Category

Expert Series

10 Questions With….. Ludo Beenen

By | Expert Series, News

Ludo works at the Academic Medical Centre in Amsterdam in the Emergency Department as a Trauma and Emergency Radiologist. We interviewed Ludo during the COVID-19 lockdown when he was working from home a lot more. Ludo told us he feels the stroke workflow will be very different following COVID-19 compared to how things were done prior. He informed us stroke care and all emergency care has been more complicated due to the worry of contamination and spread of infection.

So firstly Ludo, how did you come to be an emergency radiologist?

A little bit of chance and a bit of luck. I am a broad oriented physician meaning I’m interested in a variety of things. Radiology concerns a lot of physiological systems in the body which appeals to me. When my predecessor left the emergency department I jumped at the opportunity to become an emergency radiologist! What I love about my job is that you always know there is room for improvement and things are always evolving. If you look at stroke for example, CT perfusion was discovered, then the success of endovascular therapy followed. We then had to work out how to efficiently transfer patients to intervention centers with all the important information, and this is where Nico.lab has come in. I have been very impressed with the performance of StrokeViewer and I really believe it is doing better than the others because it encompasses all aspects of emergency stroke care, from start to finish.

That is great to hear, what’s it like being at the heart of an emergency department?

Our Prime Minister recently said during the covid outbreak “you should accept that you will make 100% of your decision based on only 50% information that you have and you should be able to cope with that problem”. I think this is a great way to look at it, even in the emergency department. It can be difficult to accept this and people often want to take their time to make a clear diagnosis, avoiding mistakes. But my state of mind is that it’s important to not see it as a problem but a challenge, and do your utmost best to make the best decision from the 50% of information you start with. Everyone knows if you only have 50% of information, sometimes you’re wrong and coping with that in the emergency department is often the main issue for emergency radiologists. You have to put the different bits of information together like a puzzle and come to a conclusion as best you can. I enjoy that challenge the most, in a timely fashion. It will be interesting to see how artificial intelligence performs. Diagnosing a stroke patient is difficult. I am an expert so most of the time I can see where the problem is but certainly for juniors they can’t identify the problem as quickly so in that aspect artificial intelligence is very important and has the potential to take the pressure off.

When have you felt most challenged at work?

Well what is happening now is people have recognized more and more how useful the CT scanner is. When I first joined the ED we only had one sliding gantry CT scanner at the emergency department. But now we have two scanners covering three shock rooms. This means we can have three acute patients arriving simultaneously to the ED when a lot of hospitals would only have capacity for one. Basically triple the pressure which can be the biggest challenge when you have to make treatment decisions and give a fast and accurate diagnosis. But this is also what I enjoy the most about my job! 

From your professional perspective, can you explain some of the challenges and gaps in the emergency department?

The challenge is of course from an operational point of view. Everyone has to be very strict and everyone involved wants to do things their own way which can result in delay. People generally think when they are busy, they are doing valuable work but that is not always true given the circumstances! It is a skill of knowing what is necessary for the patient before things have even happened, and ensuring you stick to it and do nothing unnecessary. 

The second part is the infrastructure. If there is a lot of data that has to be calculated from the scanner this can take time. We now accept this and have to wait a minute or so but we know how valuable time is. A faster calculation of data would be really helpful. Maybe that is also where artificial intelligence solutions would work and a part of that is you calculate them and then it has to be transferred to the PACS system. That is something that is not working as fast compared to what I want. I know the moment it is scanned and calculated but i can’t see it. I still have to wait some minutes. People are not always aware of how much damage is done every second in the emergency department. When you are 15 minutes slower there can be huge consequences and that is something we should share. So I would say the main challenges are time and sharing the information.

In the current stroke workflow when a stroke patient comes into the ED how do you go about making your decisions?

This is really a team effort, and we regularly discuss optimization of the workflow with all the stakeholders. When a stroke patient comes in, I try to be as clear as possible, and want other members of the team to also work as efficiently as possible, without wasting time. Non-contributing tests should be avoided. It’s a matter of making a sequence of binary decisions in a strict manner, with always the possibility to adapt the workflow to the circumstances.

Is it a stroke: yes – what do I need to know? Is it hemorrhagic or not? Is another diagnosis possible? What about brain perfusion? And are there filling defects visible on the CT Angiography? Are the radiological findings and clinical findings comparable? What type of therapy should be started, intravenous or endovascular? This should be performed in a timely fashion: Time = Brain!

In your opinion where do you think AI could be most useful for you and how could it be implemented into the radiology department? 

Although it wouldn’t be typically thought of as AI, I think the calculation of penumbra and core is very useful, but still needs some further studies. It’s great that support from AI calculations is now accepted. In my opinion, if you want to implement a new AI powered tool into a healthcare system it needs to require very little effort from the hospital’s end. It needs to be something that is easy to implement, easy to use and easy to interpret. The moment it is presented to physicians with very little effort then they can accept that they have to know what it is all about!

Can you see how AI can have a helpful role in the stroke workflow?

Yes, there are certainly a few ways in which I think AI benefits the workflow. The most obvious role is the recognition of clots which for me is not the most difficult challenge when I can correlate it with the perfusion calculation. But some of my colleagues can have some difficulties in finding the clot so there is definitely an advantage in that. 

