Selected Podcast
How AI is Improving Stroke Care
Powered by AI (artificial intelligence) on a single platform, enabling care teams to save time, and improve the lives of stroke patients.
Featuring:
Moussa Yazbeck, MD
Dr. Yazbeck is the Stroke Medical Director for Tenet Health Central Coast. He is dual fellowship trained and Board certified in Neurocritical Care and Critical Care Medicine. He is the Past President of the Society Of Critical Care Medicine Northern California Chapter. He has special interest in taking care of patients with severe stroke, subarachnoid hemorrhage and traumatic brain injury. Outside his work, Dr Yazbeck enjoys spending time with his family, traveling, skiing and biking. Transcription:
Prakash Chandran: A stroke occurs when something blocks blood supply to part of the brain, or when a blood vessel in the brain bursts. Because of the nature of this medical emergency, time is of the essence when it comes to treatment. Luckily, there have been technological advancements that leverage artificial intelligence to accelerate care coordination, reducing systemic delays that stand between patients and lifesaving treatments.
Host: We're going to talk about it today with Dr. Moussa Yazbeck. He's a stroke medical director for Tenet Health Central Coast. This is Healthy Conversations, a podcast from Tenet Health Central Coast. I'm your host, Prakash Chandran. So Dr. Yazbek, thank you so much for joining us today. I really appreciate your time. Before we get into AI and how that's leveraged in treatment, I'd love to know what the standard ways that strokes are both identified and treated today.
Dr. Moussa Yazbeck: Of course. Thank you, Mr. Chandran, for the introduction and for having me. When a patient presents with stroke symptoms such as weakness, speech difficulty numbness, new visual deficits, and many other symptoms which providers can recognize, we first perform an examination and we get a detailed history and timing of the onset is very important. After this, we do a CAT scan of the brain, which is ordered in a very expedited fashion for the obvious reasons. The reason why we do the CAT scan is to differentiate between an ischemic stroke, like you mentioned, from a clot in the blood vessel which are the majority of strokes or probably two-thirds of the stroke or hemorrhagic stroke, which is a bleeding or a blood vessel, which bursted in the brain, which constitutes a third of the stroke. Hemorrhagic strokes have a very different pathway of treatment, and we're going to discuss today mostly the ischemic strokes when you have a clot in the blood vessel.
Guest: Once you establish a, you know, potential or possibility of an ischemic stroke, we follow a very specific approach and a protocol. And as I said earlier, time is of the essence because time is brain. When you perform the we decide whether this patient qualifies for a medicine, a clot-buster so to speak. The medical term would be a thrombolytic, and some of those medications are called Alteplase or commonly known as tPA or TNK. And those medicines help with dissolving the clot, and there's specific criteria for who qualifies for this intervention or for this medicine. And usually, it has to be given within four and a half hours. Once we decide on that part, we go to the next step, which is, does the patient qualify for thrombectomy? A mechanical procedure where you go with the catheter either from the groin area or the wrist and remove the clot mechanically and open the blood vessel. And for this, we do a CT angiogram and a CT perfusion, a CAT scan also with some dye to look at the blood vessel and some fancy images where we look at the brain and which areas of the brain are still salvageable or we can help or which areas are too late to do any intervention on. So, that's the standard approach, a pretty good approach, but that's the standard approach before AI.
Host: Yeah. So, that identification process and figuring out whether a patient meets that criteria is an important piece to this. And I imagine that that whole process takes time, a lot of coordination. So now, let's talk about these new developments and technology that leverage AI for stroke identification and treatment. Talk to us about what we can expect from this new technology.
Guest: Absolutely. So now, we have artificial intelligence, AI, in medicine like many other fields which are using AI. And the idea is to detect those patients even faster than we were doing prior and see who qualifies for an intervention faster, mostly a clot removal or thrombectomy, and act accordingly and get them to the right place faster and basically, you know, this affecting their care.
Host: Yeah. That's amazing. And are you able to tell or talk at a high level around how the technology does this? Is it just that it has seen so many patterns of what, for example, a patient that qualifies for a thrombectomy looks like that it can just tell you faster? How does it work?
Guest: Absolutely. So, I mentioned earlier in the pre-AI era, which is only a few years ago, that we do a CT angiogram and a CT perfusion, basically a CAT scan with dye, and there's some fancy images we look at called CT perfusion, which is great technology, and that's how we triage patients for thrombectomies to see which patients have occlusions in some of the large vessels we can get to.
