View complete transcript
[00:00:30] Prof. Marina de Koning-Tijssen: It's a great pleasure to be here.
[00:00:32] Dr. Michele Matarazzo: So before we dive into our discussion, I want to give our listeners a bit of background of the Stanley Fahn Lecture Award. This award is one of the highest honors in the field of movement disorders, recognizing exceptional contributions to our understanding and treatment of these very complex diseases and conditions. Just to give a few names of previous recipients of this award, we can find Tony Lang, Andrew Lees, Eduardo [00:01:00] Tolosa, Kailash Bhatia, or just last year, Caroline Tanner. And I'm not mentioning many other distinguished awardees just for the sake of time. So congratulations on receiving this well deserved recognition.
What does it mean to you to be part of this highly selected list of people who contributed so much to our field.
[00:01:19] Prof. Marina de Koning-Tijssen: Yes. Well, I must say I'm extremely honored. I mean, at the end, you're doing your work and you enjoy it and many people spend a lot of time on what they're doing. And then you are the one who gets this award, which feels very much like an honor and rewarding for all the work that I've done. And I did it with great pleasure and that then my peers decide that it's.
It's the top level that really means a lot to me.
[00:01:49] Dr. Michele Matarazzo: Of course. And now, Marina, during the upcoming Movement Disorder Congress, you'll be delivering the lecture, which, as I understand, is going to be on the pleasure and importance of [00:02:00] phenotyping. Could you give us a brief overview of what you'll be discussing during the meeting?
[00:02:05] Prof. Marina de Koning-Tijssen: Yes, I've been thinking about that and I thought it's nice to sort of give a flavor of where I started, where, where the phenotyping was at that point and what I did at the start. And then in the middle of the lecture, flip over to, well, what we're doing now with machine learning and what I expect that the future will bring.
So I will, I hope to sort of give a sense of. How the field moved from, for myself and also for all other neurologists over time. That's when you get older, then there is a time difference from the beginning and the end. So that's nice to give a bit of a flavor of that.
[00:02:48] Dr. Michele Matarazzo: Well, that sounds very interesting and it's like a travel across your experience to know what you've learned from your experience. And so I think we will all be able to learn a lot from that. And you mentioned that your [00:03:00] journey also. Began from phenotyping and you're very known because you began with your research in hyperekplexia and latah, and then you did a lot of neurophysiology, electrophysiology can you share how all of this early experiences and all of this tools that you've used, how they shaped your approach to research and clinical practice?
[00:03:19] Prof. Marina de Koning-Tijssen: Yes, well for me, as you said, I started with hyperekplexia. That's a very rare disorder where patients have an excessive startle reflex and then get stiff and fall on the ground. And I had the privilege to study one of the first families that was described. It was a large family that my professor at that time, Professor Bruijn, was the original describer of that study.
And Well, what I noticed is that I learned so much from talking to the patients and looking about how they moved and what happened. And from that on, we developed a plan and we measured them [00:04:00] with electrophysiology, measuring the startle reflex. And, and then at that time, genetic was coming up. So, Around the time I started my research, they found a gene for, for hyperekplexia.
And then it was very interesting because within that family, there were two types, this full type with startle, stiffness, falls, and stiffness when they were as a baby. But there were also members in that family who only had an excessive startle reflex, and We learned from that by talking to them and then we measured them and we did genetics and it turned out that they had a different sort of reflex and they had not the gene.
And that learned me that you have to listen very carefully and look about how they walk, about how they move, and that tells you a lot about also these more complex backgrounds like electrophysiology and genetics, which you can use those. To sort of support what you're [00:05:00] finding in the clinics.
And, and then over time, because I did this startling, it was interesting because many neurologists in the Netherlands then sent me patients. Who startled in some way. So I came across many very interesting rare patients, like patients with paroxysmal kinesogenic dyskinesia that were luxated by startle, patients with epilepsy that were luxated by startle.
So I had this whole range and Just by talking to the patients and learning from them and measuring them, that gave me so much insight about what it all meant and what the difference were. And that was a great pleasure. And then after a while, I had a grant to go to Indonesia, because I've always been fascinated by the jumping Frenchman of Maine.
It's in Maine and in latah in Indonesia, and that [00:06:00] gave me the chance to measure those patients as well and talk to them with a translator and, and measuring the startle reflex. And that gave me so much insight about the whole spectrum and, and that's what I really enjoy doing.
[00:06:16] Dr. Michele Matarazzo: That's great. I think we can learn from all of this, the importance of observation, which is obviously something that all of us who dedicate ourselves to movement disorders, we know how important it is to look at patients, but also listen to them very carefully and, and the way of carefully selecting tools that can be helpful.
