Latest / The ACL Athlete Podcast / 273 | Using AI in ACL Rehab: Where AI Gets Dangerous in ACL Rehab and What It Should Never Replace - Part 2

Transcript

273 | Using AI in ACL Rehab: Where AI Gets Dangerous in ACL Rehab and What It Should Never Replace - Part 2 · The ACL Athlete Podcast

• 0:18
  SPEAKER_00: Welcome everyone to the ACL Athlete Podcast.
• 0:22
  SPEAKER_00: This is the podcast where we talk about everything related to the ACL, whether that's the injury itself, the rehab process, return to sport, and more.
• 0:31
  SPEAKER_00: I'm your host, Dr.
• 0:32
  SPEAKER_00: Ravi Patel, performance physical therapist and coach.
• 0:36
  SPEAKER_00: Between myself and fellow guests and experts, you'll learn through the lens of the patient, the healthcare professional, and the coach.
• 0:43
  SPEAKER_00: The goal of this podcast is to equip you, the athlete, with the education to make the best informed decision about your care and your ACL journey.
• 0:54
  SPEAKER_00: Thanks for joining.
• 0:55
  SPEAKER_00: Now let's dive into today's episode.
• 1:00
  SPEAKER_00: What is up, team, and welcome back to another episode on the ACL Athlete Podcast.
• 1:06
  SPEAKER_00: So today is part two of this series of using AI, artificial intelligence, in ACL rehab.
• 1:15
  SPEAKER_00: So if you have not listened to part one yet, go back and start there.
• 1:19
  SPEAKER_00: This builds directly on it.
• 1:21
  SPEAKER_00: And I really encourage you guys to listen to this because it's going to be really important as we talk through this episode, but especially just in general, this series is so important because it's going to be used so much more as time goes on.
• 1:37
  SPEAKER_00: It is going to be the forefront of how people get answers to their questions, how they search for things.
• 1:43
  SPEAKER_00: Google itself automatically has an AI-based summary overview that it creates.
• 1:49
  SPEAKER_00: So even if you Google something, you're getting things from AI.
• 1:52
  SPEAKER_00: Yes, you can go to specific links, but when people are searching these days, AI will be involved in the process and it will become second nature in terms of accessing information.
• 2:04
  SPEAKER_00: And so with this podcast episode series, this is going based on current, you know, 2026.
• 2:11
  SPEAKER_00: We'll see how this changes as time goes on because this stuff is rapidly changing so so fast.
• 2:17
  SPEAKER_00: And so this is something that I want to be helpful for you guys as you listen to this, because it is a bit of a challenging space using this.
• 2:27
  SPEAKER_00: Because as I shared in part one, it is something from an AI standpoint that is just pulling from the internet, is pulling from just information that it is fed.
• 2:39
  SPEAKER_00: It is not vetting whether it's legit.
• 2:41
  SPEAKER_00: It is not vetting whether it is not legit.
• 2:44
  SPEAKER_00: It is just information that it can pull and aggregate together based on a large language model.
• 2:51
  SPEAKER_00: That's what AI is.
• 2:53
  SPEAKER_00: And so, for a quick recap for part one, we covered essentially where AI is genuinely useful, understanding your diagnosis, preparing for surgical consultations, navigating the healthcare system, research support for clinicians.
• 3:09
  SPEAKER_00: We introduced the three-tier framework and got into tier two, where AI starts to get a little risky.
• 3:16
  SPEAKER_00: Today we are finishing tier two, going through tier three today, and closing with the three question framework you can use every single time going forward.
• 3:28
  SPEAKER_00: But first, I want to add something that came up after part one dropped.
• 3:33
  SPEAKER_00: And I think that this is something important to share with you guys as I was reflecting on this and essentially a use case I didn't cover that I think a lot of you will use it for.
• 3:44
  SPEAKER_00: And I was just thinking back to athletes working with us, and when they come in working with us, what are the things that they are sharing that they are struggling with?
• 3:53
  SPEAKER_00: And I think it's one of the most valuable things AI can actually do in this process, but it is also one of the trickiest things to get right.
• 4:03
  SPEAKER_00: So this is where the audit use case comes in.
• 4:08
  SPEAKER_00: And so, with that said, something I've been thinking about since part one is this idea of using AI as an audit tool.
• 4:17
  SPEAKER_00: And I think it applies to all three audiences the ACL athlete, the parent or supporting role, and the clinician or coach as well.
• 4:25
  SPEAKER_00: And here's what I mean: you're going through your rehab, maybe things feel off, maybe your gut is telling you something is not right, but you cannot articulate it.
• 4:35
  SPEAKER_00: Maybe your surgeon said something that did not fully land, like, you know, your three months and they said to start running, or you're six months and they cleared you for sport.
• 4:46
  SPEAKER_00: They said something that just doesn't necessarily fully land, or the PT gave you the guidance that you're just not really sure about.
• 4:53
  SPEAKER_00: It could be those same things of clearing you to start running at three months without any testing.
• 4:58
  SPEAKER_00: It could be, you know, knee extensions are very bad, but you see all this information about knee extensions being amazing.
• 5:04
  SPEAKER_00: Side note, they are very useful in rehab.
• 5:06
  SPEAKER_00: So whoever says otherwise is not really up to date, but I'll get away from that.
• 5:13
  SPEAKER_00: You are not trying to replace their judgment in these situations, right?
• 5:17
  SPEAKER_00: You are trying to just understand it better and figure out if the questions in your head are worth asking.
• 5:24
  SPEAKER_00: So it's a way for you to use AI as a soundboard, right?
• 5:29
  SPEAKER_00: Not to make the decision itself that's important here, but to help you audit the process against what the research and the science actually says.
• 5:38
  SPEAKER_00: And the thing here is that I even think about this podcast, right?
• 5:42
  SPEAKER_00: And yeah, there are a lot of things that I share my opinion about.
• 5:45
  SPEAKER_00: I share what this landscape and space of ACL rehab is, what return to performance and sport looks like, what the re-injury rates look like.
• 5:53
  SPEAKER_00: And we talk about a lot of the nuances in this process.
• 5:57
  SPEAKER_00: And I try my best to pull not just from anecdote.