So, basically, what we did before this, we do those images and then we upload those images to our systems, whether it's Central Coast or any other hospital which may have this technology. And then, the providers, specifically the stroke neurologist, me, the radiologist, the ED doctor, the interventionalist, and whoever else is involved in the care. Once those images are uploaded, they would have to look at those images separately and subsequently communicate with each other, often one-on-one by phone call or so and to come up with a plan. Now, we have artificial intelligence technology at Central Coast, which basically minutes after performing the study, the studies are downloaded to to the cloud and then pushed to all the players, us, the providers I just mentioned. Whether you're using your phone, for example, I use my phone, a tablet, a desktop, whichever device you're using. Not only this, once the study is performed, just minutes after it's uploaded to the cloud, it does prompt me, it sends a beep, prompting me that it was done. So I don't need a phone call or me keep checking. I'm able to look at the images on my phone and it's pretty impressive the quality of those images. I can zoom in, I can zoom out. There's 3D technology where I can turn the images and basically it gives me all the information I need. Not only this, the AI software is able to recognize those clots even before I make this decision with a very, very high sensitivity and specificity. And once it detects this clot, which, in our terms, we call it LVO, large vessel occlusion, it sends me multiple beeps or a sound, whichever you decide telling me that, "Doctor, there is an LVO going on," and that some often happens literally minutes after the patient gets the study done.
Of course, nothing replaces a physician or a specialized physician to look over those images and, you know, putting it in clinical context, which is the most important thing here and agreeing with the software. But you know, it's very infrequent that we don't agree. As a matter of fact, I had one case where I didn't personally appreciate the clot and the software did, and I went back and the software was correct.
So saying this, once this is all done, now we're able to communicate between all the physicians or the provider through this application, of course, it's HIPAA compliant and making decisions very, very quickly because, you know, we're all on the same page. We're communicating those studies and then we act upon it and the patient gets the care they deserve very quickly.
Host: I mean, this is absolutely incredible. So, you're saying that, you know, it doesn't make decisions necessarily for you, but it's kind of like an assistant that says, "Look, we've seen based on all of, for example, the imaging of the brain before that this is probably an LVO or something that really requires your attention. It notifies you and the rest of the people on that patient care team, so there's no time wasted in coordination. And in that way, you're cutting down your response time by an order of magnitude. Is that correct?
Guest: Absolutely. Yes.
Host: That is incredible. I'm curious, like are there any results or statistics of improved outcomes or these faster treatment times that you can share with the audience?
Guest: Oh, absolutely. So since this platform is connecting our care teams earlier, obviously, we're increasing the speed of diagnosis and subsequently we're increasing the care, you know, or getting to the care or the procedures. So by doing this and, again, I've said it earlier, time is brain. You know, you're saving trillions of neurons much faster in the brain. There's evidence that there's 40% improvement in your outcomes and length of stay in the hospital. If patients are being transferred to tertiary center, for example, you know, this happens faster as well. So, no matter from which angle you look at this, there's definitely improvement and care and better care for the patient.
Host: That's incredible. You know, I just want to make sure to focus on that time is brain piece of it because this can't be over-emphasized. When you are going through a stroke, right? Isn't it something like millions of neurons and synapses and cells that are dying. And that's why it's so important to get treatment before, you know, time runs out.
Guest: Absolutely. So, anytime there's concern for a stroke, whether it's the patient or their families, please get care ASAP, whether by calling 911, of course, or getting them to the nearest facility because time is brain and we can't overemphasize this. And once they get to our facility or facilities, you know, physicians and nurses and other providers, they're doing their best to move things very, very fast because we do appreciate the importance of timing here.
Host: Yeah. And now, leveraging this new technology, they're going to have their best chance at recovery because it's identified in a much faster and smarter way, and the care team is notified also in a more efficient way. So, you really have a leg up over other hospitals that don't have or don't leverage this technology. Is that correct?
Guest: Yes, sir. We believe we're lucky and, you know, of course we worked on it diligently to get this technology to Tenet Health Central Coast because we know our patients will benefit from it. And we're very, very thrilled to have this technology.
Host: Amazing. So just before we sign off today, is there anything else that you'd like to share with our audience?
Guest: As I said earlier, we're very thrilled to offer what we're offering at Tenet Central Coast as far as stroke care. We're very happy where we're where we are right now, and we believe we're offering excellent treatment when it comes to stroke. It's a evolving field, and, you know, we're staying at the forefront of technology and care and diagnosis. So, we're very, very happy we're able to offer this to our community.
Host: Yeah. That's incredible. Well, Dr. Yazbeck, thank you so much for your time. I really appreciate it.
Guest: Thank you.
Host: That was Dr. Moussa Yazbeck, Stroke Medical Director for Tenet Health Central Coast. For more information about the future of stroke care, you can visit our website at tenethealthcentralcoast.com. If you found this podcast to be helpful, please share it on your social channels and be sure to subscribe to never miss a new episode.
