05.21.24-Leading-Edge-Webinar-Digital-Health-and-Wellness-Beacon-Biosignals

Conference Video|Duration: 10:40
May 21, 2024
  • Video details
    Development of an EEG Neurobiomarker Platform for Neurological and Psychiatric Disease 

     

  • Interactive transcript
    Share

    SPEAKER 1: Now the next part of our webinar today, which are to start up with our Startup Exchange Program. I will hand over to Rebekah.

    REBEKAH MILLER: Thank you, Mickey. And just a note, you will have the chance to come back to those questions that we didn't have the time to answer and the upcoming panel following our startup presentations, and so now we'll move forward, as Mickey said, with introducing two fantastic startups from our Startup Exchange Program at MIT.

    I'm pleased to introduce Dr. Jacob Donoghue, CEO and co-founder of Beacon Biosignals. Dr. Donahue earned an MD from Harvard Medical School and a PhD in neuroscience from MIT developing machine learning methods for quantifying pharmacological effects on neural activity. Dr. Donahue joins us today to present on the development of an EEG neuro biomarker platform for neurological and psychiatric disease. The floor is all yours, Dr. Donahue.

    JACOB DONOGHUE: Great. Thank you so much for having me here today. It's fun to be speaking to the group as a ex-MIT or as an MIT alum who trained over in building 46 not too long ago in the neuroscience department. So, yeah, I'm excited to tell you more about what we're building here at Beacon with the next generation EEG platform for understanding neurologic and psychiatric disease.

    Our mission really is harnessing AI to focus on a platform that can transform the development of new therapies for neurologic and psychiatric disorders.

    So myself, I'm a physician scientist trained at Harvard and MIT alongside Jarrett Revels, who is at Csail, working on Julia language at MIT as well as our clinical co-founders Brendan Westover and Sydney Cash, who are based out of MGH and Beth Israel.

    So the challenge really is that diseases of the brain and mind are devastating, uniquely hard to predict, diagnose, and treat.

    To address this challenge at Beacon, we really brought together a platform to take on this opportunity to really make an impact for patients. So the Beacon platform ties together a world-class database of raw EEG and polysomnogram data for brain function-- functional recordings tied together with machine learning algorithms, many of which are FDA 510 cleared software as a medical device, integrate this via APIs through multimodal other data streams, and bring it together for our biopharma partners through web-based dashboards and analytics to be able to interrogate the heterogeneity of neurologic and psychiatric disease and measure how those changes transform with novel treatments, and as I'll talk a bit about today briefly, all of these advances are really only possible with some of the next generation medical grade wearables like the Dreem 3S headband that we've developed.

    So at our core, we believe that neurophysiology and brain activity provides this critical real time link between the molecular and cellular basis of disease and the clinical outcomes that matter to patients and often we rely on for endpoints in developing new therapies. So brain activity really forms this bridge of how patients feel, function, and survive.

    Until now, EEG, the sum of millions of neurons firing action potentials and field activity, have really when getting recordings of this of this activity in here you see an example on the right. Every squiggly line traced the electrode placed on the scalp, in this case of a child with an epilepsy, every sharp wave very likely an epileptiform discharge, which we heard about the impact just a moment ago. And so really in here in just a minute of data, you can already see the counting events of interest can be overwhelming. You can imagine as we scale up the ability of wearables to get clinical grade data, it becomes intractable for humans to be able to interrogate meaningfully and objectively and repeatedly the types of biomarkers that can be critical to understanding disease and treatment. And so this human interpretation leads to data emission subjectivity, variability, and, as I mentioned, a lack of reproducibility, and this is where machine learning tools really unlock the ability to interrogate these data sets.

    And today's neuro-diagnostic tools are inaccessible. There's often a long wait time to even getting brain monitoring and don't scale to leverage machine learning enabled biomarker discovery. Here on the right, you see a patient getting a traditional polysomnogram as a part of a sleep study. You can notice a few of the electrodes on the head and record EEG and think everyone pretty much agrees that it's hard to imagine that this-- the sleep data that you're acquire from this patient really represents what's actually going on in many especially neurologic and psychiatric disorders.

