User login
Stanford (Calif.) University’s human performance lab sits next to its physical therapy clinic, so orthopedic surgeons often stop by to request biomechanical analyses for their patients, such as athletes with repeat injuries.
“It would take us days to analyze the data, so we would only do it a handful of times per year,” said Scott Uhlrich, PhD, director of research at the lab.
Now an app can do the job in less than 10 minutes.
The motion-capture app, created by Dr. Uhlrich and fellow bioengineers at Stanford, could help clinicians design better interventions to ward off mobility problems and speed recovery. It could also help researchers fill huge knowledge gaps about human mobility.
It’s currently available free for research and educational use. Model Health, a startup affiliated with the Stanford researchers, provides licenses for commercial use and clinical practice.
Here’s how it works. Footage of human movement, recorded by two smartphones, gets uploaded to the cloud, where an algorithm identifies a set of points on the body. The app relies on computer vision algorithms, a form of AI that trains computers to “understand” visual data – in this case, a person’s pose.
Next, the app quantifies how the body is moving through three-dimensional space. Musculoskeletal system models reveal insights into that movement, such as the angle of a joint, the stretch in a tendon, or the force being transferred through the joints.
“These are the quantities that relate to injuries and disease,” said Dr. Uhlrich, co-author of a study introducing the app. “We need to get to those quantities to be able to inform medical research and eventually clinical practice.”
The conventional approach to getting this kind of analysis requires special expertise and costs $150,000. By contrast, the app is free and easy to use.
It “democratizes” human movement analysis, said senior study author Scott Delp, PhD, professor of bioengineering and mechanical engineering at Stanford. The researchers hope this will “improve outcomes for patients across the world.”
‘Endless opportunities’
A lot about human mobility remains mysterious.
In aging adults, researchers can’t say when balance starts to degrade or by how much every year. They’re also still unraveling how sports injuries occur and how degenerative joint diseases like arthritis progress.
“We don’t really understand the onset of a lot of things, because we’ve just never measured it,” Dr. Uhlrich said.
OpenCap could help change that in a big way. Although biomechanics studies tend to be small – just 14 participants, on average – the app could allow for much larger studies, thanks to its lower cost and ease of use. In the study, the app collected movement data on 100 participants in less than 10 hours and computed results in 31 hours – an effort that would otherwise have taken a year.
“Studies of hundreds will be common, and thousands will be feasible, especially if assessments are integrated into clinic visits,” Dr. Uhlrich said.
About 2,600 researchers around the world are already using the app, according to Dr. Uhlrich. Many had never created a dynamic simulation before.
“The opportunities here are endless,” said Eni Halilaj, PhD, an assistant professor of mechanical engineering at Carnegie Mellon, Pittsburgh, who was not involved in creating the app. That’s especially true for “highly heterogeneous conditions that we have not been able to fully characterize through traditional studies with limited patients.”
In one case, researcher Reed Gurchiek, a former Stanford postdoc and current professor at Clemson (S.C.) University, used the app to study hamstring strain injuries during sprinting and found that these muscles lengthen faster during acceleration, compared with running at a constant speed.
“This aligns with the higher observed injury rates when athletes are accelerating,” Dr. Uhlrich explained. “Varied-speed sprinting studies are not possible in the lab, so this was really enabled by OpenCap’s portability.”
Movement as a biomarker
The researchers are already using the app to build new tools, including metrics to identify risk for anterior cruciate ligament injury in young athletes and to measure balance.
Someday, the technology could augment annual physicals, establishing movement as a biomarker. By having patients perform a few movements, like walking or standing up, clinicians could assess their disease risk and progression or their risk of falling.
Excessive loading in the knee joint puts patients at higher risk of developing osteoarthritis, for instance, but clinicians can’t easily access this information. The disease is typically diagnosed after symptoms appear, even though intervention could happen much earlier.
“Prevention is still not as embraced as it should be,” said Pamela Toto, PhD, professor of occupational therapy at the University of Pittsburgh, who also was not involved in making the app. “If we could tie the technology to intervention down the road, that could be valuable.”
A version of this article first appeared on Medscape.com.
Stanford (Calif.) University’s human performance lab sits next to its physical therapy clinic, so orthopedic surgeons often stop by to request biomechanical analyses for their patients, such as athletes with repeat injuries.