This has been Healthy Conversations, a podcast from Tenet Health Central Coast. I'm Prakash Chandran. Thank you so much and stay well.
Disclaimer: This program is a community service and is not intended to be a substitute for medical advice. Listeners having questions about their health should make an appointment to see their personal physician. Any opinions or statements made during the program are those of the individuals or physicians making the statements and are not the opinions or statements of the hospital.
Prakash Chandran: A stroke occurs when something blocks blood supply to part of the brain, or when a blood vessel in the brain bursts. Because of the nature of this medical emergency, time is of the essence when it comes to treatment. Luckily, there have been technological advancements that leverage artificial intelligence to accelerate care coordination, reducing systemic delays that stand between patients and lifesaving treatments.
Host: We're going to talk about it today with Dr. Moussa Yazbeck. He's a stroke medical director for Tenet Health Central Coast. This is Healthy Conversations, a podcast from Tenet Health Central Coast. I'm your host, Prakash Chandran. So Dr. Yazbek, thank you so much for joining us today. I really appreciate your time. Before we get into AI and how that's leveraged in treatment, I'd love to know what the standard ways that strokes are both identified and treated today.
Dr. Moussa Yazbeck: Of course. Thank you, Mr. Chandran, for the introduction and for having me. When a patient presents with stroke symptoms such as weakness, speech difficulty numbness, new visual deficits, and many other symptoms which providers can recognize, we first perform an examination and we get a detailed history and timing of the onset is very important. After this, we do a CAT scan of the brain, which is ordered in a very expedited fashion for the obvious reasons. The reason why we do the CAT scan is to differentiate between an ischemic stroke, like you mentioned, from a clot in the blood vessel which are the majority of strokes or probably two-thirds of the stroke or hemorrhagic stroke, which is a bleeding or a blood vessel, which bursted in the brain, which constitutes a third of the stroke. Hemorrhagic strokes have a very different pathway of treatment, and we're going to discuss today mostly the ischemic strokes when you have a clot in the blood vessel.
Guest: Once you establish a, you know, potential or possibility of an ischemic stroke, we follow a very specific approach and a protocol. And as I said earlier, time is of the essence because time is brain. When you perform the we decide whether this patient qualifies for a medicine, a clot-buster so to speak. The medical term would be a thrombolytic, and some of those medications are called Alteplase or commonly known as tPA or TNK. And those medicines help with dissolving the clot, and there's specific criteria for who qualifies for this intervention or for this medicine. And usually, it has to be given within four and a half hours. Once we decide on that part, we go to the next step, which is, does the patient qualify for thrombectomy? A mechanical procedure where you go with the catheter either from the groin area or the wrist and remove the clot mechanically and open the blood vessel. And for this, we do a CT angiogram and a CT perfusion, a CAT scan also with some dye to look at the blood vessel and some fancy images where we look at the brain and which areas of the brain are still salvageable or we can help or which areas are too late to do any intervention on. So, that's the standard approach, a pretty good approach, but that's the standard approach before AI.
Host: Yeah. So, that identification process and figuring out whether a patient meets that criteria is an important piece to this. And I imagine that that whole process takes time, a lot of coordination. So now, let's talk about these new developments and technology that leverage AI for stroke identification and treatment. Talk to us about what we can expect from this new technology.
Guest: Absolutely. So now, we have artificial intelligence, AI, in medicine like many other fields which are using AI. And the idea is to detect those patients even faster than we were doing prior and see who qualifies for an intervention faster, mostly a clot removal or thrombectomy, and act accordingly and get them to the right place faster and basically, you know, this affecting their care.
Host: Yeah. That's amazing. And are you able to tell or talk at a high level around how the technology does this? Is it just that it has seen so many patterns of what, for example, a patient that qualifies for a thrombectomy looks like that it can just tell you faster? How does it work?
Guest: Absolutely. So, I mentioned earlier in the pre-AI era, which is only a few years ago, that we do a CT angiogram and a CT perfusion, basically a CAT scan with dye, and there's some fancy images we look at called CT perfusion, which is great technology, and that's how we triage patients for thrombectomies to see which patients have occlusions in some of the large vessels we can get to.