On a broad scale we have quite a few parameters which is great when trying to diagnose patients. The question we really want to know the answer to is what will the patient outcome be for a particular treatment. I know this means taking into account many parameters. Some interventionists are reluctant to treat older patients for example over 80 years old. But I believe for every individual it is different. Some 85 year olds may have at least another 15 years of valuable life. So this can be a moral problem. Hopefully in time when enough data is available we will be able to predict the probability of a beneficial outcome for a particular treatment.

What is your experience of AI in the stroke workflow and do you see it adding value in your opinion?

One of the most important things StrokeViewer does for me is the transfer of information of patients between hospitals and radiologists which can save so much time. When a patient is on its way to you, knowing all the information the primary center has found means there is no risk of wasting time repeating procedures. This has been of great value for us when trying to diagnose and deliver the best care. The system we had in place before meant that for half of the cases we were unsure whether CTA had been performed or we were unable to see the images so time would be wasted. Whereas now we always have the information beforehand and can make a quicker decision whether it is worthwhile the patient going directly to the endovascular suite. In that respect AI is great support for clot detection.

Are you excited about the potential impact of AI and the future?

Yes I think there are interesting ways in which AI can support several processes and have a positive impact, which is exciting to see. I also do have my reservations but it’s not a fear. Personally I think it is illogical to think AI will take over and humans will just be the support to AI. The human brain is very different compared to AI even though AI is the product of humans. I think the human brain functions very differently therefore AI will help support and strengthen us. 

 

Nico.lab improves clinical practice with innovative and trusted AI. Our product StrokeViewer, is an AI-powered clinical decision support system, offering a complete assessment of relevant imaging biomarkers within 3 minutes.

10 Questions With….. Bram van Ginneken

By | Expert Series, News

Bram is a Professor of Medical Image Analysis at Radboud University Medical Center and chairs the Diagnostic Image Analysis Group. He is co-founder of Thirona, a high-tech company focusing on the development of automated medical image analysis. Bram obtained his PhD on Computer-Aided Diagnosis in Chest Radiography and went on to develop software for tuberculosis and lung screening which is now used in many countries across the world.

How would you describe what you do for a living and what impact does it have?

What I typically say to people is that my team develops computer software that analyses medical images to do the same kind of thing that doctors do. At the moment, doctors are spending a large part of their job just looking at images and saying what’s wrong with the patients. They often have to search for tiny lesions like a needle in the haystack and since we don’t have enough doctors and healthcare is extremely expensive, developing this software can have a huge impact. Automating the scanning of images we can reduce costs and also make the job more enjoyable for doctors.

How did you come to what you do?

While I was studying Physics at university I saw an advertisement for a PhD position which was about writing a computer program to take a chest x-ray and predict if someone has tuberculosis. I was not planning to do a PhD but this was a very practical project and really appealed to me. I ended up becoming a professor and leading a big research group. Developing computer algorithms that do something really intelligent doesn’t have to be in healthcare but I ended up in that area and it led me to producing tuberculosis and also lung cancer screening software which is now in active use everyday all over the world.

Why do you think there is hesitancy when it comes to AI specifically in healthcare?

In my experience, doctors are open to using these kinds of tools and if they work well they are very happy to use it. If for example a license runs out and the software is no longer they will immediately call and insist it must work again. But if the software doesn’t work so well they are very quick to say it is unacceptable. But I think this is inevitable, when a new product comes on the market the very first version is usually not fantastic. The earlier doctors that have tried it are likely to say it is useless but then in time a much better product quickly becomes available and it is then taken for granted.

Machine learning was the big buzz 10 years ago but what’s your vision for the future?

I think the next big thing is actually implementing software that works. We are definitely at the very beginning. Medicine is not a sector where things move very quickly because it is heavily regulated when talking about patients and lives. So I hope the next big thing is having successful, well working software across the healthcare system but this will take longer than we think. I also hope this will keep healthcare affordable.

I think the end goal should be to make high quality healthcare accessible to more people all over the world, which can only be achieved if we keep it affordable. I think AI can help enormously with that. If you imagine our society without computers then that is the biggest difference when you think about 50 years ago. I think 50 years from now these computers will be doing the kind of tasks we are doing now. The question is then what is left for humans but of course humans should be in charge of things and should not go out of work.

If we are able to automate the simplest of tasks will this have a huge impact on healthcare systems in terms of freeing up physicians time and reducing costs?

Yes I think most research being carried out focuses on improving the quality of healthcare and a very small part is about reducing costs. If you look at what could have a big impact in society then I believe reducing costs has a greater impact than improving healthcare. If we didn’t research how to improve healthcare that would be stupid but there is so much emphasis and prestige around it and much less research looks at reducing costs.

In some cases improving healthcare leads to reduced costs for example if you automate what doctors do today then you save costs because computers can do things much faster and have a lower salary but I think a lot of the improvements in healthcare actually increase costs tremendously. For example super expensive cancer drugs improve healthcare a little, for a small percentage of the population, but at a huge cost. So I personally think improving healthcare ultimately will not always reduce costs but increase them. But if you are willing to save lives you will be willing to pay a high price for that.

