Watch the full interview with Brad Hirsch.

SignalPath CEO Brad Hirsch on Doing Less Busywork and More Science

Join Machine Meets World, Infinia ML’s ongoing conversation about AI

James Kotecki
Jul 7 · 25 min read

Episode Highlights

This week’s guest is SignalPath CEO Brad Hirsch.

“…what we initially did at SignalPath was to go through and actually hire an enormous team whose job is to literally go through a PDF, and then sit there with a computer and put it in. And so a lot of our focus on automation is how do we aggregate the information from the PDF through work that we’ve done with machine learning with Infinia and others.”

“…you look at the clinical trials that need to happen — and COVID is a great example, where we need to be doing things at incredible scale that just aren’t happening because of the complexity of execution — there are a lot of ways to make people very comfortable with the idea that streamlining workflows allows you to open more trials, get more patients on trial, answer questions quicker. So there’s so much of a need that the risk to the job is still so low.”

Watch the show above. You can also hear Machine Meets World as a podcast, join the email list, and contact the show.

Audio + Transcript

James Kotecki:

Hey, we are live from Infinia ML. This is Machine Meets World. I am James Kotecki, talking artificial intelligence today with my guest, the co-founder and CEO of SignalPath, Brad Hirsch. Welcome, Brad.

Brad Hirsch:

Thanks. Thanks for having me. I’m excited to catch up today.

James Kotecki:

Brad, there’s a lot to talk about with AI, but first I think just for context for people, we should say you are a doctor. You are in Texas. How are things going for you right now? What are you, are you doing any doctor stuff right now? And how’s that going?

Brad Hirsch:

Yeah, so I’m a cancer doc. I’m a medical oncologist. Thankfully SignalPath has grown so much and gotten so busy that I don’t work day to day, but I’m still supporting one of the hospitals here in town. And unfortunately it’s getting, it is getting pretty crazy. We’re seeing more and more exposures in quarantine among docs. We’re seeing the hospitals start to fill up. So I, like everybody else, am locked down as much as possible with my family and [inaudible] go out.

James Kotecki:

And would you, as a cancer doc, be kind of pulled into COVID cases or is it just kind of a downstream effect of how it impacts your day to day work?

Brad Hirsch:

Yeah, so we’re not taking care of them directly. I don’t think anybody, I haven’t intubated a patient in 20 years, 10 plus years. So I don’t think anybody wants to me in an ER or an ICU. But at the same time, COVID has so many sort of unique characteristics, of when people get really sick, of the immune response, so there are a lot of oncologic agents being used and tested in their relation to immune system. We’re seeing blood clots. So we’re called in, but we’re not generally having to be bedside. So most of it is remote and ending in a lot of telehealth and other mechanisms.

James Kotecki:

Okay. Well thank you for being a part of the healthcare system. And thanks for being a part of our show today. I know that you’ve got a lot going on, given those two roles that you’re playing. We’re here to talk about AI and how that plays out in SignalPath. So first, can you just describe SignalPath, assuming someone’s outside the industry? Back when we used to go to parties, let’s say you met someone at a party, how would you describe SignalPath?

Brad Hirsch:

Well, I often start by talking, our goal is to build workflows and sort of reimagine the execution of trials. First at sites, and ultimately with pharma and those conducting the trials themselves. And often I start by describing the incredibly antiquated nature of clinical trials today. I mean, people don’t understand, they think clinical research, they think it’s sort of very advanced and next generation stuff going on. But in reality, if I’m a clinical research site, which I work at one, as a doc, we get sent 400 page PDFs of how to conduct a trial by FedEx coming in the mail. We have disparate systems that are different for every trial. It’s full of manual processes. They send out coordinators or monitors, excuse me, from the companies running the trials to figure out how they’re being executed. So it’s this incredibly antiquated approach.