• 6:02
  SPEAKER_00: Yes, I share those things from you guys, but if you look at our foundation of what we do, it is rooted in research.
• 6:09
  SPEAKER_00: It is rooted in science, it is rooted in data.
• 6:14
  SPEAKER_00: And this is where we have to make sure that this is the thing that we operate from.
• 6:19
  SPEAKER_00: It's not just, I had a hamstring graft, and therefore hamstring grafts are better, right?
• 6:25
  SPEAKER_00: Or I did this one type of therapy and this is what got me back to my sport.
• 6:30
  SPEAKER_00: Or I did this one exercise and it was the magic bullet for everything, right?
• 6:35
  SPEAKER_00: It's not picking all these methods, but it is being principles first.
• 6:39
  SPEAKER_00: First principles where you can distill this down into the very core basic thing.
• 6:43
  SPEAKER_00: Like physics is physics.
• 6:45
  SPEAKER_00: We're not changing physics.
• 6:46
  SPEAKER_00: And so when we go to manipulate exercise selection, like we are not changing what physics is defined as, the principles of it.
• 6:53
  SPEAKER_00: We're just using it as our principles to guide us in exercise selection, constraints, the way that we navigate this process.
• 7:01
  SPEAKER_00: And so that is the core of what we're looking at.
• 7:04
  SPEAKER_00: We use different models in this process, none of them are perfect, but the thing is that they help guide foundational principles for us to understand how we are rebuilding you from the ground up and back to being an athlete, being active.
• 7:18
  SPEAKER_00: And so these are so important in this process to understand what it is rooted in versus all these different methods or these particular ways of doing things.
• 7:29
  SPEAKER_00: There's so many different ways to get there.
• 7:31
  SPEAKER_00: And so I think, you know, having a soundboard is helpful and AI can help just process through this versus making the ultimate decisions.
• 7:39
  SPEAKER_00: And so here's some concrete examples of what that looks like here.
• 7:44
  SPEAKER_00: So you ask the AI, what does the research say about criteria for starting running after ACL reconstruction?
• 7:51
  SPEAKER_00: Not what does my PT say, or not what does my surgeon say.
• 7:55
  SPEAKER_00: What does the research say?
• 7:57
  SPEAKER_00: What is the aggregate of looking at different studies that have put data together, have you know had controls, have had different experimental groups, and they're looking at different biomechanical factors, different, you know, psychological factors, they're looking at lifestyle factors, they are looking at strength factors, human anatomy factors, all these things that are going to be variables that contribute to successful outcomes and maybe not so successful outcomes.
• 8:29
  SPEAKER_00: But we don't go based on what they just say.
• 8:32
  SPEAKER_00: We need to make sure that it is rooted in what the data and research shows.
• 8:37
  SPEAKER_00: And then, you know, you ask what the research says, and then you compare to what you're actually being told.
• 8:42
  SPEAKER_00: And this is probably the biggest dilemma that a lot of ACLers face is that I will hear so many people who reach out and they're like, my PT says this, my surgeon says this, and they're just not sure.
• 8:58
  SPEAKER_00: And it's okay to be not sure.
• 9:01
  SPEAKER_00: The thing is, is like, how do you become equipped to make sure you have more education and understanding to be able to go with what they say, to push back on what they say, or to just ask more questions, right?
• 9:15
  SPEAKER_00: And I think that's the most important thing is to inform yourself.
• 9:18
  SPEAKER_00: And that's what this podcast exists for.
• 9:20
  SPEAKER_00: It's not to replace anything, it's to help inform you in making better decisions because we know that ACL injuries are at an all-time high.
• 9:30
  SPEAKER_00: And we know that return to sport and performance is at an all-time low.
• 9:35
  SPEAKER_00: It just continues to get worse in both directions.
• 9:37
  SPEAKER_00: And we have not changed that yet.
• 9:39
  SPEAKER_00: Even with the most research and the most information about a particular injury, we're still not getting there yet.
• 9:47
  SPEAKER_00: And the thing that I've been thinking about too is that the information exists.
• 9:52
  SPEAKER_00: It exists.
• 9:53
  SPEAKER_00: Like, there's not going to be another study that shows, like, oh, doing this other training is going to be better than doing really good rehab and strength and conditioning and getting your quad stronger and all these other variables that are important to athletic performance that is going to get you there.
• 10:07
  SPEAKER_00: Like we know that the basics are going to get there.
• 10:09
  SPEAKER_00: There's not these magical exercises that is going to drastically change this process.
• 10:14
  SPEAKER_00: And so the thing is that most of this stuff exists, including testing.
• 10:20
  SPEAKER_00: If we did more of the evidence-based testing that exists, guess what?
• 10:24
  SPEAKER_00: We would have a drop in re-injury rates.
• 10:26
  SPEAKER_00: We have more people getting back to the things that they love.
• 10:29
  SPEAKER_00: Those are the two main variables that influence this process overall.
• 10:34
  SPEAKER_00: And a lot of it comes down to the education that surgeons and physical therapists share and the process and the frameworks that they use, more so the timeline-based protocols that are not helpful for these ACLers.
• 10:50
  SPEAKER_00: And a lot of this comes down to insurance-based limitations or healthcare limitations and the model itself.
• 10:57
  SPEAKER_00: So the model is what kind of stabs ACLers in the foot, if you will, because it's just hard because it is not designed for this rehab.
• 11:06
  SPEAKER_00: It is not designed for it.
• 11:08
  SPEAKER_00: And so this is where it is really important to root these things on understanding the research, science, and comparing these things to what you're actually being told.
• 11:19
  SPEAKER_00: So let's say also another example, you ask, you know, what does proper quad strength testing look like after ACL reconstruction?
• 11:26
  SPEAKER_00: And what are the accepted benchmarks before return to sport?
• 11:30
  SPEAKER_00: And the thing is, is that if you ask your physical therapist this or your surgeon when you first start, and this is a part of vetting the process, vetting them as being on your team, right?
• 11:41
  SPEAKER_00: So, so important.
• 11:44
  SPEAKER_00: That there are going to be a ton of PTs and surgeons who are going to tell you, oh, well, we'll check that when we get there, right?
• 11:52
  SPEAKER_00: They will kind of overlook the question because they can't answer it because they don't do it.