    And so at Dreem, we've built the Dreem-- at Beacon, we've acquired and taken forward the Dreem 3S scalable sleep EEG medical device built and tailored for clinical trials.

    And so our EEG headband is really optimized for patient comfort, really important to us, patients with high disease burden that we can be utilized to get clinical grade brain data at home over multiple nights. So can this be set up without a technologist or even by yourself or by a caregiver and enables clinical grade sleep staging at the most granular level equivalent to in-lab polysomnography. And that is our FDA clearance.

    So this device-- and I have one in my hand. Folks can see me. The hours that it takes to set up a traditional sleep lab, all of a sudden, in just a few seconds can be set up and then paired to a smartphone to enable clinical grade brain monitoring from home. But you can see here that we have electrode locations in many of the traditional 1020, the system that traditional eegs measured locations from the scalp-- the nasion is the nose-- so that you can be able to better interrogate and transform traditional brain monitoring data to data that we're capturing from our device.

    And what's so powerful is that by bringing this device forward through this medical device pathway was the ability to see that these dry EEG electrodes while wearable and comfortable also acquire a gold standard level EEG. And so here on the right, you can see both the Dreem alongside a PSG, polysomnogram EEG like we saw, being co-applied at the same time. And so these applications really get unlocked now that we have the ability to reach patients that traditionally you can't send to a sleep lab.

    In addition to this, we've built software as a medical device for automated sleep staging so that these algorithms can interrogate the same level of granularity within a polysomnogram but at home and at scale. And actually, Beacon just received the first predetermined change control plan from the FDA, meaning we can start to tailor and improve our algorithms without submitting new 510Ks so continually improving our ability to get better generalization for patients.

    And so these processes really unlock the labor intensive manual process, and here's 1,000-- couple of 10,000 polysomnograms that we were able to stage in just a few hours that might otherwise take human months, removing that subjectivity and getting really fast turnaround when months for a new treatment getting to patients really matters.

    And so the types of things we now can get is not just sleep and wake times but really the whole granular structure as I mentioned of sleep stages but at home and over multiple nights, so you can get already validated and accepted endpoints that are utilized in clinical development for new therapies looking at features like sleep efficiency and interrogating REM sleep in particular, which is important to characterizing diseases like narcolepsy and depression and can change with interventions like SSRIs. As well as because we're getting raw EEG data, we can really look at interesting features linked to cognition like sleep spindles and power spectral. The amplitude of slow waves might be linked to neurodegenerative disease and the clearance of tau and amyloid from the brain.

    And this is all tied together through a platform that enables sleep lab quality at home where we can see the quality of the EEG that's coming in from each recording that the recordings and the participants are utilizing the devices on the head. And then, of course, all these analytics are available for our for our partners, both in biopharma but as well as academia.

    And so really this scalable sleep EEG at home and at scale accelerates the development of precision medicines by being able to look at the heterogeneity of these otherwise complex syndromes and understand why certain patients might respond to certain therapies being better quantitative and objective manner of understanding how a treatment might affect brain activity. And what's critical is sleep is important to patients, to providers, to payers, and to regulators and offers a really important way that now we can bring to patients at scale.

    And to close, what's really we think important about sleep is it provides this amazing surface area for understanding biomarkers about brain health. It's not just about sleep for sleep sake. That sleep, of course, is critical in disorders like narcolepsy and insomnia where the primary disease of sleep. We know that sleep is important and a core diagnostic criteria in diseases like bipolar disorder, PTSD, and depression and just as importantly in neurodegenerative disease. The way that sleep changes, especially in non-REM 3 and deeper sleep is linked to neurodegenerative disease processes, and, for example, in Parkinson's disease, the onset of rem behavior disorder so movement during REM sleep when you're supposed to be atonic without muscle contractions can often predict symptoms five to seven years in advance. And so now with tools like this, we're able to start screening and think about earlier interventions for the next generation of therapies.

    And lastly, of course, sleep is important outside of CNS disorders and looking at in pain and rheumatoid arthritis, ulcerative colitis, and Crohn's disease, diseases where the primary disorder is disturbing your nighttime sleep and affects the quality of life where you really just would have a challenge of sending these patients to a sleep lab. But now we get to unlock those lab quality endpoints at home for our biopharma partners.