Brad Hirsch:

And so what we first did at SignalPath, over the last five or six years, is to build workflow tools that really digitize that trial. So it takes the PDF and allows you to understand as a financial person on the trial, as a coordinator on the trial who meets with the patients, these are the things I need to do. This is what’s happening today and tomorrow. I need to get this lab and once that’s completed, this [inaudible] person needs to do something. And so it’s really about the automation and digitization of those workflows.

Brad Hirsch:

And for us, that’s sort of SignalPath 1.0. SignalPath 2.0 is, then how do we now use a national network of sites on our platform that are operationalizing these trials to really reimagine the trials themselves. And so it’s been a, it’s been a long process, but a lot of fun.

James Kotecki:

This is interesting because what you’re talking about with medical trials being really antiquated and then starting to digitize, and now kind of getting into the realm of AI, which we’re going to talk about. It seems like that kind of traces business digitization overall over the last couple of decades. And maybe medical trials, from what you’re saying, may be a little bit behind the curve compared to some other industries. But it’s that idea of first you start with just sheets of paper and a completely disorganized system. Then at least you start to try and digitize and standardize some of that data. And then you start to use AI to kind of go at the data in interesting ways and see what you can automate. Is that kind of been your experience?

Brad Hirsch:

Yeah. And on the electronic health records side, in medical clinics, that happened based on legislation back 15, 20 years ago, where everybody started having electronic health records and automating binders that were in the clinics and that data got generated. And then that went to a whole bunch of interesting things that are being done today that I did as a part of a prior company called Flatiron, and others who do it. Clinical trials are behind that, first bit because legislation wasn’t passed to impact them. But second, because everybody is so risk averse in the space. The FDA watched it so closely and no drug wants to put their — no company wants to put their new agent, their new drug at risk. And so it’s, I would argue even further behind others. And additionally it’s, even before the data approach of how do you aggregate the data and do the analysis differently, it’s even the operational piece, which is where we focus first. Of how do you just move the operational piece outside of PDFs and Excel files and sticky notes.

James Kotecki:

So you’re definitely in the weeds of this kind of digital transformation. I don’t know if every industry is still calling it that, but you guys are definitely in the weeds of that. And so as a CEO of a company going through this process, just give me the high level. How do you think about automation and from, at a strategic kind of C-Suite level and how that plays into the work you’re trying to do?

Brad Hirsch:

There’s so many levels of opportunity for automation in clinical trials and clinical research. The first is on the operational side. How do we move away from those PDFs and Excel files and calendars, and how do we put those into a digital format? And you can automate that by, what we initially did at SignalPath was to go through and actually hire an enormous team whose job is to literally go through a PDF, and then sit there with a computer and put it in. And so a lot of our focus on automation is how do we aggregate the information from the PDF through work that we’ve done with machine learning with Infinia and others? Are you losing me again?

James Kotecki:

No, no, no. I’m just saying it’s funny, actually I was responding to what you were saying. Because it’s funny that when you talk about this great digital revolution, it’s still so people intensive, at least at the beginning. Because it’s like, okay, it’s digitizing things, but it’s people sitting there typing in the information.

Brad Hirsch:

Well, and it’s the same thing. So Flatiron Health is a big company in the oncology space and the data space, that I worked with a couple of years back before coming to SignalPath. And we literally had a thousand tumor registrars, who were people that sit in the basements of hospitals, looking at the medical record and then putting things in another chart. And what we did there is we applied machine learning, it was Google and Palantir and folks that came over. And we literally just apply, we made the ability to say, “Is this a stage four patient?” Instead of searching through the electronic health record, we made it easy. But the last mile is still human because of the risk aversion, the ethical complexity of not getting it right.

Brad Hirsch:

And so we’ve applied that now at SignalPath. Of how do we make it a much more streamlined process through automation and through ML, but still at the end of the day, it’s a person saying, “Yes, that is correct.” And really QAing at the last mile, at least for now. Hopefully we’ll get past even that, but I don’t think the healthcare spaces is ready for that quite yet.