You know, because in your case, electrophysiology and depending on the phenotype and on the observation, electrophysiology, neuroimaging or other different tools can be can be applied and useful to what we are seeing and listening to. And, and actually you were mentioning before that now you're, these days you're applying something newer and, and very promising, which is and artificial intelligence and machine learning. [00:07:00] To the phenotyping process.
Now that's obviously quite a leap from traditional clinical observation or even neurophysiology. How do you think that machine learning has influenced or, or is influencing your research and or even your clinical practice? How do you see, now that we are talking about that, how do you see the role of clinicians evolving alongside these new technologies?
[00:07:23] Prof. Marina de Koning-Tijssen: Yes, well, there is a very nice paper from Mark Edwards and Anna Sadnicka, and they wrote about the importance of phenomenology, so the clinical phenotyping, because these days, We all tend to say, okay, we go for the genetics because from, from genes you can get different phenotypes, but different genes can give the same phenotypes, so we should stick to the genetics.
But in that paper, Mark really nicely describes that's at this point, it's extremely important that we [00:08:00] phenotype patients very carefully, also the motor phenotype, because although there are hundreds, perhaps even thousands of causes for movement disorders, there are only a limited number of phenotypes like myoclonus, chorea, dystonia, parkinsonism, and so somehow all these different causes converge to these few phenotypes.
And at this moment, it's hard to, to develop like treatments based on the genetic part that might come in the future, but not now. At this moment, we mainly focus on treatment on the clinical phenotype, and that can learn us a lot about the circuitries that are involved. So, I think it's very important to, to classify the patients properly with their phenomenology.
And saying that, I think that what we are currently doing is a big [00:09:00] project where we phenotype groups of patients where we do EMG accelerometry and video. And then we have in each group we have 20 patients with a single phenotype like cortical myoclonus, essential tremor, or dystonia, and we focus on the arms.
And I don't think that this tools will replace the phenotyping that the clinician does. But the aim is to support a clinician because, as you also know, it can be quite hard to differentiate. And then if you're on a clinical basis, think, well, I think this is a tremor. And the machine tells you, well, I think for 80 percent this is a tremor based on different features that they found.
Then you are supported in your clinical idea. So I don't think in the next 5 or 10 years that the machine will replace the clinician, but it can help you to support you in your [00:10:00] idea, so more neurologists are able. to correctly phenotype a patient. And another thing that is very important from the clinical side is when you learn the machine, and if you put it simple you have to put in the right patients.
And, and what we do now is that we send all these patients with videos to experts worldwide. And there are three experts per patient and they must agree before we put them in for that phenotype. So the machine must learn. from the clinical view. So I think that replacing is not the word, but it can help you support in what you think is the phenotype.
And we need the clinicians, as I said in the beginning, with the curiosity of what am I seeing. And we have to start with the straightforward phenotypes, the simple phenotypes, and then the next phase will be the more complex with more than one [00:11:00] phenotype, and that will be another next challenge, because we noted already that if we ask experts to classify combined phenotypes, It's very difficult to find the consensus there.
[00:11:15] Dr. Michele Matarazzo: The agreement is not so high, right?
[00:11:17] Prof. Marina de Koning-Tijssen: No, it's not so high. So there is really a challenge. So we, we start with the single phenotypes because we must start somewhere to learn the machine and what will be interesting from the machine learning that you can, it can happen that. At the end, when we have put in all the patients, also the combined, that he makes new groups, right?
So not only the groups that we label, but also subgroups. And that would give a lot of insight as well. So I think we can also learn from the machine learning. If, if the machine says, well, this is a different group, we can go back to the patients, have another look and figure out based on the features, why that is a different group and what can that [00:12:00] learn us.
From what's going on in the brain, the circuitry, so that will be very interesting as well. So, no, not, it will not be a replacement, but it can be a help. And where it can also be helpful is once you say, well, it's, it's like myoclonus, it can be helpful in seeing if the treatment works.
So if you do an intervention and you have a very objective measure, at the moment it's usually scales that we use, clinical scales, but I think their machine learning could help to quantify the severity of the disorder. So I think that's another field where it will help us.
[00:12:42] Dr. Michele Matarazzo: Well, a lot of possibilities. I really liked the idea of what you were telling me before about teaching the machine, but also learning from the machine and then go back and forth. I really think If an artificial intelligence find a different group, which you haven't [00:13:00] thought about before, it's really kind of a new person who is, or a new entity who is making a different type of observation and listening to it could could give us new ideas when we try to help our patients.
