Roy Jakobs

So we use models from partners. Concrete example, we have a strong development partnership with AWS, where actually we are looking into imaging, as you mentioned. For example, the image acquisition system, the PACS, needs to be taken to the cloud. That's an effort that actually we both are looking into from our own perspective and from their perspective, how we can best support that.

So we use models from partners. Concrete example, we have a strong development partnership with AWS, where actually we are looking into imaging, as you mentioned. For example, the image acquisition system, the PACS, needs to be taken to the cloud. That's an effort that actually we both are looking into from our own perspective and from their perspective, how we can best support that.

What are the models that actually can help do so? So that's one of the example cases that is out there. We also use some of the models and then train on our own kind of data the model to kind of make it specific for a clinical application area. So give the example when we wanted to develop the smart speed, making an MR to scan faster, in essence, three times faster.

What are the models that actually can help do so? So that's one of the example cases that is out there. We also use some of the models and then train on our own kind of data the model to kind of make it specific for a clinical application area. So give the example when we wanted to develop the smart speed, making an MR to scan faster, in essence, three times faster.

What are the models that actually can help do so? So that's one of the example cases that is out there. We also use some of the models and then train on our own kind of data the model to kind of make it specific for a clinical application area. So give the example when we wanted to develop the smart speed, making an MR to scan faster, in essence, three times faster.

That's something that you do with your own data. So you use a base model, then you're going to train it based on your own data. And actually, we did it with a provider. specific provider in the Netherlands, Leiden Medical University, where they had 200 researchers on it. We had a team of 200 on it.

That's something that you do with your own data. So you use a base model, then you're going to train it based on your own data. And actually, we did it with a provider. specific provider in the Netherlands, Leiden Medical University, where they had 200 researchers on it. We had a team of 200 on it.

That's something that you do with your own data. So you use a base model, then you're going to train it based on your own data. And actually, we did it with a provider. specific provider in the Netherlands, Leiden Medical University, where they had 200 researchers on it. We had a team of 200 on it.

Together, you develop a model that actually makes sure and turns into an algorithm that is very specific for the use case. So it depends also on kind of what is the application area, what model you use. So we have a mix. We are not, I believe, the party that develops the foundation model. They're the big technology players that do that, and we partner with them.

Together, you develop a model that actually makes sure and turns into an algorithm that is very specific for the use case. So it depends also on kind of what is the application area, what model you use. So we have a mix. We are not, I believe, the party that develops the foundation model. They're the big technology players that do that, and we partner with them.

Together, you develop a model that actually makes sure and turns into an algorithm that is very specific for the use case. So it depends also on kind of what is the application area, what model you use. So we have a mix. We are not, I believe, the party that develops the foundation model. They're the big technology players that do that, and we partner with them.

But we are the ones that actually develops that into the clinical application layer that actually has the workflow solutions, that has the specific software solutions to improve the performance of products, to make sure that it actually lands in healthcare practice in a meaningful manner. And there I give the nurse example, the practical example of the increased scan time.

But we are the ones that actually develops that into the clinical application layer that actually has the workflow solutions, that has the specific software solutions to improve the performance of products, to make sure that it actually lands in healthcare practice in a meaningful manner. And there I give the nurse example, the practical example of the increased scan time.

But we are the ones that actually develops that into the clinical application layer that actually has the workflow solutions, that has the specific software solutions to improve the performance of products, to make sure that it actually lands in healthcare practice in a meaningful manner. And there I give the nurse example, the practical example of the increased scan time.

We use AI to actually help take noise out of the images when people go through a scan so that they don't have to retake a scan. We make sure when a scan is taken that actually we can take the most urgent and critical cases and put them in front of a queue for radiologists to look at.

We use AI to actually help take noise out of the images when people go through a scan so that they don't have to retake a scan. We make sure when a scan is taken that actually we can take the most urgent and critical cases and put them in front of a queue for radiologists to look at.

We use AI to actually help take noise out of the images when people go through a scan so that they don't have to retake a scan. We make sure when a scan is taken that actually we can take the most urgent and critical cases and put them in front of a queue for radiologists to look at.

So there are real tangible examples happening today, and that's indeed using LLM models for relevant use cases, but also still using traditional AI, numerical, kind of to come to certain predictions when we are in the ICU. When we look at patient deterioration in the ICU, our monitors can actually predict in 24 or 36 or 48 hours if there's an event going to happen

So there are real tangible examples happening today, and that's indeed using LLM models for relevant use cases, but also still using traditional AI, numerical, kind of to come to certain predictions when we are in the ICU. When we look at patient deterioration in the ICU, our monitors can actually predict in 24 or 36 or 48 hours if there's an event going to happen

So there are real tangible examples happening today, and that's indeed using LLM models for relevant use cases, but also still using traditional AI, numerical, kind of to come to certain predictions when we are in the ICU. When we look at patient deterioration in the ICU, our monitors can actually predict in 24 or 36 or 48 hours if there's an event going to happen