James Kotecki:

And you mentioned Infinia ML is working with SignalPath. So I should probably should just say that for disclaimer purposes or whatever, but we enjoy the work we’re doing with you guys a lot. And I’m interested in these high- level conversations because when we talk about the impact that automation is having on humans, you’re saying that there still needs to be human involved at the last mile, especially for medical and ethical and compliance reasons. And also just to make sure that this stuff is actually working, but are you able to use fewer people? Are you able to reassign people to other tasks? How do you think about the human side of the equation in your case?

Brad Hirsch:

Yeah, so for us, it is a good problem to have, but we have such scale pressure right now, to really scale out the organization. We went from, we needed to digitize, bring a hundred trials in the system, to a thousand trials in the systems, to thousands of trials in the system. And so there’s never been a moment in our experience where there was a risk of anybody losing a job or re-imagining. It really is how do we drive it from a 50 hour process to a 20 hour process to a five hour process. So that it really is around QA and it really is about making sure that we can do this right and at scale versus risk to jobs of which they’re have been none.

James Kotecki:

I sense that there, maybe not in your company, but I sense there might be someone listening to this who says, “Yeah, right. A CEO is always going to say that. And eventually there will be some kind of impact to people.” Do you have a communication challenge or maybe a communication strategy of how you talk to the frontline workers whose jobs are not maybe being disrupted, but are certainly changing based on this new technology? Are people reticent to use it? Are people embracing it? How do you deal with the frontline folks?

Brad Hirsch:

Well, it depends on who you define as frontline folks. So first are the people… Digitization, this idea of taking a trial and putting it in the system is not like a, that’s a very SignalPath specific thing. Or the role of the tumor registrar at Flatiron.

James Kotecki:

You said tumor, you said tumor registrar is the name of the title.

Brad Hirsch:

That’s, folks that, there, it is a very obscure title, but it is something, it is a government mandated thing that large hospital health systems have to have an oncology registry. And the person that manually enters that data in a hospital system is called a tumor registrar.

James Kotecki:

Okay.

Brad Hirsch:

For us a digitizer is just our way of saying the person who’s digitizing the trial on our behalf. And so for folks on our team, it hasn’t been an issue because, again, they’ve seen us continue to hire through it. And so it was really about explaining how it makes their lives easier. So you don’t have to go in and enter 50 activities for a trial. We can bring those in and then you can make sure they’re right. And so that hasn’t been an issue.

Brad Hirsch:

The frontline workers, right? The folks in a hospital who are doing the research, they’re often called coordinators, who sit with the doc and the patient and make sure the trial is getting executed. They have not seen as much automation in the workflow today. But again, when you look at what’s possible, you look at the clinical trials that need to happen — and COVID is a great example, where we need to be doing things at incredible scale that just aren’t happening because of the complexity of execution — there are a lot of ways to make people very comfortable with the idea that streamlining workflows allows you to open more trials, get more patients on trial, answer questions quicker. So there’s so much of a need that the risk to the job is still so low.

James Kotecki:

And as a sidebar, by the way, do you see a, what do you see as far as a forecast for demand of medical trials? Let’s even set aside COVID-19. Are medical trials just growing? As I assume that the medical industry just continues to grow and pharmaceutical industry continues to grow in complexity?

Brad Hirsch:

There will never be a lack of clinical trials out there, right? Every pharma company has their pipeline and their pipeline is dependent on the execution of those trials. So we’ll continue to grow. What we have seen recently is that the pressures of COVID-19 has meant that many have stopped the new trials that they were getting ready to launch. Many of them are not allowed to do accruals onto the trials that are open, new patients to go on to those trials. Because of the risk of that patient getting COVID-19 and having a bad outcome, or not going to get procedure because the hospital can’t take care of them. And so we have seen, and unfortunately that’s led to furloughs and complexities.

Brad Hirsch:

But there’s no question that that’s going to take off and that medical… And so not only our medical trials, in a historical perspective, are going to continue to grow in how they were previously done. But there’s this entire new opportunity around, how do we use real world evidence from electronic health records? How do we reimagine the execution of trials? Some of the early COVID data is a great example of, they randomized 11,000 patients to the recovery trial in the UK within a couple of months, to what’s called a basket trial with multiple arms and roles of machine learning and have answered — so that’s where the insight that steroids helped and can decrease mortality by a third in COVID have come from, sort of novel trials and the way to do these things differently.

