adaptive immunity. These are mostly T cells and B cells, which are subtypes of lymphocytes. They can very specifically react to certain pieces of pathogens. From this first line and non-specific immune defense, you get to a particular immune defense targeting only certain bacteria, surface structures, or also viruses.

Benjamin: You already mentioned some very important points, which lead me to my next question. There is this hypothesis about the blood carrying an information-rich signature. What is this about and why is it not yet routinely used for precision diagnostics?

Johannes: When there is an invasion by pathogens, leukocytes release certain signaling molecules, which also signal the bone marrow to release more leukocytes. Something that we see very early in infections is that young neutrophils are released from the bone marrow and sent to the site of inflammation or attack to defend against the invaders. And simultaneously, these molecules that are released, activate the leukocytes and other cells.

These activation states are usually not routinely evaluated in blood tests. Typically, people are just looking at what is the percentage of the different subtypes of white blood cells. They are not checking the specific activation states or expressed molecules. For example, lymphocytes split up into T cells and B cells. They are just summarized by the term lymphocyte. But T and B cells have completely different roles. T cells can even further be subcategorized into TH1 cells, TH2 cells, and cytotoxic T cells, and they also have different roles. In routine analysis, all these different states are not analyzed, because it’s not possible to do that with traditional brightfield microscopy.

Benjamin: And how do you think a better understanding of the role of blood might contribute to the development of new treatments for inflammatory conditions?

Johannes: I think that it is very important to discriminate between chronic and acute inflammation because acute inflammation is the physiological response of the organism that is transient. That goes back to a kind of quiescent state. That is just normal. But the problem comes with chronic inflammation.

There are different theories. It’s believed that, for example, repetitive infections over the lifespan finally lead to an increase in this general inflammatory state. From a normal, quiescent situation, you’ll get more and more into a chronic inflammatory state. Being aware of the role of blood cells is important to understand this shift from normal physiologically acute inflammation towards chronic inflammation, which then causes problems such as cardiovascular diseases, thrombotic events, increased tendency to form blood clots, and so on. We would like to understand that better.

However, the current problem, I would say, is that the markers usually used to depict inflammation, which is, for example, a blood marker called C reactive protein, cannot tell you whether it’s an acute phase or a chronic phase of inflammation. If that marker is elevated, it could be the sign of chronic inflammation, which is kind of slowly boiling, as we say, in the scientific community, or it could also be the declining phase of acute inflammation. Therefore, a more precise analysis of leukocytes from standard blood smears could then probably also identify the specific inflammatory states and distinguish between chronic and acute inflammation.

Benjamin: Okay, as you are doing research on blood inflammation, what is your experimental design? Do you use any specific mouse models to do your research?

Johannes: Yeah, when we set up the project, we said first, okay, it’s important to contribute a better understanding of what is going on in humans. But in humans, there is a high variability from one person to the next person, due to the different genetic backgrounds, the different health states of the people, and so on. Sometimes with humans, it’s difficult to see the pattern in the noise.

To cope with that issue, it is very good in a complementary manner, to build up mouse models, because then you have the same genetic background of the mouse strain that you’re investigating. You just change one element like one certain gene, which means you have a stable background, and then you’re looking at what is happening if you turn that specific gene on or off.

We have a mouse model where we can induce chronic inflammation specifically in arterial endothelial cells, by genetic means. That has a big advantage because we are not just injecting, for example, an inflammatory molecule, likely a lipopolysaccharide or something like that. But we are turning on inflammation very specifically, and only in the endothelial cells of arteries. That is something we have already successfully done to study atherosclerosis. But we have a second mouse model where we can do this genetic trick in all the endothelial cells including the veins. We think that this is very interesting because the recruitment of white blood cells to sites of inflammation is actually from veins and not from arteries. So in the case of arteries, it’s important for atherosclerosis, but in the case of veins, it’s vital for the recruitment of leukocytes to sites of injury or inflammation. Now we have mouse models for both situations.

Diving into the Complexity of Blood Cell Analysis

Elisa: Wow, that’s impressive. So now, we have some information about the role of blood and the techniques that are going to be used. I would like to dig a bit deeper into blood cell analysis, as we have shortly mentioned already. Can you describe the standard method used for white blood cell differentiation?