Do you think that within the healthcare sector the implementation of AI should or could be faster?

Well it really depends, some tasks are simpler and can be implemented faster but others are more complicated and therefore take longer. I think right now we should aim to implement the simple things and then hopefully society will accept that we need to develop these algorithms and there will be funding to sort out the more difficult cases. A lot can be done but it is still a lot of work and time to develop these systems and it’s a big investment one has to make. It is not that implementation is technically difficult, like some radiologists may think because their IT system has said ‘our PACS system is not compatible’. Implementation is easy for example with StrokeViewer which has been cleverly designed to be easily implemented around the current workflow and PACS system. So technically it can often be done, but there must be an incentive, demand and a funding scheme so that it is worthwhile for companies to develop this software for radiology departments to buy.

Would you be able to comment on general AI and narrow AI?

So at the moment we only have narrow AI. They are often seen as separate things but in reality there is a continuum between narrow and general AI. Eventually we will get collections of narrow AI solutions which will make up general AI, for example finding all different relevant diseases on a chest x-ray. So for us, as software developers, it will become easier and easier. Even now I can see that what I am able to do with a student in 3 months used to take us 3 years. We used to train on data sets of 100 images and are now training on data sets of 100,000 images. Facebook is training networks with 2 billion images. Things are definitely moving and developing, it is very exciting.

There has been more and more focus on prevention and early detection rather than cure, if we continue to work with narrow AI will we be able to detect multiple diseases someone may face at one time?

Well preventing someone from getting a disease is the most important but that is more to do with lifestyle and is outside the realm of what I’m doing with image analysis. My work is early detection which I think is a completely different thing. But yes, AI products developing on the market today are for the most common abnormalities because there’s a market for it. I hope in the future we will have implemented AI for the more common diseases and can develop software for more rare diseases. In some areas of early detection we will have better scans although CT scans is a fully developed technology. It has not improved for the last 15 years, they are so good they measure almost every photon of radiation you give to the patient. To get a better scan you have to give the patient more radiation which is harmful so we won’t do that. In the future, yes I think things will be totally transformed by AI. We will be able to see diseases earlier if we want to, but then we have to decide what we want to do with that. For patients, people get very scared if they hear they have early stage cancer but we all have cancer, we just don’t detect it until is something that could be serious. In the future if we get better detection mechanisms we have to deal with that and I think it’s going to be very difficult for people to handle. If there are more possibilities of detecting disease early, do you really want to know that? Is it always necessary to know that?

Can you think of which living person you admire most and why?

Very difficult question, one of the big nice things working in science is you see and interact with very clever people. In many ways more clever than I am which is very inspiring. I am using deep learning to analyse medical images. But I didn’t invent it, those that did are so clever and I think it is so fantastic what they have developed. There are a lot of people I admire in the field.

Finally, are you excited about the future of AI within medical imaging?

What I find exciting and also scary is the development of the last couple of years. We now have these AI systems that can actually create images. So for example, a simple application is a very low dose image and you improve the quality of the image. There are also systems where you take an X-ray and predict a 3D scan and it actually looks very realistic. We know this from deep fake images, you may have come across this online when someone is made to look like they are talking and it’s actually completely computer generated and looks really realistic.

I think this has great applications in medical imaging analysis. We will be able to generate complete 3D scans from very little data but it’s also very scary. We won’t know whether a computer has made up certain structures in an image or not and whether we can actually trust the data that the computer generates. I think at the moment that is one of the most exciting areas in deep learning and image analysis.

 

Nico.lab improves clinical practice with innovative and trusted AI. Our product StrokeViewer, is an AI-powered clinical decision support system, offering a complete assessment of relevant imaging biomarkers within 3 minutes.

10 Questions with….. Fabien Scalzo

By | Expert Series, News

Fabien Scalzo is an Assistant Professor at UCLA and Director of the AI in Imaging and Neuroscience Lab there. He develops Machine Learning and Computer Vision algorithms to better our understanding of neurological disorders particularly Stroke and traumatic brain injuries (TBI). He does this using brain mapping of imaging and biosignals.

So Fabien, how did you come to do what you do today?

I was trained in computer vision and machine learning at the University of Liege in Belgium and carried out research to build new algorithms for visual learning. I became fascinated by how we perform these visual tasks as humans and interested to learn about the learning mechanisms in the brain. This was a big inspiration for my PhD work and one of the reasons I later joined   the department of neurology and neurosurgery at UCLA. This was 10 years ago, when I realised that machine learning, although had its limitations, could really make a difference for clinical care. 

Why do you think AI has proven so transformative to medical imaging and why is this particularly the case in something like stroke?

With stroke, it wasn’t even clear 10 years ago that imaging would be helpful in guiding decisions and there was a lot of controversy regarding the value of it. With machine learning, we let the computer figure out the patterns that really matter by retrospectively analysing the outcome of hundreds of patients. It reduces the manual information and the prior assumptions that we need to input and because we make better assessment of the patient, thanks to the algorithm, the patient may have access to the best treatment option; the one with the highest likelihood of good outcome. Machine learning streamlines automation and has an ability to extract information that we would not be able to comprehend or even visualize, making it accessible to clinicians.