James Kotecki:

So is the speed and urgency being driven by COVID-19 going to then kind of spill out into things that other trials can learn and adopt in the future? Is it actually driving the overall industry forward?

Brad Hirsch:

No question. So there are things you hear about all the time, like telehealth, and things like that, that have really just taken off. And there’s a lot of question of the sustainability of that. Are doc’s going to embrace it when they don’t have to? Is it going to be paid for? Things like that. But in the trial side, there’ve been so many changes. So examples would be, before it was impossible to get a trial through regulatory and to a site and everything in a couple of days. Now it’s happening all the time, early on in COVID. People have been embracing remote tooling to be able to keep patients out of sites, because either the site’s closed or patients are afraid to go in.

Brad Hirsch:

And so there is a reimagination, I always, just, for some context, there was a big push over the last couple of years for what are called virtual trials. This idea that a trial can come to you in the house and that you don’t actually need to go to a clinic. Of course, in oncology, you’re not going to try a new cancer drug in the house. You’re not going to, a cardiac device isn’t going to be deployed in your house. And seen an interesting mixture and mold of those two, of traditional trial at a site with a doc and a coordinator, and this fully virtual piece that’s been driven by COVID-19.

James Kotecki:

All right. So let’s get back to the idea of the people working with increasing augmentation and automation technologies. So you’ve got people, as you mentioned before, kind of buying in philosophically at least, on the concept that this is a really good thing for their job and for this company. What do you need to do then tactically, as the CEO, to make sure that this stuff actually works? We talked to a Forrester analyst last week, J.P. Gownder. And he said, a lot of the problems with AI adoption is not the technology. It’s the human side of the equation. It’s making sure that people can actually use this stuff, and it’s done in a way that people can actually adapt and build into their workflows. So how do you think about that, not strategically, but now tactically getting people able to use these systems in a way that actually is making them more efficient.

Brad Hirsch:

Yeah. And the positive in our world, in clinical research, is that there are truths that you can compare it to. So it’s not a black box where you’re saying yes or no. “I have some algorithm and I’m supposed to do something based on that algorithm.” It is tangible objective output.

Brad Hirsch:

And so for us an example would be, you have the protocol. You have the PDF as a coordinator that you’ve used traditionally, to know what you need to do today. We’ve now used ML to be able to take out the key pieces, what are the instructions? What are the activities and visits? What are the relationships? To populate our system. But at the end of the day, you have our system on one screen and you have your PDF next to it. And you can say, “Is it objectively correct?” And where it’s not, what do I need to do to correct for it?

Brad Hirsch:

So when we get new sort of iterations on the model, we can also, as a company, go through and look at utility, and look at the where to deploy it, how to deploy, and how to impact that workflow, because it is such an objective outcome on the operational side. Similarly, on the data side, as you start to say, “How do I optimize the design of my trials?” “How do I use historical data, understand my trial portfolio, and the elements of that trial, and the execution of that trial, and start to get insights into how to better optimize it?” Again, it’s tangible objective outcome that’s not a black box. And also while we might not know every consideration on the ML side, we know the implications on the output that make it much easier to understand how to trust it and how to implement it.

James Kotecki:

Mm-hmm (affirmative). And so let’s talk about what is the line right now, or that you see, and what is the near term forecast for the amount of automation that you’re going to be able to do in your business? So right now we’re talking about things like data, you might call it data extraction, knowledge extraction. Having a machine that reads documents, and then finds information, pulls that out, and populates that into another system. And that’s plugging along. But then what other kinds of automations are you maybe looking forward to in the near term that you don’t have yet, but are theoretically possible?

Brad Hirsch:

Yeah. So next big one that I think is an exciting place for us to focus on is on the trial side of it. So obviously we’ve been focused on the operational side of how do we get rid of antiquated systems? How do we optimize it? But on the trial side, there are sort of innumerable ways that we can help to make people’s lives easier. So a couple of examples.