So that's, that's extremely interesting apart from all the classification and biomarker things that you were mentioning.
[00:13:19] Prof. Marina de Koning-Tijssen: To add there, what I'm really excited about is it would be amazing if we could either with EMG or accelerometry or with the video analysis, if we could find biomarkers for dystonia, you know, that would be fantastic. Because I mean, if we discuss dystonia and we see a patient, I try to learn young people to tell them what I see, but it's usually the whole pattern, right, and that's something you do by training, seeing a lot, but at the end of the day, someone has to say, is it dystonia, yes or no, and what you notice, the older you get, they give you more credits, right? [00:14:00] But I'm not sure if I'm always right. If we could have biomarkers there to support us, that would be very helpful as well, I think.
[00:14:07] Dr. Michele Matarazzo: Great. Now, changing a little bit the topic I want to talk about a little bit about your career. I mean, you're receiving this award for all you've done so far. So now looking back over your career, how do you think the field of movement disorder has evolved since you first started? And where do you see it's headed in the next decade and how do you think you will contribute to this, to this change?
[00:14:37] Prof. Marina de Koning-Tijssen: Well, if I think from when I started till now, I think what has been a revolution is the genetics. There is so much more known about, I mean, When I started, there was just a Mendelian, and then we looked for a gene in one family, and then we've tried to well, we still do that, but I mean, that's, for the geneticists, it's easygoing.[00:15:00]
But now it's all so complex, and I think there has been a major advantage. And for treatment, I think what I've seen from when I started is that the DBS is something evolved over time and it's nice to see now that with. And their machine learning can perhaps also be helpful that we can sort of have these feedback systems to improve the programming, because for quite some time, there nothing changed that much, although it was very effective.
I think there will be a major improvement possible And also I realized that when I looked at my old papers, when I started, we had only the CT scan, you know, can you imagine, and no MRI at all. So now, if you see all these imaging tools, it's so much more clear to see what the effect is in the brain.
So I think for all these additional [00:16:00] tools, there has been a lot of development over time. That's really amazing actually. And I think the challenge will be for the next periods, are we able to integrate all that information, you know, because that will learn us a lot. And as I said, I think we should Also from tradition, but also because we know a lot about it and it will learn us about the brain, start with the clinical phenomenology, but we will be able to add all these things to it.
And that's, too much for a human brain to do that. So if we could use tools there to add that all together, that will really improve our diagnostics and also insights in the pathophysiology, I expect. Then you also have the metabolomics. I'm not really into that, but that's another field that's really interesting.
So I think that those Bioinformatic [00:17:00] tools and machine learning that's something that will develop over time and, and that also illustrates for me that we will need people to do that because that's too, I think there will not be many. Perhaps I'm underestimating people, but this is not really something that clinicians can oversee, I'm afraid.
So we have to collaborate more in a multidisciplinary way. And because as I noticed when I work with people from the machine learning department in our university, it takes about a year before you understand what you're talking about. For them, features is something totally different than what a feature is for me.
And that's In the beginning, I
[00:17:42] Dr. Michele Matarazzo: speak different languages, right? So we have to get to a point where we speak the same language to, to then advance together. Yeah.
[00:17:49] Prof. Marina de Koning-Tijssen: Yeah, because then you can really make steps forward, because then you can tell them what your problem is. They can understand and they can help you. And then once they have some [00:18:00] results, you can talk back to them because you understand what they're doing. But that's really, that's really a new field.
I find it very exciting. And I think that the next generation will be much better in that than me, because they are much more into computers and things when they're very young.
[00:18:16] Dr. Michele Matarazzo: So yeah, then the next step forward is to, as you were mentioning, to try to integrate all that information, thanks to all these new tools that you are also pioneering with the application of machine learning to phenotyping and electrophysiology. And I think there is something that you mentioned before that is very important talking about this next step, which is you were feeding the algorithm information from high level specialists, three different high level specialists. So the information that we feed this new algorithm that we want to train is very, very important because if not, there is this concept of garbage in, garbage out, which
means if you feed information that is not correct to the machine learning, then you will never be able to trust what comes [00:19:00] out. So, and I think that in, from that point of view, it still is very important the origin of all this talk is, which is the phenotyping and observation and listening to patients, right?
[00:19:11] Prof. Marina de Koning-Tijssen: Yes, and I think that part is really our part, right? So we should also be proud of that. And I fully agree. I'm always saying the same things. We have seen that with databases in the past. If you just let it all go in. And there is always garbage in, garbage out. So you also with databases, you must have good clinicians putting them in because then you know what goes in is right for as far as you can know.