What do you think about the hesitation to AI in the medical field? Where do you think it comes from?

One part of this is tradition. The medical field has operated in a certain way for so many years; very successfully in many cases. There is also a hierarchy amongst doctors and bringing on board a machine that is part of the decision process can be met with scepticism and in some cases it’s valid. There is a need for understanding what that machine is doing, I don’t mean being able to code it but get a grasp of the structure and the rationale behind its diagnostic output. 

What do you think needs to be done to reduce this?

The clinicians need to be able to give trust to the machine learning model which comes with having a good general understanding of it. Also regulation through the FDA is becoming even more important because they are the only organisation that the clinician can trust to adopt and regulate these tools. I think we need to customise special tracks for approving machine learning tools. These are very different from anything we’ve worked with before so even though it may fall within the category of a medical device there is another dimension to it. I see far too often clinicians taking for granted the ability of machine learning to always give the perfect answer and then if it gives a wrong prediction they immediately lose their trust in it. There is always some uncertainty and that’s ok as long as we are aware of it. I try to tell people it’s not perfect, it has its own weaknesses but as long as you know when that happens and what to look for, it is still useful.

Do you think there are issues with transferability from different technical perspectives?

Yes totally. There was a very big hurdle because we have two very different worlds trying to come together. On one side we have the clinical world which measures the value of the technology by how much it improves the care of the patient and the workflow. On the other side, the machine learning scientists normally measure usefulness  by how well the model performs with certain types of accuracy metrics. Even though these metrics may work well in the experimental setting it may have zero impact when transferred into clinical care. This is something I’m working on, trying to evaluate the machine learning not only with respect to the accuracy of the models but with respect to the clinical value in terms of decisions and workflow.

Do you think in the next 5 years we are going to see big changes in the merge of these two worlds?

Yes machine learning is very young and novel and will continually improve. At some point, machine learning will be at least on par with the best neurologists in the world for diagnosis so in that sense it’s going to be very interesting to see how it develops. At the moment all the models we have are extremely specialized for specific tasks but I think eventually models will develop to be more versatile and general so a neurologist will be able to see things that previously would have been missed. And some of that is not part of machine learning, it is common sense so eventually we want to bring common sense to the machine!

Is the best of AI currently in medical imaging or is it to any extent behind because of the difficulty with patient data and accessing it?

Today if you compare the magnitude of data in medical imaging to marketing/advertisement, we have extremely small data sets and they are for very heterogeneous diseases, where every patient is different. There are a lot of issues regarding privacy which need to be addressed. We are behind because we rely on data coming from a medical provider, such as a hospital, but across the world they do not have the infrastructure to deal with large amounts of data in an efficient way so that slows things down.

It is great that there are so many new companies like Nico.lab that build around existing infrastructure in the hospital, which makes it easily accessible to all physicians and renders the process of using data in a transparent and secure way which is what we need right now. 

If patients were more comfortable with sharing their data for research would this have a positive impact on AI?

Yes millions of people will benefit from sharing your data so it’s increasing awareness of that and working out an infrastructure that will make people feel comfortable about sharing it. But I think it will be a gradual cultural shift. 15 years ago Google was using our data and there was a big backlash but now most people don’t mind about that to some degree. It is part of the world we live in, as long as it is recognised and we know how it is used in a positive way toward betterment of society I think it is acceptabe. But some would disagree and that is also ok! One should have the right to not share their data. 

On a personal level, what is the most exciting aspect of your work?

Well two things. You can use machine learning to do finance and make a lot of money but you can also use it to save lives and that is what excites me the most. To be more precise I am very interested in stroke because it is a disease that is extremely hard to understand – some patients may return home with almost no side effects and others may not survive. It is phenomenal that soon we will be able to predict accurately the outcome of a patient for a given treatment course. This works because machine learning can incorporate data from so many patients with different treatment options so eventually the model will be able to make simulations of the likely outcomes for different treatments including the different benefits versus risks. It will be truly personalised medicine!

As well as prediction, all the different elements of AI streamlining and speeding up care will improve healthcare systems as a whole but do you think that AI will also save money?

Yes, AI will make the whole process from reading images to treatment much more efficient therefore reduce time and make work easier and more pleasurable for neuroradiologists which will of course reduce costs. It seems obvious to me but i guess it depends on how versatile the model is. If it is only for stroke, you need another one for cancer and screening brain injuries then you will be surrounded by AI. Versatility and some AI is a good direction in the future where we will have a model not just for stroke. But we are not there yet, we are in different compartments, but eventually I hope we can bring it together.

Nico.lab improves clinical practice with innovative and trusted AI. Our product StrokeViewer, is an AI-powered clinical decision support system, offering a complete assessment of relevant imaging biomarkers within 3 minutes.

10 questions with….. Albert Yoo

By | Expert Series, News

Dr. Albert Yoo is currently an interventional neuroradiologist, Medical Director and Director of Endovascular Fellowship & Research at the Texas Stroke Institute. He is an incredibly knowledgeable figure in the stroke world after his many years of research and training. He has participated in several major trials of intra-arterial stroke treatment, including MR CLEAN, THERAPY, MR RESCUE, DAWN, ARISE II, and GOLIATH trials. He is currently involved in the TESLA trial which he sheds more light on during this interview. 