Brad Hirsch:

When a clinical trial is done at a site today, there’s, you come in and you say, “I need to do these ten lab tests.” I have to go into the EHR, the electronic health record. I have to enter those. I then have to wait for the results to come back. I have to take those results and I have to put them into a case report form that’s another digital form. So there are all these ways to say, how can I look at what’s in the trial? How can I automate, put some automation in place to understand how to order these things at sites and then make sure that they’re right? And as the results come back, how can I take those out of the electronic health record? How can I normalize them?

Brad Hirsch:

Because one lab test, one albumin, one whatever can have 50 different ways that it’s captured in one electronic health record. So how can I take those out? How can I normalize them, and then put them in the CRF, in the report forms that the trialists use. And so, again, these are objective things where there’s incremental learning we can do across the board to get it right. But the end result is a very tangible where the coordinator is sitting there saying, “Yes, that was the right lab that was ordered. Yes, that the result is available. And yes, the result that ended up in the form that goes back to pharma is also accurate.”

Brad Hirsch:

And so those are much more complex problems of how do you understand lab tests? How do you understand normalization? It’s something, it’s things we’ve done at groups like Flatiron and elsewhere. So it’s not outside the realm of what’s possible in the near term. But that’s sort of the next level of complexity that we’re going to be dealing with at SignalPath and elsewhere.

James Kotecki:

So it’s sometimes dangerous to personify AI, but if you could personify it, it’s basically, in some ways you’re talking about an assistant, right? That’s doing all this kind of a digitization, then giving it to a human to check over it. And it’s as if you had another person there, or you cloned yourself or whatever, and now you just have the opportunity to check it over instead of having to do the grunt work of putting it into the system.

Brad Hirsch:

Correct. There’s still a lot of mapping involved. There’s still a lot of manual work. There’s still a lot of review. But the question is, “How do we have more trials? How do we have better trials? And how do we get people away from the busywork that doesn’t need to occur that has no real utility other than making sure that it’s correct?” And so that the patient, the coordinator can be with the patient to make sure that they’re getting what they need done. They can be opening more trials. They can be doing the things that they’re really qualified to do, that advances the science, advances the case of patients.

Brad Hirsch:

A great example, oncology patients are in incredible need of being a part of research. I never met a cancer patient that doesn’t want to get the best possible care on the best trial. And yet less than 3%, depending on the number, either 3 or 5% of patients ultimately get on a trial. So how do we make the trials better, make them more efficient, and get that 3% to 50% so that people are really a part of advancing the science in the way that they can?

James Kotecki:

Oh, we’re getting a live question now from Katherine James on the LinkedIn live stream, who says, “With wearables offering a more continuous means of collecting some bio data, are they being included in medical trials yet?”

Brad Hirsch:

So there are a lot of organizations that are spending a lot of time thinking about that. Again, we come back to the idea that this is a antiquated, sort of services-oriented, approach to clinical trials today. And people have a hard time figuring out how to link all these things together. So one of the big things, one of the big things that we’re pushing at SignalPath is, so now we’ve digitized the trial, right? So you know what visits are happening, what the relationships are. If they develop liver issues, you should do this. And that infrastructure is now allowing us to tie into patient surveys. So there’s direct patient engagement. It allows us to tie into sensor data and be the infrastructure, EHR, data pulls from the electronic health record, and be able to put together the data set to answer those questions.

Brad Hirsch:

There are groups like Elekta Labs. There are groups like DiMe, the digital medicine association, that are really pushing that agenda and it’s gaining steam. But it’s not yet at the point where that pivotal trial, that the pharma client sees the utility, that they understand what to do with it. So it’s at the fringes, it’s more of the periphery. But we, like everybody else, would love it to become increasingly essential to everyday. So you need to understand the patient experience. Not only did they live or die, or did we extend their life or their time to progression, with cancer as an example, but also did we make their quality of life better? Did we make them more active? Were they able to go to the graduation or the wedding when they wanted to do that?