    So with that, thank you so much. Very excited to tell you more about what we're doing at-- doing at Beacon.

  • Video details
    Development of an EEG Neurobiomarker Platform for Neurological and Psychiatric Disease 

     

  • Interactive transcript
    Share

    SPEAKER 1: Now the next part of our webinar today, which are to start up with our Startup Exchange Program. I will hand over to Rebekah.

    REBEKAH MILLER: Thank you, Mickey. And just a note, you will have the chance to come back to those questions that we didn't have the time to answer and the upcoming panel following our startup presentations, and so now we'll move forward, as Mickey said, with introducing two fantastic startups from our Startup Exchange Program at MIT.

    I'm pleased to introduce Dr. Jacob Donoghue, CEO and co-founder of Beacon Biosignals. Dr. Donahue earned an MD from Harvard Medical School and a PhD in neuroscience from MIT developing machine learning methods for quantifying pharmacological effects on neural activity. Dr. Donahue joins us today to present on the development of an EEG neuro biomarker platform for neurological and psychiatric disease. The floor is all yours, Dr. Donahue.

    JACOB DONOGHUE: Great. Thank you so much for having me here today. It's fun to be speaking to the group as a ex-MIT or as an MIT alum who trained over in building 46 not too long ago in the neuroscience department. So, yeah, I'm excited to tell you more about what we're building here at Beacon with the next generation EEG platform for understanding neurologic and psychiatric disease.

    Our mission really is harnessing AI to focus on a platform that can transform the development of new therapies for neurologic and psychiatric disorders.

    So myself, I'm a physician scientist trained at Harvard and MIT alongside Jarrett Revels, who is at Csail, working on Julia language at MIT as well as our clinical co-founders Brendan Westover and Sydney Cash, who are based out of MGH and Beth Israel.

    So the challenge really is that diseases of the brain and mind are devastating, uniquely hard to predict, diagnose, and treat.

    To address this challenge at Beacon, we really brought together a platform to take on this opportunity to really make an impact for patients. So the Beacon platform ties together a world-class database of raw EEG and polysomnogram data for brain function-- functional recordings tied together with machine learning algorithms, many of which are FDA 510 cleared software as a medical device, integrate this via APIs through multimodal other data streams, and bring it together for our biopharma partners through web-based dashboards and analytics to be able to interrogate the heterogeneity of neurologic and psychiatric disease and measure how those changes transform with novel treatments, and as I'll talk a bit about today briefly, all of these advances are really only possible with some of the next generation medical grade wearables like the Dreem 3S headband that we've developed.

    So at our core, we believe that neurophysiology and brain activity provides this critical real time link between the molecular and cellular basis of disease and the clinical outcomes that matter to patients and often we rely on for endpoints in developing new therapies. So brain activity really forms this bridge of how patients feel, function, and survive.

    Until now, EEG, the sum of millions of neurons firing action potentials and field activity, have really when getting recordings of this of this activity in here you see an example on the right. Every squiggly line traced the electrode placed on the scalp, in this case of a child with an epilepsy, every sharp wave very likely an epileptiform discharge, which we heard about the impact just a moment ago. And so really in here in just a minute of data, you can already see the counting events of interest can be overwhelming. You can imagine as we scale up the ability of wearables to get clinical grade data, it becomes intractable for humans to be able to interrogate meaningfully and objectively and repeatedly the types of biomarkers that can be critical to understanding disease and treatment. And so this human interpretation leads to data emission subjectivity, variability, and, as I mentioned, a lack of reproducibility, and this is where machine learning tools really unlock the ability to interrogate these data sets.

    And today's neuro-diagnostic tools are inaccessible. There's often a long wait time to even getting brain monitoring and don't scale to leverage machine learning enabled biomarker discovery. Here on the right, you see a patient getting a traditional polysomnogram as a part of a sleep study. You can notice a few of the electrodes on the head and record EEG and think everyone pretty much agrees that it's hard to imagine that this-- the sleep data that you're acquire from this patient really represents what's actually going on in many especially neurologic and psychiatric disorders.

    And so at Dreem, we've built the Dreem-- at Beacon, we've acquired and taken forward the Dreem 3S scalable sleep EEG medical device built and tailored for clinical trials.