How did you come to do what you do?

I stumbled upon it. I was always interested in Neurosciences. In medical school it was a very hard decision between becoming a neurologist, a neurosurgeon or a neuroradiologist. But during my neurosurgery orientation I saw the interventional suite and it was pretty soon after that I realised it was my path because it’s a nice intersection between imaging and therapy. 

After the dust settles, which stroke biomarker do you think will prevail? 

I am impressed with the simpler imaging approaches. A non-contrast CT and CT angiogram are very powerful tools. Part of their power is that they are available virtually everywhere and as we think about moving the field forward not only should we ensure that our imaging selection does not exclude patients who might benefit, we should keep an eye towards standardizing treatment decision making and delivery across all stroke centres.

Do you expect that collateral assessment like multiphase CTA will ever go out of fashion, and if so do you ever think that’s justified? 

Obviously collaterals are important because they sustain the brain tissue until the occlusion is recanalized. However, the clinical question is whether we need to be so fine in grading the strength of the collateral perfusion. I don’t think so. If you have a measure of the infarct size with ASPECTS and the infarct is not too big, then as far as collateral evaluation, all you need to exclude is the complete absence of collaterals. If there are no visible collaterals, patients will not do well regardless of rapid reperfusion as MR CLEAN and other studies have shown. Otherwise, it doesn’t matter whether the collaterals are intermediate or robust because in either case there is treatment benefit. I believe that traditional single-phase CTA is sufficient to say whether the collaterals are absent or at least partially present.

If there was a large increase in the use of MRI in acute stroke care across the world, due to increased availability for example, would you be happy with this and why?

Yes I think MRI is the most accurate imaging modality for determining the size of the infarct at presentation and to me that’s a key imaging finding to decide whether or not to pursue treatment, but as you are alluding to there needs to be improved availability of MRI as well as faster MRI workflow. With clinical experience, centers can have treatment times with MRI that are just as fast as with CT. This was shown in the recent GOLIATH trial that was conducted at Aarhus University hospital which uses MRI as their primary imaging approach. 

Do you think there is still a space for EVT patient selection within the 6-hour time window?

Within the 6-hour time window you need to have a very strong reason not to treat. That being said imaging is still necessary. You need a CT scan to rule out hemorrhage, and this same scan will tell you whether the ASPECTS is too low. You also need a CTA to evaluate the level of occlusion, and this scan will tell you whether the collaterals are absent. When it’s early you pretty much have to have evidence of zero collaterals or a very large stroke to not go to the angio suite. Fortunately, CT/CTA is very quick to perform. 

You are working on the TESLA trial, a very exciting project which could potentially have a huge impact for the field, what outcome are you hoping for and why? And what do you think the outcome will mean for the role of ASPECTS in stroke?

As an interventionist my hope is to provide evidence that expands the treatment population. We have very strong evidence in patients who have small infarcts and good collaterals. What is uncertain is whether patients with moderately large to very large infarcts will have a benefit with rapid reperfusion. I think there is a subset there which will benefit just based on the existing data, and probably many of these patients are not being treated currently.

Paradoxically, it will make ASPECTS more important because the trial will have established a threshold ASPECTS value that is critical for determining treatment response. At the same time, there would be a rationale to test whether more advanced imaging modalities can identify subgroups within the low ASPECTS population that might respond to treatment.

Recently there was a large Chinese trial that showed the non-inferiority of Direct EVT which was a pretty big deal, do you think this result will hold up in other trials? 

It’s important to remember that this trial only applies to the subset of patients who present directly to the endovascular centre where the time period over which IV tPA has a chance to work is very small, so it’s a result that is not surprising and I’m sure it will be confirmed in future trials. I think the key point to remember is that IV tPA is still useful in patients who are being transferred (drip-and-ship patients) for consideration of endovascular therapy. The real question is how does this trial result change practice for patients who present front door (direct admit to Endovascular Center)? There is a rationale based on cost to just forgo the tPA but cost issues aside there were still signals of benefit for tPA for early recanalisation obviating the need for intervention and overall reperfusion. In addition, there may be patient subgroups that might benefit from IV tPA that could not be detected with adequate power such as those in whom the occlusion cannot be reached. Also remember this is a moving target. It is likely that we will be using tenecteplase instead of alteplase soon. Do these results apply to tenecteplase? Probably not because we know that tenecteplase performs better than IV tPA for endovascular eligible patients. About 75% of subjects in EXTEND-IA TNK received intravenous thrombolysis at the endovascular centre.

EXTEND-IA TNK showed significant benefit of tenecteplase over IV tPA as a bridge to endovascular therapy. In more general stroke populations, there is strong evidence for the noninferiority of tenecteplase. Taken together, there is sufficient evidence to support using tenecteplase in place of alteplase. The fact that tenecteplase can be given as a single bolus dose further adds to its clinical utility.

What are your thoughts on nerinetide? Do you think it has the potential to be a game changer?