• 11:58
  SPEAKER_00: And so the thing is, is that point blank, I'm just going to tell you guys if a surgeon doesn't tell you what they're looking for, but especially your physical therapist, if they don't have a very specific, this is what strength testing looks like.
• 12:13
  SPEAKER_00: We're going to put you in an isokinetic machine, we're going to kick out, we're going to pull back, we're going to put you in isometric testing, get an objective value there.
• 12:20
  SPEAKER_00: We're going to do a three to five repetition maximum to be able to express that.
• 12:24
  SPEAKER_00: Whatever method they use, they need to say objectively, this is what we're looking for for running.
• 12:29
  SPEAKER_00: This is what we're looking for for being able to return to sport.
• 12:32
  SPEAKER_00: These are our benchmarks.
• 12:33
  SPEAKER_00: This is when we look at it.
• 12:34
  SPEAKER_00: This is how often we test.
• 12:36
  SPEAKER_00: These are things that need to be explicitly said.
• 12:40
  SPEAKER_00: Every single one of our athletes knows exactly what these are.
• 12:43
  SPEAKER_00: Every single meeting we have with them, we revisit these KPIs and benchmarks because it's what the research shows, but it lets them know where they are and what they're working towards.
• 12:54
  SPEAKER_00: Also, being able to just map out the plan and make it specific to them.
• 12:59
  SPEAKER_00: I can't stress this enough.
• 13:00
  SPEAKER_00: And so this is something where you can ask AI to be like, what are the proper strength testing for after ACL reconstruction and what are those benchmarks?
• 13:09
  SPEAKER_00: You look at the answer it gives you and ask yourself whether anyone has actually tested you that way.
• 13:15
  SPEAKER_00: And the thing is with the current landscape, don't be surprised if they haven't.
• 13:22
  SPEAKER_00: I would just kind of throw it out there.
• 13:24
  SPEAKER_00: Whenever I think about this space, what we learned in school, what it looks like is that when you think about the entire landscape of physical therapy, surgeons, and just for us working with athletes all over the world, getting such a massive sample size of different parts of the country, different parts of the world, and seeing what surgeons are communicating and physical therapists and what's going on and what people are cleared by, and especially with healthcare and insurance being in the mix.
• 13:55
  SPEAKER_00: I mean, guys, I would say this is a generous, generous percentage.
• 14:00
  SPEAKER_00: I would say 80 to 90 percent of physical therapists and surgeons are not testing ACLs.
• 14:07
  SPEAKER_00: That means 10 to 20 percent are doing it, and that's generous.
• 14:13
  SPEAKER_00: And I would almost say it's less than that.
• 14:16
  SPEAKER_00: And so that means that majority of physical therapy clinics, and I looked at all the ones here, even in Atlanta, and I like put them all up on the list, and I was like, okay, how many of these actually have an isometric dynamometer?
• 14:29
  SPEAKER_00: And it was less than 5% within you know, putting 20, 30 PC clinics around me, less than 5%.
• 14:38
  SPEAKER_00: So then therefore it's like, but people are going to all these different places to do ACL rehab.
• 14:43
  SPEAKER_00: So then we have to kind of think about like, okay, the reality is that it's the minority.
• 14:48
  SPEAKER_00: So then therefore, using AI to be able to just kind of process through a question like this to know what it should look like.
• 14:56
  SPEAKER_00: So then therefore you can compare and say, like, well, was I actually tested this way?
• 15:01
  SPEAKER_00: Another example here is you ask what is the evidence for, you know, return to sport after ACL reconstruction.
• 15:08
  SPEAKER_00: You get the list, strength symmetry, hop testing, you know, vertical hop testing, psychological readiness, a graded sport specific exposure, looking at power-based uh performance, looking at range of motion, knee symptom profile, all these different things.
• 15:24
  SPEAKER_00: So then you ask yourself whether any of those boxes have been checked and actually been evaluated in your own care, especially if you're nearing the end of this process.
• 15:34
  SPEAKER_00: That is a legitimate use of AI, using research and science as your foundation, not an opinion, not anecdote, and holding your care up against it.
• 15:45
  SPEAKER_00: The proper standard.
• 15:46
  SPEAKER_00: And the standard needs to be, you know, the bar, the bar is on the floor, floor, y'all.
• 15:51
  SPEAKER_00: That's how it just kind of feels a lot of the times.
• 15:53
  SPEAKER_00: Every time I take a call with people, I'm like, the bar is on the floor.
• 15:56
  SPEAKER_00: So we just need to, we need to pick it up.
• 15:58
  SPEAKER_00: We need to raise the standard, we need to elevate it, we need to make sure we're in a better place because you guys deserve better, and that's how we're gonna move things forward.
• 16:07
  SPEAKER_00: But here's the critical thing I have to say about this because it is easy to get it wrong.
• 16:13
  SPEAKER_00: AI is a quote unquote yes man.
• 16:17
  SPEAKER_00: That is not an exaggeration.
• 16:19
  SPEAKER_00: It is something that has been reported repeatedly, it does not push back on you nearly as much as it should.
• 16:25
  SPEAKER_00: It wants to give you what you want to hear.
• 16:27
  SPEAKER_00: It will shape its responses based on how you ask the question, especially.
• 16:31
  SPEAKER_00: I think I shared before garbage in, garbage out.
• 16:34
  SPEAKER_00: So you need to make sure if you ask it in ways that leads towards the answer you want, it will get you there.
• 16:40
  SPEAKER_00: It's just like I said, if you want to prove smoking will benefit your health, there is actual research out there to show that people have manipulated the process to get it published, right?
• 16:52
  SPEAKER_00: And it was just an experiment.
• 16:55
  SPEAKER_00: The thing is, like you can find the answers there.
• 16:59
  SPEAKER_00: And so then, therefore, it will give you that.
• 17:02
  SPEAKER_00: But the thing is, if you ask it neutrally, you might get a more honest answer.
• 17:06
  SPEAKER_00: And the thing that could be helpful with it is to make sure to give it something that's like, you know, be a devil's advocate, you know, push against this, be a contrarian, if you will.
• 17:17
  SPEAKER_00: So then, therefore, it's not necessarily always giving you everything you want.