James Kotecki:

And when you zoom out even more, do you think about the kind of meta studies that maybe some AI systems are starting to look at? I saw something where a group was using AI to kind of look at a whole bunch of different COVID studies, for example, and maybe try and draw out some insights. But it was unclear if they could actually draw out insights or just kind of figure out what most people were saying, which is not necessarily the same thing as like a true insight from the data. But if you think about having a whole bunch of different trial data collected together and pooled together, and using AI to kind of look at that massive amount of data and pull out true information, where are we on that?

Brad Hirsch:

Yeah. So on the trial side of it’s complex, of course, because every company’s trial is 100% proprietary and everybody’s signed away. Nobody’s able to talk about what’s actually happening and nor would you want them to, because people aren’t making… You’ve seen this with different agents and COVID, where people have beliefs that one thing works or another doesn’t work, and it changes the way that people care for patients. And so the trial information isn’t there.

Brad Hirsch:

But where you see it a lot and it’s something… The woman who was my mentor for many years at Flatiron, who’s now the deputy commissioner at the FDA, a woman named Amy Abernethy, has has really been driving this real- world evidence approach, where you can aggregate huge data sets among health systems, out of electronic health records. And really be able to have forward looking insights into what’s working, what’s not. You can integrate social media to understand where hotspots are happening.

Brad Hirsch:

There’s so much out there. There’s the risk. And the risk that we saw with, again, everybody’s talking about COVID right now so that’s an example of, with COVID-19 there were two big studies that were put in the leading journals in the world, with Lancet and New England Journal of Medicine, that were ultimately retracted because the people that wrote the article didn’t have access to the data. And there’s risks of not really understanding the provenance of the data and exactly what it means. But at the end of the day, we’re seeing huge success of aggregating this data in the right way and making sense of it.

Brad Hirsch:

And the last piece of it is stuff that we’re really excited about, being able to do things like randomize and real world practice, right? So we are the operational infrastructure to say, that COVID patient came in, that there are five options available to them put them on one of those, and then let’s use sensors, and let’s use the personal survey information, and let’s use the electronic health record, and the application of ML across it, or AI across it, to really drive the insights into what’s possible, but also having the right sort of ethical constraints on it to make sure that we truly understand the data, that we’re doing it in the right way. And that it doesn’t lead to publications that are off base.

James Kotecki:

Yeah, we should definitely talk a little bit more about ethics in the final few minutes here. Because obviously medical ethics have been known and discussed for quite a long time, many decades, if not centuries. We now talking about AI ethics quite a bit, especially even on this show. How do those two worlds overlap for you? Do you see any specific, I mean, obviously there’s issues of data privacy, right? But beyond that, are there any specific issues, ethical issues of kind of medical/AI concern for you? Either that you don’t think people are taking seriously enough or you think people are taking too seriously and it’s impeding the progress that needs to be made?

Brad Hirsch:

Well, so in our world at SignalPath, it’s not as much of an issue because everything that we do for our sites is very much in line with the trials that they’re conducting. Our application of AI is more how do you take a PDF and put it into a system and QA it? How do you take data out of the system, put it [inaudible] system, and QA it? And so in that respect, the data is not being shared. It’s not being reused. All these patients are consented. And so from our day to day, it’s not as much of an issue. On the electronic health records side, and what that looks like, it is more of an issue because often patients are aren’t consented. It’s de-identified data.

Brad Hirsch:

Although there has been a lot of success in putting the right models in place to make sure the patients are empowered. They control their data. They can take it out of the data set either proactively or, there are a lot of ways that they can control how it’s used and what it looks like. And there’s a big push, including out of the Office of the National Coordinator, so the ONC, that oversees this to increasingly empower patients to have access to their data, to aggregate that data, and do what they want with it.