So, basically, what we did before this, we do those images and then we upload those images to our systems, whether it's Central Coast or any other hospital which may have this technology. And then, the providers, specifically the stroke neurologist, me, the radiologist, the ED doctor, the interventionalist, and whoever else is involved in the care. Once those images are uploaded, they would have to look at those images separately and subsequently communicate with each other, often one-on-one by phone call or so and to come up with a plan. Now, we have artificial intelligence technology at Central Coast, which basically minutes after performing the study, the studies are downloaded to to the cloud and then pushed to all the players, us, the providers I just mentioned. Whether you're using your phone, for example, I use my phone, a tablet, a desktop, whichever device you're using. Not only this, once the study is performed, just minutes after it's uploaded to the cloud, it does prompt me, it sends a beep, prompting me that it was done. So I don't need a phone call or me keep checking. I'm able to look at the images on my phone and it's pretty impressive the quality of those images. I can zoom in, I can zoom out. There's 3D technology where I can turn the images and basically it gives me all the information I need. Not only this, the AI software is able to recognize those clots even before I make this decision with a very, very high sensitivity and specificity. And once it detects this clot, which, in our terms, we call it LVO, large vessel occlusion, it sends me multiple beeps or a sound, whichever you decide telling me that, "Doctor, there is an LVO going on," and that some often happens literally minutes after the patient gets the study done.
Of course, nothing replaces a physician or a specialized physician to look over those images and, you know, putting it in clinical context, which is the most important thing here and agreeing with the software. But you know, it's very infrequent that we don't agree. As a matter of fact, I had one case where I didn't personally appreciate the clot and the software did, and I went back and the software was correct.
So saying this, once this is all done, now we're able to communicate between all the physicians or the provider through this application, of course, it's HIPAA compliant and making decisions very, very quickly because, you know, we're all on the same page. We're communicating those studies and then we act upon it and the patient gets the care they deserve very quickly.
Host: I mean, this is absolutely incredible. So, you're saying that, you know, it doesn't make decisions necessarily for you, but it's kind of like an assistant that says, "Look, we've seen based on all of, for example, the imaging of the brain before that this is probably an LVO or something that really requires your attention. It notifies you and the rest of the people on that patient care team, so there's no time wasted in coordination. And in that way, you're cutting down your response time by an order of magnitude. Is that correct?
Guest: Absolutely. Yes.
Host: That is incredible. I'm curious, like are there any results or statistics of improved outcomes or these faster treatment times that you can share with the audience?
Guest: Oh, absolutely. So since this platform is connecting our care teams earlier, obviously, we're increasing the speed of diagnosis and subsequently we're increasing the care, you know, or getting to the care or the procedures. So by doing this and, again, I've said it earlier, time is brain. You know, you're saving trillions of neurons much faster in the brain. There's evidence that there's 40% improvement in your outcomes and length of stay in the hospital. If patients are being transferred to tertiary center, for example, you know, this happens faster as well. So, no matter from which angle you look at this, there's definitely improvement and care and better care for the patient.
Host: That's incredible. You know, I just want to make sure to focus on that time is brain piece of it because this can't be over-emphasized. When you are going through a stroke, right? Isn't it something like millions of neurons and synapses and cells that are dying. And that's why it's so important to get treatment before, you know, time runs out.
Guest: Absolutely. So, anytime there's concern for a stroke, whether it's the patient or their families, please get care ASAP, whether by calling 911, of course, or getting them to the nearest facility because time is brain and we can't overemphasize this. And once they get to our facility or facilities, you know, physicians and nurses and other providers, they're doing their best to move things very, very fast because we do appreciate the importance of timing here.
Host: Yeah. And now, leveraging this new technology, they're going to have their best chance at recovery because it's identified in a much faster and smarter way, and the care team is notified also in a more efficient way. So, you really have a leg up over other hospitals that don't have or don't leverage this technology. Is that correct?
Guest: Yes, sir. We believe we're lucky and, you know, of course we worked on it diligently to get this technology to Tenet Health Central Coast because we know our patients will benefit from it. And we're very, very thrilled to have this technology.
Host: Amazing. So just before we sign off today, is there anything else that you'd like to share with our audience?
Guest: As I said earlier, we're very thrilled to offer what we're offering at Tenet Central Coast as far as stroke care. We're very happy where we're where we are right now, and we believe we're offering excellent treatment when it comes to stroke. It's a evolving field, and, you know, we're staying at the forefront of technology and care and diagnosis. So, we're very, very happy we're able to offer this to our community.
Host: Yeah. That's incredible. Well, Dr. Yazbeck, thank you so much for your time. I really appreciate it.
Guest: Thank you.
Host: That was Dr. Moussa Yazbeck, Stroke Medical Director for Tenet Health Central Coast. For more information about the future of stroke care, you can visit our website at tenethealthcentralcoast.com. If you found this podcast to be helpful, please share it on your social channels and be sure to subscribe to never miss a new episode.
This has been Healthy Conversations, a podcast from Tenet Health Central Coast. I'm Prakash Chandran. Thank you so much and stay well.
Disclaimer: This program is a community service and is not intended to be a substitute for medical advice. Listeners having questions about their health should make an appointment to see their personal physician. Any opinions or statements made during the program are those of the individuals or physicians making the statements and are not the opinions or statements of the hospital.