Neuroprotection as a bridge to intra-arterial reperfusion is a very attractive idea. This concept was tested in ESCAPE-NA1. Unfortunately, the primary analysis of this trial was neutral. The secondary finding that benefit was seen only in patients who did not receive bridging alteplase must be regarded as exploratory and requires confirmation. I believe that a subsequent trial is being planned in the IV tPA-ineligible population.

How do you think artificial intelligence benefits acute stroke care?

I think it works brilliantly as a screening tool to alert the team that there is high suspicion of LVO. The communication gets everyone coordinated in a parallel fashion. From the early experience at our center, it seems to really help with the workflow and speeds the time to intra-arterial treatment. Of course, the algorithms are not perfect but they don’t need to be. They just need to screen and capture most patients so that the LVO can be confirmed by the physicians. It is likely that AI will become increasingly involved in the actual selection of patients. We are seeing this in the realm of CT perfusion. In my opinion CT perfusion over selects — it selects patients away from treatment that should be treated and there is evidence to that effect. But I think AI algorithms as applied to imaging findings on noncontrast CT and also CTA collaterals will be very helpful. If you have a tool that has been validated to accurately tell you that the ASPECTS score is this or the CTA collaterals are not absent then it becomes a tool that can standardize treatment decision-making across centres. I think there are a lot of things that will continue to develop as these tools are utilised more and more.

Do you think covid-19 will have a positive or negative effect on progress of stroke research in the future?

I think it’s too early to say. Certainly, in the short-term there’s been a major impact. What we’ve seen in TESLA is that stroke volumes are down broadly. I think the likely reason is that patients aren’t seeking care as aggressively as they used to out of fear of coming to the hospital. But the crisis has highlighted opportunities to improve trial methodology in the future. As one example, we are finding that centers are now adopting electronic consent as a means to enroll patients whose LARs are not able to be in the hospital. This will help enrollment beyond the Covid-19 era.

Do you think it’s better to focus on complex stroke imaging techniques which will have a greater effect on the Western world or is it better to focus on more pragmatic techniques that would benefit more broadly including developing countries?

This is an issue that I think is so important. Many physicians in First World industrialized nations get very enamoured with advanced imaging technologies just because they are advanced and provide an “easy” answer. But I’ve studied the spectrum of acute stroke imaging over the past decade, and I find that the “easy” answer particularly as it relates to CT perfusion is often wrong and tells the physician to not treat when in fact the patient should be treated. The major misconception among physicians that use CTP is that perfusion can tell you whether the tissue is dead. This is obviously wrong. CTP forecasts tissue viability but just like the weather, the reality is often very different. Numerous studies have demonstrated this. 

The simple imaging techniques are powerful for multiple reasons. First, because noncontrast CT images ischemic edema which is a sequelae of neuronal death, it can tell you with a high degree of confidence whether the brain tissue is dead. Second, CTA is an efficient means of identifying the location of the occlusion and whether the collaterals are futile. Third, these tools are low cost and available to any hospital. I believe as a field that we should be rooting for these simple, pragmatic imaging approaches if we are going to impact care around the world, particularly in less developed countries. Clearly, doing a noncontrast CT and CTA is much more feasible than doing advanced imaging like CT perfusion or MRI. Committing resources towards developing AI algorithms that are based on the simple, pragmatic imaging approaches I think is very important as they will help to standardize treatment delivery worldwide.

 

Nico.lab improves clinical practice with innovative and trusted AI. Our product StrokeViewer, is an AI-powered clinical decision support system, offering a complete assessment of relevant imaging biomarkers within 3 minutes.

10 Questions with….. Ivana Išgum

By | Expert Series, News

Ivana Išgum is the University Professor of AI and Medical Imaging at the Amsterdam University Medical Center.  Graduating in Mathematics from the University of Zagreb in 1999, Išgum became a PhD student in the Netherlands in 2001 at the Image Sciences Institute. Obtaining her PhD degree in 2007, titled ‘Computer-aided detection and quantification of arterial calcifications with CT,’ Ivana then worked at both Leiden University Medical Center and UMC Utrecht, before embarking upon her position at the AMC.

Why do you do what you do?

Growing up in Croatia, both my mother and father were engineers. I remember being taken to my father’s work as a little girl and finding the whole experience and atmosphere so exciting. I knew I wanted to follow in my parents’ footsteps. I  came to the Netherlands 20 years ago as part of my journey to pursue that goal and never left.

Studying mathematics was quite abstract for me. I think that math is everywhere around us but we’re often not aware of it. It is fulfilling for me to connect knowledge with something that is applicable and visible in everyday life. So, I wanted to do something medical.  I wanted to strive for an academic career and surround myself with inspiring people. Of course getting funding for the research we want to do can be a struggle, but ultimately my job comes with the capacity to work on what I wish to work on, there is such creative freedom.  

How has AI proved so transformative to medical imaging? 

It has been the appearance of deep learning and the ability of algorithms to now be able to learn from the data themselves that has boosted the field enormously. Whilst AI is everywhere and indeed everyone is excited about it, for some this may appear as ‘hype’, but I really try to avoid using the word ‘hype’.  It implies that after a few years AI will somehow go away. That’s not the case, I am convinced that AI is here to stay, it will develop further, and it will radically change the way we work. 