• 17:21
  SPEAKER_00: And so you can think on both sides of this versus it being like, tell me what I want to hear, right?
• 17:27
  SPEAKER_00: So if you are going to use it to audit your care or your guidance, your prompt has to be neutral as possible, but also give it the guidance and the specifics it needs.
• 17:39
  SPEAKER_00: Do not ask, my PT has not tested my quad strength yet.
• 17:42
  SPEAKER_00: Is that bad?
• 17:43
  SPEAKER_00: That is a leading question, and AI will validate your concern whether it is warranted or not, just because you said, is that bad?
• 17:50
  SPEAKER_00: Right.
• 17:50
  SPEAKER_00: You could have said, is that good?
• 17:52
  SPEAKER_00: Right.
• 17:52
  SPEAKER_00: And then it could have maybe flipped the answer.
• 17:55
  SPEAKER_00: Ask instead, what does the research say about when and how quad strength should be assessed during ACER rehabilitation?
• 18:02
  SPEAKER_00: Now you are asking AI to report what the science says, not to validate a feeling you already have.
• 18:08
  SPEAKER_00: That is a much more useful answer.
• 18:11
  SPEAKER_00: Stick to research, data and science is your foundation.
• 18:15
  SPEAKER_00: That is where AI is and has the most to offer in this role.
• 18:20
  SPEAKER_00: The moment you start asking it to evaluate your specific situation, your specific PT, or your specific surgeon's guidance, you're moving into territory where it does not have enough information or context to give you a truly reliable answer.
• 18:36
  SPEAKER_00: Use it to get educated.
• 18:38
  SPEAKER_00: Use it to build better questions.
• 18:40
  SPEAKER_00: Use it to walk into your next appointment, knowing what the evidence says, so you can have a real conversation with the human who is actually responsible for your care.
• 18:51
  SPEAKER_00: That is the audit.
• 18:52
  SPEAKER_00: Done right is so genuinely powerful and so useful.
• 18:58
  SPEAKER_00: Done wrong, it is just another way to find the answer you are already looking for.
• 19:04
  SPEAKER_00: So be mindful of that.
• 19:06
  SPEAKER_00: But I think that that was very important to share with you guys because with this process, knowing how to use it and audit it is so, so valuable.
• 19:17
  SPEAKER_00: All right.
• 19:18
  SPEAKER_00: So that was a way longer tangent for sharing more part one, more than I was expecting.
• 19:26
  SPEAKER_00: But I felt like I had to share that with you guys because that felt like it was kind of missing a little bit as I was reflecting.
• 19:32
  SPEAKER_00: Um, and I think that that was important to share.
• 19:34
  SPEAKER_00: So now, picking up where part one left off, we talked about the confidence gap and the confirmation bias.
• 19:41
  SPEAKER_00: The athlete who asked AI the same question with different framing until they get the answer they want.
• 19:47
  SPEAKER_00: We just talked about this a little bit.
• 19:49
  SPEAKER_00: I want to add one more pattern here before we move into tier three.
• 19:53
  SPEAKER_00: The athlete who builds a program from AI and then uses AI to troubleshoot it.
• 19:59
  SPEAKER_00: No human ever in the loop, no one watching them move, no one tracking their quad, no one to catch the compensations or the patterns they develop in their gate, no one to notice that the swelling they keep describing as a little bit is actually telling them something important.
• 20:17
  SPEAKER_00: AI does not accumulate knowledge about you between sessions.
• 20:22
  SPEAKER_00: It resets every time.
• 20:23
  SPEAKER_00: It cannot connect the dots across weeks because it has no memory of your weeks, especially more of the nuanced details that are important here.
• 20:31
  SPEAKER_00: And it will not tell you it does not have enough information to answer.
• 20:35
  SPEAKER_00: It'll just answer it, right?
• 20:36
  SPEAKER_00: It'll go based on what it has available.
• 20:38
  SPEAKER_00: That combination is where things go wrong quietly over a long period of time, and you do not realize it until something forces the issue.
• 20:47
  SPEAKER_00: Don't be that person.
• 20:49
  SPEAKER_00: Now, the filter question for tier three what AI should never replace.
• 20:56
  SPEAKER_00: Does this require someone who can actually assess me, watch me move, and take responsibility for the outcome?
• 21:04
  SPEAKER_00: If yes, AI is not your answer.
• 21:09
  SPEAKER_00: And so here's one I want to share with you building your ACL rehab program.
• 21:13
  SPEAKER_00: AI can produce something that looks exactly like a program.
• 21:16
  SPEAKER_00: Phases, exercises, sets, reps, progressions, it will look legit.
• 21:21
  SPEAKER_00: That is the problem.
• 21:23
  SPEAKER_00: And I can't tell you guys the number of people that I've had where they have said, oh, well, I just kind of put in an AI and it built out something.
• 21:31
  SPEAKER_00: And so I just kind of worked with it.
• 21:33
  SPEAKER_00: And don't get me wrong, y'all.
• 21:34
  SPEAKER_00: I get it.
• 21:35
  SPEAKER_00: Like if you are, you know, dealing with bad PT or you just don't have enough resources, maybe visits ran out and you're trying to make it work, I get it.
• 21:46
  SPEAKER_00: So I'm not going to fault anyone here for doing it.
• 21:49
  SPEAKER_00: The thing is, is that my goal with this is to point out the just to be careful because it shouldn't replace it.
• 21:56
  SPEAKER_00: Because this is just something that you can't leave to raise.
• 22:00
  SPEAKER_00: Risk.
• 22:00
  SPEAKER_00: You can't leave it to chance.
• 22:01
  SPEAKER_00: It you just can't.
• 22:03
  SPEAKER_00: And so what the thing about this AI or putting it into whatever chat bot you have, it does not know is your graph type.
• 22:14
  SPEAKER_00: Sure, you can give it to it.
• 22:15
  SPEAKER_00: And what that means for tissue maturation, the loading timelines, your quad deficit, like how is it testing it?
• 22:22
  SPEAKER_00: Like, is it actually giving you valid ways to test it and assess this?
• 22:26
  SPEAKER_00: Are you doing it the right way?