But that's very important. And I think in that way, clinicians are very. And also, like I said, in the beginning, just before that we, we treat based on the phenotype, right? So we also have this experience about what is the effect and I think that's very [00:20:00] valuable and that is not replaceable by a machine, but it can help and support us.
That's, I think that is the main thing. Yeah.
[00:20:08] Dr. Michele Matarazzo: Now, as someone who has certainly inspired many in the field, what advice would you give to young researchers and clinicians who are just starting their careers?
[00:20:18] Prof. Marina de Koning-Tijssen: I would say enjoy yourself. I think that's always the main thing to say. Find something that really comes to your heart and that's feasible. So focus you don't have to focus on a very small part, but focus either on a specific part in Parkinson's or on DBS, or on rare movement disorder.
So try to find somewhere a niche that you're really interested in. So then it's easier to approach people in the field who are doing that as well. Collaborate. Find a network. I think that's a very important one. And and also, well, that's a tough one, but [00:21:00] try to keep a bit of your autonomy, so that you can do what you like to do.
And, I notice, I have quite some fellows and young people coming to me from other places all over the world, but also from the Netherlands. And sometimes they're so forced to do what the supervisor wants to do. So, And I would say, well, be a bit, not always following, let's put it that way.
Because if you have a nice idea, you should go for it. And and that gives a lot of pleasure if it works out. As I noticed with this startling, because I did that, Then people like it that you do that, and they send you patients, and then at some point you broaden your field, like when I did this hyperekplexia, then I moved to myoclonus, and then I moved to myoclonus dystonia, and because you're so much into something, then you can broaden up later.
[00:21:59] Dr. Michele Matarazzo: I think [00:22:00] that's that's really important. It goes with the first thing you were saying, which is enjoy what you're doing. Right. So I think one way of really enjoying what you're doing is that you have your idea and you go for it.
[00:22:11] Prof. Marina de Koning-Tijssen: Yes, absolutely. And another very practical thing is these days as well, both for men and women, I mean, it's also the time that you get your family. And get it well organized. Even if it costs you money, just do it. Because well, then you have quality time with your children. And and I think that's also an important part these days.
Because in the old days it was more traditional. But these days it's both people, both mom and dad are working. And then well, make it yourself as easy as you can for the practical part. And that's, I think that's a very important thing.
[00:22:52] Dr. Michele Matarazzo: Oh, definitely.
[00:22:53] Prof. Marina de Koning-Tijssen: Yeah.
[00:22:54] Dr. Michele Matarazzo: Now and, and I know that you have still a lot to contribute but I would like to ask you one [00:23:00] question which is in the future, what do you hope your legacy will be?
[00:23:06] Prof. Marina de Koning-Tijssen: Well, what I hope is that I've trained a lot of people over the years and that they will bring the field forward. I think that's my main goal. And I also notice now, because, well, I have to get this price now, so this is really high level that I sort of start doing more and more. I always did that to facilitate young people.
And to help them get grants and, and things like that, positions because I think that will be my legacy that they learned from me and take it forward because you can never have a legacy. I mean, things change over time, so I hope they remember me as, well, that I push them forward, facilitate.
I think that's the main word. I like that very much, yeah.
[00:23:58] Dr. Michele Matarazzo: That's great Marina. Thank you so [00:24:00] much for sharing your insights and experience with us today. It's been a fascinating conversation and I'm sure that our listeners have gained a lot from it. Is there anything else that you want to share with our listeners before we close the interview?
[00:24:14] Prof. Marina de Koning-Tijssen: Well, I would say that you are a great example of how it should go, right? You're ambitious, you're having fun. I've seen you once in Groningen, dancing with all of us, and I think you're good. You're doing great and you're also having a family and you're still doing all these things and and especially like I said you show that you're having fun and I think that's the main thing so I would like to add that.
[00:24:41] Dr. Michele Matarazzo: Thank you so much. I mean that means a lot to me, obviously. So thank you very much. You definitely have been an inspiration for myself personally. So thank you personally. So I know there are listeners and audience here, but it's, it's good to have this personal moment between you and us.
So thank you. Before we sign off, I want to remind [00:25:00] our listeners that Professor Marina de Koning-Tijssen will be delivering the Stanley Fahn Award Lecture at the upcoming Movement Disorder Congress in Philadelphia on September 28th, 2024. Be sure to attend if you can, if you're there, it definitely promises to be an enlightening session. Thanks for listening and thank you, Marina.