Unlike in other industries, in the field of medicine we seem to not be as far as we could be. The main reason for this is probably accessibility of data. AI requires large sets of data that represents different populations. Furthermore, responsible application of AI in clinic while adhering to privacy legislation and ethical standards is far from trivial.   We are actively researching these issues so that the medical field can benefit from current AI much more than it’s already doing.

In a broad sense, how can AI currently benefit healthcare systems? 

Current healthcare systems are not sustainable as they are. The number of medical experts is limited and AI can help relieve their burden. AI might save costs, but it will more likely allow for a more sustainable system and it will certainly enable expert healthcare to be brought to parts of the world where there is a shortage of expertise. 

Where does reservation around AI stem from, and how can it best be addressed? 

One of the fears of AI comes from the caution of un-transparent decision-making. There are big, important and needed conversations currently going on about this. As experts, we do have to look at causality and make sure we validate thoroughly. I think the general public have a fear of AI being depersonalised but I don’t think it necessarily is, or should be. 

Are mindsets becoming more confident in AI powered tools in clinical practice? How does that feel for you at the forefront of research?

More people are seeing the potential benefits of AI to the future of healthcare, and resistance is decreasing. But  people generally resist change, if you look back at history it’s always been the way. Five years ago, people were panicking that radiologists would be replaced, and therefore not trained and that we’d see the takeover of robots. Now that hasn’t happened and undoubtedly never will. The medical world moves slowly and AI brings added value to routine work that radiologists can partner with and check. Ultimately AI makes it faster, more reproducible and cheaper than clinicians doing it all themselves, and that creates impact. 

Physicians understand AI in medical care is here to stay and people want to ride the wave. It opens the door for so many opportunities and for us who have been in the field for a long time, it’s rewarding. 

Where are we currently in terms of the potential of AI being utilised in clinical practice and what are the key challenges when it comes to implementation? 

From a research perspective, you see that a lot is possible. I would say the technology to answer many questions is available. But if you look at the products, the diversity doesn’t match up. For example the number of AI powered radiology tools used on a large clinical scale, is not very many, if at all. So the translation from research to clinical practice is really important. Regulatory bodies such as the FDA for example, are changing their strategy quite considerably in order to keep up with regulations that enable these tools to come into use.

Careful implementation is crucial. We cannot make mistakes. It’s important to consider all aspects of AI application including associated ethical and legal issues.  We are working with an extremely powerful tool and sensitive data; there is a lot to be discovered. Whilst we want to utilise AI to find information which is important to be found, we do not want to generate investigations which are not needed. These are all issues that are fundamental as we explore the possibilities of translating research into clinical practice. 

What do you think are the main technical challenges related to applying AI to medical imaging in comparison to using AI in other fields?

Medical image data is large and high dimensional, often 4D and 5D. At the same time, there are small sets of labelled representative data to learn from and the labels are often noisy. This can be challenging for analysis. Furthermore, data is quite diverse regarding image acquisition protocols, variability in anatomy and pathology. Designing methods that are robust to all these is not straightforward.

After the introduction of deep learning, there was a boom of papers applying similar techniques to different domains. Has the rate of progress slowed down or are there new interesting challenges that (if solved) could represent a major milestone in the field of medical imaging processing?

I am not sure whether the rate has been slowing down. We see that methodological developments from one research domain are being adjusted for application in others, which is good. However, we also see enormous competition in publishing where outperforming previous work, even if very slightly, is expected. While improving on the performance is possibly important and relevant, I think it is even more important that these are reproducible and that we try to understand why some approaches work better than others. 

What’s next in terms of key medical areas to benefit from AI?

It’s such an exciting time, we’re at the tip of what’s possible. There are very many possible applications of AI in medicine.Not only analysis of medical images but  the merging of different types of information can be very impactful and could help deeper understanding greatly.

As a successful woman in the field of AI, you must get asked about gender dynamics a lot, do you feel this question is still useful? 

Absolutely, it’s an incredibly important but sensitive issue. I dislike when people say it’s not a problem or that it’s not a problem for me any more. That is just not true. Ultimately we do not have enough women in science. As a society we are educating women but missing out on benefiting from their talent. Why? It’s such a complicated myriad of things at play (such as PhD and PostDoc timing coinciding with starting families, judgement in the workplace, at home, in the school playground etc) that ultimately work together to prevent society not utilising the full capacity of experts trained in science. 

Diversity is key for benefiting science. Some people say we need to encourage women, I think it would be a start if we didn’t discourage them.

Nico.lab improves clinical practice with innovative and trusted AI. Our product StrokeViewer, is an AI-powered clinical decision support system, offering a complete assessment of relevant imaging biomarkers within 3 minutes. 

10 Questions with….. Wouter de Monyé

By | Expert Series, News

Dr. Wouter de Monyé is the Medical Director of Radiology at Spaarne Gasthuis. Having graduated from the University of Utrecht in 1996, de Monyé went on to carry out research at Leiden University Medical Center and trained as a radiologist at Leiden University Medical Center and Royal Liverpool University Hospital.

Why do you do what you do? 