• 22:28
  SPEAKER_00: Which is the single most important modifiable risk factor for reinjury and which AI cannot truly measure, right?
• 22:35
  SPEAKER_00: Your swelling pattern week to week, how you actually move, your concomitant injuries, your sport demands, your psychological readiness.
• 22:42
  SPEAKER_00: The program AI builds is the average of a generic ACL content on the internet, which is protocol-driven, time-based approach that already fails athletes.
• 22:53
  SPEAKER_00: It is not built from the assessment.
• 22:56
  SPEAKER_00: It is not responsive to how you are actually doing.
• 22:59
  SPEAKER_00: Yes, you can give it those contexts and it might kind of layer some of that stuff in.
• 23:04
  SPEAKER_00: But again, it's pulling from especially the information from just the general ACL content that lives out in the world.
• 23:12
  SPEAKER_00: And like I said, if you Google ACL rehab on the internet, it is going to show up with the most generic protocols in this process and most of the surgeons and hospital sites that are just general stuff.
• 23:27
  SPEAKER_00: So that's stuff to factor into this, unless you feed it particular sources, right?
• 23:32
  SPEAKER_00: Like that could influence things.
• 23:34
  SPEAKER_00: But at the same time, do you really want to leave that to chance in this situation?
• 25:42
  SPEAKER_00: And so I wanted to say something directly to athletes using AI to build programs because they had, you know, no other options, right?
• 25:50
  SPEAKER_00: And I kind of shared this earlier, but the system failed you, right?
• 25:54
  SPEAKER_00: If that's what you're doing, that is real.
• 25:56
  SPEAKER_00: But you deserve more than something generic, template from a language model or AI.
• 26:02
  SPEAKER_00: You know, there are different in-person options that exist that you could explore.
• 26:07
  SPEAKER_00: There's, of course, remote coaching options that exist.
• 26:10
  SPEAKER_00: We do those things.
• 26:11
  SPEAKER_00: There are so many other amazing folks in the space who do this.
• 26:14
  SPEAKER_00: And they could do it from anywhere, right?
• 26:16
  SPEAKER_00: This is something that you can reach out to us and we're happy to point you in the right direction.
• 26:20
  SPEAKER_00: But this is exactly the situation we built our services for to help fill in the gap.
• 26:26
  SPEAKER_00: And serving hundreds and hundreds of ACLers currently and in the past to be able to fill in these gaps where people just kind of fall through, right?
• 26:36
  SPEAKER_00: For a number of different reasons, because our healthcare system, insurance, because of just poor education from a physical therapy or surgical standpoint is hard.
• 26:47
  SPEAKER_00: Another area, the return to sport decision.
• 26:52
  SPEAKER_00: Return to sport is a criteria-based decision.
• 26:56
  SPEAKER_00: There's quad strength symmetry we look at.
• 26:58
  SPEAKER_00: There is hopping ability from a vertical hop testing.
• 27:02
  SPEAKER_00: We look at deceleration mechanics, we look at different movement quality under load.
• 27:07
  SPEAKER_00: We look at different positions, we look at power output, we look at dynamic tasks like jumping, cutting, different things like that, agility, psychological readiness.
• 27:19
  SPEAKER_00: And then we're also looking at graded sport and activity exposure.
• 27:22
  SPEAKER_00: All of it assessed by a human who has been watching you progress.
• 27:26
  SPEAKER_00: Yes, and we do this remotely too.
• 27:27
  SPEAKER_00: This is something that video and technology is so helpful for.
• 27:31
  SPEAKER_00: You get such a detailed look at things that we get to share with our athletes that is so incredible.
• 27:37
  SPEAKER_00: Um, I had someone who reached out was like, how do you even do this remotely?
• 27:40
  SPEAKER_00: And I'm like, it's it's possible with the way that technology is.
• 27:44
  SPEAKER_00: And then also when we look at potentially the amount of time or actual eyes on you at an insurance-based clinic, it's usually not as much as people think.
• 27:56
  SPEAKER_00: And so then therefore, you take the generic in-person or the individualized remote, right?
• 28:01
  SPEAKER_00: And it could be a blend of both.
• 28:03
  SPEAKER_00: We've got people who are doing both.
• 28:04
  SPEAKER_00: And the thing is, is that it can be done.
• 28:07
  SPEAKER_00: And so the thing is that you just need to make sure that people are testing based on what the research shows.
• 28:13
  SPEAKER_00: You need to make sure that there is an individualized plan and program that is progressed each and every week based on your goals and criteria, and making sure that you have expert guidance, people who know this process, who have a rehab background, who have a performance background, and is able to get you back to the end goal and explain how we're going to get you there.
• 28:33
  SPEAKER_00: That is so important.
• 28:35
  SPEAKER_00: The thing is, is that it needs to be looked at from all these different areas for return to sport decisions.
• 28:42
  SPEAKER_00: AI can tell you what that list is, but it can't tell you whether you meet it.
• 28:47
  SPEAKER_00: Those are completely different things.
• 28:49
  SPEAKER_00: An athlete who asks AI if they are ready and gets a permissive answer has not been cleared.
• 28:57
  SPEAKER_00: They had a conversation with a uh with an AI language model, ACL reindury rates, and athletes returning to pivoting sports sits between 15 to 25%.
• 29:08
  SPEAKER_00: I'd probably say 20 to 30 percent is usually the main kind of like thing that you will hear.
• 29:14
  SPEAKER_00: That number does not come down by skipping the testing, it comes down by doing it right, y'all.
• 29:21
  SPEAKER_00: The other thing that you can look at, the ongoing coaching relationship, the human in your corner accumulates knowledge about you that no AI interaction can replicate.
• 29:30
  SPEAKER_00: They know your history, your tendencies, your patterns.
• 29:33
  SPEAKER_00: They notice things that are changing week over week that you cannot see yourself.
• 29:38
  SPEAKER_00: AI may save some of that stuff, but it will also kind of like forget some of that stuff.
• 29:43
  SPEAKER_00: It cannot do what a good physical therapist and coach does every month.
• 29:47
  SPEAKER_00: These are not comparable things.
• 29:49
  SPEAKER_00: If I could just kind of share with you guys like the things that we have to like kind of navigate for our athletes in this process, and people all often think like, oh, it's an exercise program.