Johannes: The standard methods are various blood counters that can differentiate between the different leukocyte subtypes. That is usually then done complimentary by so-called flow cytometry. In flow cytometry, you have a cell suspension that is focused. One cell after the other is flowing through a measurement region, where a laser is shining onto the cell and then you can record the reflection of the lights from different angles. By that, you can get information on the size of the cell and the granularity.

And in addition, you can label the cells with antibodies that are linked to a fluorescent marker, and by that, you are able to discriminate different subtypes of cells. These are the standard methods, apart from the kind of routine blood smear microscopy that is done. Blood smear microscopy with brightfield microscopy is usually done, when there is something strange in the cell count or flow cytometry analysis. Then the pathologist would like to see the cells, meaning the morphology of the cells. By doing so, the pathologist might be able to say whether that is a state of leukemia where the cells change their appearance.

Elisa: And what are the advantages or limitations that come with flow cytometry compared to a blood smear method?

Johannes: Flow cytometry is a quite sophisticated technique. It requires rather expensive equipment and it is not available in standard diagnostic labs. Standard diagnostic labs have cell counters or they just do brightfield microscopy, the simple one. The limitation of flow cytometry is that it’s complicated and special training is required.

Also, you only get quantitative information about the cells such as the percentages or the expression of certain activation markers. There is no information on the cell shape. This is a major limitation of flow cytometry. The advantage is that you can analyze 1000s of cells within a second. You can stain with antibodies, as I had mentioned before, and with that, you can reach a high precision and a very good qualification.

Elisa: Okay. Could you explain the difference between the quantification of individual cell populations and capturing the different activation states of immune cells for instance?

Johannes: Of course. In routine health checks, for example, there is just this kind of raw quantification of leukocyte subpopulations concerning their percentages. That is done based on their size and the content of the granules. With that, you can discriminate between neutrophils, lymphocytes, and monocytes.  But beyond that, we cannot make any clear statements.

If you capture specific activation states using antibodies that are binding to specific activation markers on the cell surface, you can get a much more precise assessment of what the different cells are doing in that state. Are they activated? And if so, in which direction are they activated?

We do not only have one activation state, we have different activation states that need to be discriminated to then really assess what is the current function of the cell. What I forgot to mention is that for the capture of specific activation states, you need to use something where you can measure several parameters in parallel. And fluorescence measurements are ideally suited for that because you can make use of the different colors of the fluorophores. You can use various types of light to create or generate fluorescence. Then you can look at the colors quantitatively and achieve multifactorial quantification.

Benjamin: If I might jump in here with a question. Johannes, if I remember right, you also mentioned that sometimes white blood cells are covered with those platelets in specific inflammatory episodes. Is this measurable with flow cytometry? Or do you think that the blood smear analysis is the better option?

Johannes: That is a very good point. I think that is something that has been recognized in more detail just recently. These are the so-called platelet leukocyte aggregates. Platelets can bind to monocytes, lymphocytes, and neutrophils, and then they interact with the cells not only physically but also “talk” with the cells.

It’s known that platelets can, for example, activate neutrophils, that they perform a very specific function. They can expel their DNA in a process that is called NETosis, so they expel NETs. This is an abbreviation for neutrophil extracellular traps with which they can capture bacteria. Vice versa neutrophils can also activate platelets, and when platelets are activated, they contribute to blood clot formation.

There is a strong interconnection between immune defense mechanisms and blood clotting during pathologic processes. And this kind of sophisticated crosstalk between these two different biological processes is still not completely understood in all the details that would be necessary. These platelet leukocyte aggregates can be analyzed much better through microscopy because then you see a leukocyte decorated with platelets. You can also assess how many platelets are sitting on the leukocyte, what is the activation state of the platelets, what is the activation state of the leukocyte, and so on.

In flow cytometry, with some tricks, you can look for so-called duplicates (when platelets and leukocytes are detected simultaneously). You can also detect platelet leukocyte aggregates, but you do not have the detailed information that you would have with microscopy.

Elisa: Okay. I’ve already mentioned in the beginning AI plays an important role in the analysis. So, how will this multi-fluorescence staining and artificial intelligence be used to take this methodology to the next level?

Johannes: We believe that fluorescence technology for blood smears will raise the whole analysis to a completely new level. It allows you to analyze the fluorescence in different channels. You can examine many different groups and parameters in parallel. That is not possible with a standard technique. It will be necessary to use artificial intelligence to train the system and save time.