    And so our EEG headband is really optimized for patient comfort, really important to us, patients with high disease burden that we can be utilized to get clinical grade brain data at home over multiple nights. So can this be set up without a technologist or even by yourself or by a caregiver and enables clinical grade sleep staging at the most granular level equivalent to in-lab polysomnography. And that is our FDA clearance.

    So this device-- and I have one in my hand. Folks can see me. The hours that it takes to set up a traditional sleep lab, all of a sudden, in just a few seconds can be set up and then paired to a smartphone to enable clinical grade brain monitoring from home. But you can see here that we have electrode locations in many of the traditional 1020, the system that traditional eegs measured locations from the scalp-- the nasion is the nose-- so that you can be able to better interrogate and transform traditional brain monitoring data to data that we're capturing from our device.

    And what's so powerful is that by bringing this device forward through this medical device pathway was the ability to see that these dry EEG electrodes while wearable and comfortable also acquire a gold standard level EEG. And so here on the right, you can see both the Dreem alongside a PSG, polysomnogram EEG like we saw, being co-applied at the same time. And so these applications really get unlocked now that we have the ability to reach patients that traditionally you can't send to a sleep lab.

    In addition to this, we've built software as a medical device for automated sleep staging so that these algorithms can interrogate the same level of granularity within a polysomnogram but at home and at scale. And actually, Beacon just received the first predetermined change control plan from the FDA, meaning we can start to tailor and improve our algorithms without submitting new 510Ks so continually improving our ability to get better generalization for patients.

    And so these processes really unlock the labor intensive manual process, and here's 1,000-- couple of 10,000 polysomnograms that we were able to stage in just a few hours that might otherwise take human months, removing that subjectivity and getting really fast turnaround when months for a new treatment getting to patients really matters.

    And so the types of things we now can get is not just sleep and wake times but really the whole granular structure as I mentioned of sleep stages but at home and over multiple nights, so you can get already validated and accepted endpoints that are utilized in clinical development for new therapies looking at features like sleep efficiency and interrogating REM sleep in particular, which is important to characterizing diseases like narcolepsy and depression and can change with interventions like SSRIs. As well as because we're getting raw EEG data, we can really look at interesting features linked to cognition like sleep spindles and power spectral. The amplitude of slow waves might be linked to neurodegenerative disease and the clearance of tau and amyloid from the brain.

    And this is all tied together through a platform that enables sleep lab quality at home where we can see the quality of the EEG that's coming in from each recording that the recordings and the participants are utilizing the devices on the head. And then, of course, all these analytics are available for our for our partners, both in biopharma but as well as academia.

    And so really this scalable sleep EEG at home and at scale accelerates the development of precision medicines by being able to look at the heterogeneity of these otherwise complex syndromes and understand why certain patients might respond to certain therapies being better quantitative and objective manner of understanding how a treatment might affect brain activity. And what's critical is sleep is important to patients, to providers, to payers, and to regulators and offers a really important way that now we can bring to patients at scale.

    And to close, what's really we think important about sleep is it provides this amazing surface area for understanding biomarkers about brain health. It's not just about sleep for sleep sake. That sleep, of course, is critical in disorders like narcolepsy and insomnia where the primary disease of sleep. We know that sleep is important and a core diagnostic criteria in diseases like bipolar disorder, PTSD, and depression and just as importantly in neurodegenerative disease. The way that sleep changes, especially in non-REM 3 and deeper sleep is linked to neurodegenerative disease processes, and, for example, in Parkinson's disease, the onset of rem behavior disorder so movement during REM sleep when you're supposed to be atonic without muscle contractions can often predict symptoms five to seven years in advance. And so now with tools like this, we're able to start screening and think about earlier interventions for the next generation of therapies.

    And lastly, of course, sleep is important outside of CNS disorders and looking at in pain and rheumatoid arthritis, ulcerative colitis, and Crohn's disease, diseases where the primary disorder is disturbing your nighttime sleep and affects the quality of life where you really just would have a challenge of sending these patients to a sleep lab. But now we get to unlock those lab quality endpoints at home for our biopharma partners.

    So with that, thank you so much. Very excited to tell you more about what we're doing at-- doing at Beacon.

    Download Transcript