My motto first and foremost is to make people better. It’s a broad ambition but importantly it’s not just patients, but people, I strive to better. That’s why I love to teach and train within the hospital. I want to help my fellow medical colleagues come to patient diagnoses faster and more efficiently. 

What was your first job?

Funnily enough, cleaning a hospital when I was 15 years old.  My mother was a nurse and set me up waxing those long corridors. As a child I wanted to be a pilot, but this soon developed into the desire to become a flying doctor. You need to become a doctor first and the aviation part comes later. In the process, I discovered that radiology was my calling. I built computers on the side for a bit of extra money during my studies, but soon had to give this up when I started my research at Leiden University. 

What has been the highlight of your career so far? 

Becoming the Medical Director of my department. This has allowed me to influence systems and process and improve the circumstances we work in as doctors. At some point you think you’re doing your job and you’re being a professional, but ultimately it’s the same process time and time again. If you manage to get a helicopter view, you can see structural changes that could benefit not only one person, but a cohort of patients. I find being able to influence such change extremely rewarding. 

What is the most challenging aspect of your job? 

Getting the diagnosis right, particularly in acute situations. As a radiologist, you need to be able to ascertain what’s normal and abnormal and stick to it. But you also need to be able to know when to convey doubt. That’s the tricky part, especially when you don’t have time for a colleague to have a second look. Sometimes doctors put you on the spot – is it this or is it this? You find yourself having to give black or white answers in situations which are often pretty grey. 

Do you have any coping mechanisms for dealing with the stress of your profession?

That’s difficult to answer. I guess realising life can be over in the blink of an eyelid makes you focus on trying to build its most important elements,  like good relationships and friendships. 

Why is timely diagnosis so urgent in stroke?

The challenge to get things right is critical. Stroke is pretty common and its impact can be huge – everlasting, for the rest of a patient’s life. Quick diagnosis is key, but more and more research suggests even in cases with a larger diagnosis time interval, there is still benefit from intra-arterial thrombolysis. So having been proven in the Mr Clean trial that in acute situations this type of treatment option works, I think we’re now working towards expanding this approach to more patients. We’ll be seeing more situations whereby a possible clot has a chance to be treated in this novel way. I think the Mr Clean trial gets more relevant every year. 

Are there any challenges in the current stroke workflow? 

Getting the knowledge to the image, or the image to the knowledge. Doing the actual scans is no longer a problem. These days the CT scanners are able to produce high quality images in a very short amount of time. The challenge is interpretation. Those many images need to be analysed by an expert. Such professionals (ideally a neuroradiologist) are not always on call and a patient’s scan will often have to be interpreted by colleagues with less professional experience. 

Why is improving image sharing so vital? 

A waste of time is a waste of brain. If we can get an image to an expert as quickly as possible, that can be life altering for a patient. If a colleague needs help with interpretation, the current standard way of sharing images is difficult if you’re relying on your old fashioned Pacs system. You receive a call when you’re out and you have to go home, load up your computer, get the images up-  this can easily lose you 15 minutes which is absolutely critical. The shorter the door to needle time, the better the result. With StrokeViewer, I can immediately receive a patient’s CT scan on my mobile phone wherever I am. That saves time by enabling me to interpret the high quality images on the spot. It’s amazing and something I’ve never experienced with the regular Pacs system. Seamless image sharing makes everything so much easier. 

What are your thoughts on the added value of assisting image interpretation via artificial intelligence? 

 Assisting interpretation of CT images via artificial intelligence brings the bottom of quality up. It raises the bar.  If for whatever reason (timing, location etc.) a less experienced radiologist is interpreting the scans (which is common), AI can help him or her perform like an experienced radiologist. That’s something which is really becoming clear from Strokeviewer and is so helpful. As with any relationship, it takes time to build a relationship of complete trust. And similarly with any new technology, AI will need to prove itself in clinical practice. But we are currently seeing real life applications of artificial intelligence and it is amazing how fast my colleagues are beginning to rely and trust in it. It’s definitely going in the right direction.

How would you describe the current relationship between radiology and artificial intelligence? And how do you see this progressing? 

In radiology we’ve had digital images for quite a long time, more than 15 or 20 years. Even in the early stages there were various computer aided detection promises. But it’s only very recently, in the past 3 or 4 years, that deep learning and artificial intelligence, actually make possible what was promised all that time ago. I’ve worked with systems which have tried to do similar things but in a poor way. As with any relationship, it takes time to build a relationship of complete trust. And similarly with any new technology, AI will need to prove itself in clinical practice. But we are currently seeing real life applications of artificial intelligence and it is amazing how fast my colleagues are beginning to rely and trust in it. It’s definitely going in the right direction.

 

Nico.lab improves clinical practice with innovative and trusted AI. Our product StrokeViewer, is an AI-powered clinical decision support system, offering a complete assessment of relevant imaging biomarkers within 3 minutes.

 

Cookies helpen ons om inzicht in het gebruik van deze website te krijgen. Meer informatie

We gebruiken Google Analytics om inzicht te krijgen in het globale gebruik van deze website. We kunnen niet zien wie u bent, we kunnen alleen (anonieme) individuele gebruikers onderscheiden om te zien wie wat doet en wie bijvoorbeeld terugkomt op de website.

Sluiten