Brad Hirsch:

So the only time that I think it crosses the, the ethical consideration is if folks are aggregating data and trying to make assumptions without truly understanding it and truly knowing… As soon as it leaves the hospital or the health system where it was collected and it gets aggregated and aggregated again, is it really understood where it came from, how complete is it? Amy and the group at Flatiron put out a great article a couple of years ago about sort of the 10, the 10 elements of a high quality data set, and what you need to look at to make sure that you understand it and you can use it correctly. And so my biggest concern is just making sure that we know what to do with the data and how to interpret it correctly.

James Kotecki:

You’ve mentioned government entities a couple of times in this conversation. You also mentioned that a medical, electronic medical records were digitized because of a law that Congress passed 15, 20 years ago. I think that was George W. Bush, if I’m not mistaken. Does, is there something that you want the government to do to help your industry or make your life easier? Do you have any major kind of policy prescriptions that need to happen to advance your industry further?

Brad Hirsch:

Well, so interestingly, I think a lot of them are happening through the FDA. I think that it predated COVID and the pressures that COVID had put in place, but Amy and others really looking at, how do you look at the automated, how do you look at where regulation makes sense, where it doesn’t? How do you look at how to use real world evidence in the conduct of trials? How do you integrate sensors and other technologies?

Brad Hirsch:

How do you look — so adverse events are great examples. So every time an adverse event happens on a trial to a patient, meaning they have their blood pressure raised or something happened, their fatigue, everything gets reported. How do we really make sense of those elements, and only make sure that we’re collecting the ones that are critical to the content of the trial and your interpretation of it. And so there’s a lot of movement to, how do you leverage the technologies out there and the advances that are out there to reimagine drug development? And I think that’s the biggest area of pressure. And I think there’s already a ton of visibility thanks to, 21st Century Cures Act as an example of something that was pushing real-world data and how do we reimagine things?

James Kotecki:

Brad, let’s try to end on an optimistic note. You are a doctor in Texas, you’re the CEO of a company that sits at the center of kind of medical and AI issues. When you look at clinical trials, what’s something that you’re optimistic about?

Brad Hirsch:

Oh, I’m incredibly optimistic of where the space is going, of the ability to say, to reimagine the model to make sure that patients get access to trials, because they’re no longer just at predominantly large academic centers where you have to be there in person with a doc and a patient and a coordinator and a whole team behind it. I think increasingly we’re going to get to the model where you go into the clinic, you get the key thing that you need done. You have a relationship with the doc, they put you on the trial, but then you have sensors and surveys and you go to the local LabCorp to get your lab drawn.

Brad Hirsch:

That there’s a reimagination of a model that gives access to folks. So it’s much more ubiquitous and more people can get on trials. You decrease the burden on patients to get on those trials. And you have much deeper data sets to ask a lot more nuanced questions of not only how long until that patient progressed, but what was their experience. And so I’m incredibly bullish and it’s where we’re going with SignalPath 2.0, and why we spend so much time focused on it, because I think that it’s, the industry will change an incredible amount in the next five to 10 years.

James Kotecki:

Brad Hirsch, CEO of SignalPath. Thanks so much for joining us today on Machine Meets World.

Brad Hirsch:

Yep. Thanks, James. It was great to talk to you.

James Kotecki:

And thank you so much for watching. Machine Meets World is a production of Infinia ML. So whether you’re watching the video or listening to the podcast, you can do the other one because we distribute it as much as we possibly can on all possible channels. You also email the show. It’s M-M-W, Machine Meets World, @InfiniaML.com. I am James Kotecki, and that is what happens when Machine Meets World.

Image for post
Image for post

Originally published at https://infiniaml.com on July 7, 2020.

Machine Meets World from Infinia ML

Weekly Interviews with AI Leaders

Medium is an open platform where 170 million readers come to find insightful and dynamic thinking. Here, expert and undiscovered voices alike dive into the heart of any topic and bring new ideas to the surface. Learn more

Follow the writers, publications, and topics that matter to you, and you’ll see them on your homepage and in your inbox. Explore

If you have a story to tell, knowledge to share, or a perspective to offer — welcome home. It’s easy and free to post your thinking on any topic. Write on Medium

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store