• 30:00
  SPEAKER_00: And it's like, no way, it's so much more than that.
• 30:03
  SPEAKER_00: Like, we are there from being able to answer any and every question, navigating graph types, whether people should or should not have surgery, when they should have surgery from an optimal timing and life standpoint, to understanding yes, what rehab things are going to be important, the goals and criteria.
• 30:23
  SPEAKER_00: But you talk about someone who goes through a family loss, someone who's going through challenges at school, a breakup, a relationship, right?
• 30:32
  SPEAKER_00: We are there to help them with trying to work through the rehab process while understanding these things, these contextual factors, because people's lives don't just exist in a vacuum for ACL rehab.
• 30:44
  SPEAKER_00: You know, they are dealing with all these mental hurdles with this rehab and then all of a sudden have to deal with also life itself, right?
• 30:52
  SPEAKER_00: You know, we have the mom who has a six-month-old who can't carry her six-month-old because she's non-weight bearing for six weeks, right?
• 30:59
  SPEAKER_00: You have someone who can't drive for two months because they can't get their quad going and is on the driving leg side, right?
• 31:07
  SPEAKER_00: So that really influences people, but then you start start talking about like how they get in their head, how that influences so much of this.
• 31:15
  SPEAKER_00: And just being able to also just look out for things that are coming up that we're noticing this athlete says, Oh, well, this bugged my knee, this bugged my knee, or maybe this day they notice something.
• 31:25
  SPEAKER_00: And then we start to see the accumulation and we try to get ahead of the problem before it becomes, you know, this big fire.
• 31:32
  SPEAKER_00: Maybe at first it's just this small little like fire, or you see some smoke.
• 31:37
  SPEAKER_00: Guess what?
• 31:38
  SPEAKER_00: We could put it out quick.
• 31:39
  SPEAKER_00: And that's our goal with this.
• 31:41
  SPEAKER_00: And to be honest, like it's one of those things with AI that it's just not going to pick up on some of this stuff that is so detailed and nuanced that is so important in this process because we're working on the whole person.
• 31:53
  SPEAKER_00: It's not just an exercise, it's not just this test, right?
• 31:57
  SPEAKER_00: And that brings me to psychological support.
• 32:00
  SPEAKER_00: You know, being able to explain kinesophobia, for example, like the fear of movement, sure, it can help normalize fear of reinjury, maybe.
• 32:09
  SPEAKER_00: That's a tier one that I had talked about, and it can be useful.
• 32:13
  SPEAKER_00: But what it can't do is work through it with you.
• 32:16
  SPEAKER_00: The identity disruption, the grief, the fear that lives in your body when you try to push, that requires a real human, real expertise.
• 32:23
  SPEAKER_00: Sport psychology is the most underutilized part of this process, and using AI as a substitute for it is a gap that shows up in our outcomes.
• 32:32
  SPEAKER_00: And so we do need to be very, very, very mindful of that.
• 32:36
  SPEAKER_00: And so one of the things I want to share here is in terms of how we're using it at the ACL athlete.
• 32:43
  SPEAKER_00: And I want to be specific here because I think it matters.
• 32:46
  SPEAKER_00: And we're not here to say no AI, like we hate it.
• 32:50
  SPEAKER_00: We are a tech forward company.
• 32:52
  SPEAKER_00: We use it, we use our technology.
• 32:55
  SPEAKER_00: That is how we work with our athletes.
• 32:58
  SPEAKER_00: And it's something that I think is really important to have it built intentionally because it is here to stay.
• 33:05
  SPEAKER_00: And when used correctly, it makes us better at what we do as a team individually, to show up better for our athletes, to give them what we need.
• 33:15
  SPEAKER_00: It helps us so much.
• 33:17
  SPEAKER_00: But I will be clear here, it is not a replacement for what we do here at the ASI athlete from coaching to communications to anything that we do here.
• 33:27
  SPEAKER_00: That is something that is very important to me as the owner and our team as we talk about it.
• 33:33
  SPEAKER_00: We are so intentional about the use cases for it.
• 33:37
  SPEAKER_00: Now, here's where we can really allow it to help us, right?
• 33:41
  SPEAKER_00: Research deep dives.
• 33:43
  SPEAKER_00: When a complex case comes in, we can move through the evidence faster.
• 33:47
  SPEAKER_00: We have brains that can kind of filter through seeing some papers, looking at the way that it consolidates information, looking at the resources.
• 33:56
  SPEAKER_00: The judgment is still ours, of course.
• 33:58
  SPEAKER_00: It just allows us to aggregate the data and information faster to make better decisions and also potentially share some of this stuff with our athletes.
• 34:07
  SPEAKER_00: It allows us from an analyzing testing data and building out dashboards.
• 34:12
  SPEAKER_00: When we're working through performance metrics and KPIs and looking at compliance data, AI helps us to identify patterns efficiently.
• 34:20
  SPEAKER_00: Again, it's not making the decisions for us, it's helping us to look at these things and be like, okay, here's what we see from you know athlete A and here's what we see from athlete B.
• 34:30
  SPEAKER_00: You know, the testing still happened.
• 34:32
  SPEAKER_00: We're still doing that.
• 34:33
  SPEAKER_00: A human designed it, and now, you know, we're open owning the interpretation of what this information is.
• 34:40
  SPEAKER_00: It's just collecting some of this data to help us to be like, okay, let's see this cleaner and let's build this out better for our athletes and share this with our athletes.
• 34:49
  SPEAKER_00: So then that way they also have value from it too.
• 34:51
  SPEAKER_00: And that's the two sides of this.
• 34:53
  SPEAKER_00: We're not just trying to use it just from our side of things, but also trying to deliver as much value to our athletes as much as we can when we're utilizing things like this.
• 35:04
  SPEAKER_00: Translation, we work with athletes in other countries.
• 35:06
  SPEAKER_00: I've had MRI reports in so many different languages, like German, surgical notes in other languages, that I'm like, I don't I can't read this.
• 35:14
  SPEAKER_00: And I've tried to put in a Google Translate and it comes through and it's okay.
• 35:18
  SPEAKER_00: Guess what?