You don’t have to do it on an individual basis where you would need hours or maybe weeks to analyze all the data that is there. All you need is a computer that can do that in an automated manner. For that, you have to train it. The computer can do what humans can do but with much higher speed and precision. The computer has quantitative information not only on, for example, the level of expression of a certain marker, but also on the dimensions.

As humans, when we look at something, we can give a qualitative assessment. But we cannot say that the diameter of the nucleus is five micrometers or 6.5 micrometers. We can just say this nucleus looks a little bit bigger than the other nucleus. And this is where a machine can be much better than a human. I think it’s important to keep in mind that machines are not replacing humans or pathologists in this concrete application. But machines can help us to achieve our tasks much faster and with higher precision than what we could do alone as humans.

Elisa: Yes, and human error and human bias can be reduced as well. 

Johannes: I agree absolutely.

Elisa: And how do you see this ongoing research project impacting the field of medical diagnostics in the future?

Johannes: Of course, for the phase in which we are right now, we have a research application of all these different new techniques and technologies. But there is a long way to go to bring something from a research application to a real medical application. It has to be approved by different authorities as a medical product. But I think that we can try to contribute to that development of the field by providing information on how that can be achieved.

We can also give information on the limitations and the strengths and where we have to finetune the system to get robust results. I think in general this research can have a huge impact on the field of medical diagnostics in the future. But it will take its time of course, because there will be certain variabilities. There will be errors that we hope to eliminate during the project.

But my vision is that in the future every general practitioner could have such a system in his practice. GPs could actually when they are doing a blood routine test run an AI system over that sample. Maybe the images are acquired automatically and then the computer tells them “You should have a more detailed look at this or that patient” because this or that patient shows a specific situation. And again the machine is not replacing the doctor but the machine is helping the doctor to identify those patients who they should have a closer look at.

One of the important aims of our project is to get all this information with simple fluorescence microscopes, which can be afforded by general routine diagnostic labs or by a normal medical doctor such as a general practitioner. He or she can then apply these techniques and technologies without the need to get very special training or to purchase an instrument like a flow cytometer which costs up to 100.000 Euros. The entry-level for precision diagnostics should be as low as possible to provide precision health care.

Elisa: This brings me back to when you talked about looking at patients individually. You said differentiating between acute and chronic inflammation states is vital to provide specific and targeted therapy that a patient needs. You mentioned the time factor as well. Can you give us an estimate of how long it will take to have that in a GP practice? Can this be done in a human lifetime?

Johannes: This is always difficult to answer because we do not know how fast the development will be. My impression is that we have a very fast development right now. Many companies and research institutes are going in that direction as we have seen from ChatGPT and AI in Language models that were not that much discussed and only in specialist communities. Then suddenly it became super hype and everybody is talking about that.

That’s why I would not dare to give an estimate of how fast that will get into practice. But certainly not within a lifetime but rather within the time of five to ten years. I expect it to be first applied in the research field, in the academic field and when it proves to be valuable there then it will slowly move into the health care system. Maybe first in hospitals and university clinics that have an affinity towards newer research techniques and so on. Only after that it will kind of slowly shift over to smaller labs. I expect that this will be a rather fast process.

Elisa: Ok. Well, I am hoping to witness some of this.

Johannes: Yes, I am sure you will.

Elisa: That brings me to our final question for today. Just to round it all up. Could you briefly, in three sentences max, tell us what the future of multifactorial fluorescence microscopy and AI-assisted analysis of blood smears will look like? And especially focus on what you would like others and our listeners to take away from your research.

Johannes: I think that multifactorial fluorescence microscopy and AI will allow an easy assessment of health status and the diagnosis of chronic inflammatory states before real diseases come up and develop. It should be possible to counteract it early enough.

My takeaway message is to be aware that it is crucial to prevent the onset of disease and not to wait until you are sick and then try to get good treatment. You need a precision diagnosis of your organism early enough and if necessary you have to change your lifestyle or you take early actions to extend a healthy lifespan. So that you can stay healthy as long as possible with precision diagnostics and precision healthcare and you do not develop diseases that you would otherwise develop.

Elisa: It’s all about living a longer life with healthier years.

Johannes: Exactly.

Elisa: Thanks again for taking your time and for talking to us.

Johannes: Thanks a lot for your interest in our project and it was a very nice possibility to talk about it a little bit.

Benjamin: Thank you, Johannes, also from my side. It was a pleasure to have you here and we look forward to the second interview to talk about the results of your research and the upcoming publication in autumn.