• 35:19
  SPEAKER_00: Now, with these models, like I can throw it into one, I can blur out the personal information, which we always do, and then being able to take that information into our clinical judgment and to have that context, especially in a different language.
• 35:34
  SPEAKER_00: Um, complex imaging and reports, it helps us parse through language and detailed, you know, kind of clinical documents faster so we can focus on that interpretation versus man, you guys have read MRI reports, radiologists like to get nitty-gritty detailed, and it's almost kind of weird how detailed they get of certain things.
• 35:52
  SPEAKER_00: And you're just like, this has no importance here, but obviously they are doing everything to read the ins and outs of the details.
• 35:59
  SPEAKER_00: And then summarizing coaching feedback after detailed sessions, AI helps us to produce a clean reference for our athletes to be able to share with them.
• 36:08
  SPEAKER_00: Okay, this is what we talked about.
• 36:10
  SPEAKER_00: The coaching happened, the relationship happened, you know, those are the things that are important.
• 36:15
  SPEAKER_00: It allows us to just kind of share a summary.
• 36:17
  SPEAKER_00: You guys know how you get off a call, for example, and you're like, okay, we talked about a lot.
• 36:22
  SPEAKER_00: So how could I get that consolidated into a review and action items?
• 36:27
  SPEAKER_00: Well, guess what?
• 36:27
  SPEAKER_00: We can share that with our athletes, and we do tend to do that with a lot of our ACLers.
• 36:32
  SPEAKER_00: So then, therefore, they know exactly what did we review?
• 36:36
  SPEAKER_00: What is it that we're putting into play?
• 36:37
  SPEAKER_00: What are some action items?
• 36:39
  SPEAKER_00: So that allows us to be organized, but then also be very intentional because there's a lot of information that we have to process.
• 36:47
  SPEAKER_00: It could be helpful in finding surgeons, you know, when an athlete in a new city needs a referral.
• 36:52
  SPEAKER_00: Sometimes AI helps, sometimes it's not great, to be honest.
• 36:55
  SPEAKER_00: We still vet everything, but there are some cities where I'm like, I know no one close to you within two to three hours.
• 37:03
  SPEAKER_00: And so then, therefore, it can help with just pulling information from other areas outside of just a Google search.
• 37:09
  SPEAKER_00: So there could be value in things like that.
• 37:12
  SPEAKER_00: What AI does not do in our practice, it does not build programming.
• 37:17
  SPEAKER_00: That is such a hard and fast role for us.
• 37:19
  SPEAKER_00: We do the programming, we are the coaches, we're the physical therapists, we do our programming that we individualize to each athlete.
• 37:27
  SPEAKER_00: It does not make any type of return to sport decisions or calls.
• 37:31
  SPEAKER_00: As I shared, we can aggregate some data with it to help us put some things together and clean it up, but it makes no decisions.
• 37:38
  SPEAKER_00: It allows us to synthesize things better.
• 37:41
  SPEAKER_00: It does not replace the coaching relationship.
• 37:43
  SPEAKER_00: Y'all, this is by far the most important thing for us.
• 37:47
  SPEAKER_00: Could we automate a bunch of stuff?
• 37:49
  SPEAKER_00: 100%.
• 37:51
  SPEAKER_00: Would it reduce time for us?
• 37:52
  SPEAKER_00: 100%.
• 37:54
  SPEAKER_00: Would it remove what we care about and value so much, which is that coach athlete relationship?
• 37:59
  SPEAKER_00: 100%.
• 38:00
  SPEAKER_00: And so we're not going to do that.
• 38:02
  SPEAKER_00: The coaching relationship is the core of what it is.
• 38:05
  SPEAKER_00: And when I meet athletes and when they connect with our team, that's so key.
• 38:09
  SPEAKER_00: And it's what drives us to be able to connect with these people and build relationships.
• 38:14
  SPEAKER_00: The check-ins we have with them, the real conversations, getting to know them, you know, their kids, you know, where they're going to school, what kind of sports, you know, whether they have a dog, like, you know, traveling, like it's more than just a rehab process.
• 38:28
  SPEAKER_00: And that's important to us in terms of making sure this is a human-to-human thing.
• 38:33
  SPEAKER_00: And that is one thing AI is not going to replace for us here at the ACL athlete.
• 38:38
  SPEAKER_00: The week over week knowledge that builds about who an athlete is and what they need, that is irreplaceable.
• 38:45
  SPEAKER_00: And we are not trying to replace it.
• 38:49
  SPEAKER_00: That's important to us.
• 38:51
  SPEAKER_00: We want to make sure this is a human-to-human thing and we allow AI to be something that helps us improve the value of what we're doing here, but never replace it.
• 39:02
  SPEAKER_00: The goal is never to shortcut the things that matter.
• 39:05
  SPEAKER_00: The goal is to be efficient on the surrounding work so we have more capacity to do the important work better.
• 39:11
  SPEAKER_00: No AI can do that.
• 39:12
  SPEAKER_00: What a skilled clinician and coach does for an athlete going through this.
• 39:16
  SPEAKER_00: And the thing that I've had people ask me is like, are you scared about AI replacing what you're doing, especially from a remote standpoint?
• 39:23
  SPEAKER_00: And I have no concerns whatsoever.
• 39:26
  SPEAKER_00: And I know that that sounds very maybe overly confident, if you will, but it's nothing about that.
• 39:34
  SPEAKER_00: It's just so much of this process is embedded in, like, I feel it in my bones in terms of what this entails and the calls I have with athletes and the people who enter our ecosystem and what this injury does to them, because I've felt this twice and I've had numerous injuries to recover from.
• 39:53
  SPEAKER_00: So I know what this is like on the personal level.
• 39:56
  SPEAKER_00: And the thousands and thousands of conversations I've had with ACLers and working with athletes, that it's just something that it's just not going to replace what we do because of the detail that goes into this process.
• 40:10
  SPEAKER_00: And the biggest thing here that I want to express, because you have to be careful with AI, is that the risk is just too high.
• 40:17
  SPEAKER_00: Whenever one in every three to four athletes are going to go on to get injured again, the risk is way too high.
• 40:24
  SPEAKER_00: So to use, you know, AI or social media or YouTube or combinations of these things to kind of formulate your rehab, it's just not going to cut it.
• 40:35
  SPEAKER_00: Because at the end of the day, when you're looking at a 20 to 30% re-injury risk, the detail of expertise and the nuances of this process, even the graph type, is going to matter so, so, so, so much.
• 40:50
  SPEAKER_00: And so that is something that I want to really stress here because it's so important from an AI standpoint, and whether I truly think like if you're really good at this process and you really care about it, it is not going to come in and replace things.
• 41:05
  SPEAKER_00: But I will see it replacing generic rehab.
• 41:08
  SPEAKER_00: It will come a day where people will be pushed towards apps and doing things at home more and more through an automated process.
• 41:16
  SPEAKER_00: And we're starting to see this even with some of these more online telehealth-based companies.
• 41:22
  SPEAKER_00: But with this ACL rehab, the nuance is so important and it needs to be dialed in.
• 41:27
  SPEAKER_00: And so the thing is that for us, we want to be skilled clinicians and coaches and be there for our athletes and do the things that are important.
• 41:37
  SPEAKER_00: And I promise you that, and we are not interested in pretending otherwise in this process.
• 41:44
  SPEAKER_00: So now to close this out, three questions, use them every time from a decision framework standpoint in terms of vetting AI.
• 41:55
  SPEAKER_00: One, am I using this to get smarter and more prepared, or am I using it to make a clinical decision?
• 42:01
  SPEAKER_00: Getting smarter, that is what AI is for.
• 42:05
  SPEAKER_00: The moment you're using it to decide something that requires clinical assessment, you have crossed the line.
• 42:11
  SPEAKER_00: Two, if AI gets this wrong, what is the actual consequence?
• 42:16
  SPEAKER_00: This is a good one.
• 42:18
  SPEAKER_00: Wrong about a term?
• 42:19
  SPEAKER_00: You can ask your PT to clarify.
• 42:22
  SPEAKER_00: That's fine.
• 42:23
  SPEAKER_00: Wrong about whether you are ready to return to sport, you retain at nine months.
• 42:28
  SPEAKER_00: The consequence tells you how much human expertise needs to be in the loop.
• 42:33
  SPEAKER_00: You cannot risk another re-injury, y'all.
• 42:35
  SPEAKER_00: You just cannot.
• 42:36
  SPEAKER_00: And the thing is, the way to stop another re-injury, the best reduction of risk, the best rehab individualized to you, and the best testing that is going to make sure to get you there.
• 42:48
  SPEAKER_00: I just had to tell an athlete that we're not there yet as much as tryouts are coming up, and we had to talk through it and with our parents.
• 42:57
  SPEAKER_00: And one of the things that came up with the parents, especially is they're like, Well, how much risk?
• 43:03
  SPEAKER_00: I said, What's the cost of another ACL injury?
• 43:06
  SPEAKER_00: Especially after a long enough and hard enough process already.
• 43:10
  SPEAKER_00: So this is the actual important thing.
• 43:12
  SPEAKER_00: If AI gets this wrong, what is the actual consequence?
• 43:16
  SPEAKER_00: Number three, last one, does this require someone who can actually assess me, watch me move, and take responsibility for the outcome?
• 43:24
  SPEAKER_00: If yes, that is a human job.
• 43:26
  SPEAKER_00: AI supports it.
• 43:27
  SPEAKER_00: AI does not do it itself.
• 43:31
  SPEAKER_00: So if you're an athlete right now, you deserve someone who knows you, tests you, watches you move, and builds your program around your actual data and individualized to you.
• 43:39
  SPEAKER_00: AI can help you find that person and show up prepared to work with them.
• 43:44
  SPEAKER_00: It cannot be that person.
• 43:47
  SPEAKER_00: If you're a parent or a supporting role, use AI to get educated and ask better questions.
• 43:52
  SPEAKER_00: If something feels wrong, seek another qualified human opinion.
• 43:56
  SPEAKER_00: That is the right move here.
• 43:58
  SPEAKER_00: If you're a clinician or a coach, AI is leveraged.
• 44:01
  SPEAKER_00: The filter is you.
• 44:03
  SPEAKER_00: If you remove yourself from that filter, you are not using AI as a tool.
• 44:08
  SPEAKER_00: You're asking it to do your job.
• 44:11
  SPEAKER_00: AI is a powerful tool, y'all, and it's a powerful crutch.
• 44:16
  SPEAKER_00: I'm gonna say that again because it's good.
• 44:19
  SPEAKER_00: AI is a powerful tool and it is a powerful crutch.
• 44:25
  SPEAKER_00: The difference is whether you're using it to show up better for the humans in your corner, or are you trying to use it to replace them?
• 44:33
  SPEAKER_00: If you want an actual human in your corner from surgery through return to sport and beyond, you can reach out to us.
• 44:40
  SPEAKER_00: We are going to express working with folks more.
• 44:45
  SPEAKER_00: Um, that is something that I've kind of shied away from, but I realize that it is something that is not serving the greater majority of you that don't necessarily know that we do remote coaching, that we work with people all over the world.
• 44:58
  SPEAKER_00: So if you need help, please find us um at theathlete.com.
• 45:03
  SPEAKER_00: You can also look in the show notes for one-on-one remote coaching.
• 45:07
  SPEAKER_00: We have worked with so many different sports, different athletes, different ages.
• 45:13
  SPEAKER_00: And a lot of times people are skeptical of the remote process.
• 45:16
  SPEAKER_00: And that's why we get to do a free consultation together to be able to decide if this is a good fit.
• 45:22
  SPEAKER_00: And if not, I promise you that you will have next steps and clarity in the direction you go, whether it's with us or whether it is in a different direction.
• 45:30
  SPEAKER_00: I just want to make sure you have more power in your story, more clarity.
• 45:34
  SPEAKER_00: So then that way you have next steps.
• 45:36
  SPEAKER_00: And that's so important.
• 45:38
  SPEAKER_00: What I want from you guys is to share this series with someone who needs it, another athlete, a parent, a clinician.
• 45:45
  SPEAKER_00: This conversation matters so much, especially for the future ahead.
• 45:49
  SPEAKER_00: See you in the next episode.
• 45:51
  SPEAKER_00: This is your host, Robbie Patel, signing off.