“It would take us days to analyze the data, so we would only do it a handful of times per year,” said Scott Uhlrich, PhD, director of research at the lab.
Now an app can do the job in less than 10 minutes.
The motion-capture app, created by Dr. Uhlrich and fellow bioengineers at Stanford, could help clinicians design better interventions to ward off mobility problems and speed recovery. It could also help researchers fill huge knowledge gaps about human mobility.
It’s currently available free for research and educational use. Model Health, a startup affiliated with the Stanford researchers, provides licenses for commercial use and clinical practice.
Here’s how it works. Footage of human movement, recorded by two smartphones, gets uploaded to the cloud, where an algorithm identifies a set of points on the body. The app relies on computer vision algorithms, a form of AI that trains computers to “understand” visual data – in this case, a person’s pose.
Next, the app quantifies how the body is moving through three-dimensional space. Musculoskeletal system models reveal insights into that movement, such as the angle of a joint, the stretch in a tendon, or the force being transferred through the joints.
“These are the quantities that relate to injuries and disease,” said Dr. Uhlrich, co-author of a study introducing the app. “We need to get to those quantities to be able to inform medical research and eventually clinical practice.”
The conventional approach to getting this kind of analysis requires special expertise and costs $150,000. By contrast, the app is free and easy to use.
It “democratizes” human movement analysis, said senior study author Scott Delp, PhD, professor of bioengineering and mechanical engineering at Stanford. The researchers hope this will “improve outcomes for patients across the world.”
‘Endless opportunities’
A lot about human mobility remains mysterious.
In aging adults, researchers can’t say when balance starts to degrade or by how much every year. They’re also still unraveling how sports injuries occur and how degenerative joint diseases like arthritis progress.
“We don’t really understand the onset of a lot of things, because we’ve just never measured it,” Dr. Uhlrich said.
OpenCap could help change that in a big way. Although biomechanics studies tend to be small – just 14 participants, on average – the app could allow for much larger studies, thanks to its lower cost and ease of use. In the study, the app collected movement data on 100 participants in less than 10 hours and computed results in 31 hours – an effort that would otherwise have taken a year.
“Studies of hundreds will be common, and thousands will be feasible, especially if assessments are integrated into clinic visits,” Dr. Uhlrich said.
About 2,600 researchers around the world are already using the app, according to Dr. Uhlrich. Many had never created a dynamic simulation before.
“The opportunities here are endless,” said Eni Halilaj, PhD, an assistant professor of mechanical engineering at Carnegie Mellon, Pittsburgh, who was not involved in creating the app. That’s especially true for “highly heterogeneous conditions that we have not been able to fully characterize through traditional studies with limited patients.”
In one case, researcher Reed Gurchiek, a former Stanford postdoc and current professor at Clemson (S.C.) University, used the app to study hamstring strain injuries during sprinting and found that these muscles lengthen faster during acceleration, compared with running at a constant speed.
“This aligns with the higher observed injury rates when athletes are accelerating,” Dr. Uhlrich explained. “Varied-speed sprinting studies are not possible in the lab, so this was really enabled by OpenCap’s portability.”
Movement as a biomarker
The researchers are already using the app to build new tools, including metrics to identify risk for anterior cruciate ligament injury in young athletes and to measure balance.
Someday, the technology could augment annual physicals, establishing movement as a biomarker. By having patients perform a few movements, like walking or standing up, clinicians could assess their disease risk and progression or their risk of falling.
Excessive loading in the knee joint puts patients at higher risk of developing osteoarthritis, for instance, but clinicians can’t easily access this information. The disease is typically diagnosed after symptoms appear, even though intervention could happen much earlier.
“Prevention is still not as embraced as it should be,” said Pamela Toto, PhD, professor of occupational therapy at the University of Pittsburgh, who also was not involved in making the app. “If we could tie the technology to intervention down the road, that could be valuable.”
A version of this article first appeared on Medscape.com.
Stanford (Calif.) University’s human performance lab sits next to its physical therapy clinic, so orthopedic surgeons often stop by to request biomechanical analyses for their patients, such as athletes with repeat injuries.
“It would take us days to analyze the data, so we would only do it a handful of times per year,” said Scott Uhlrich, PhD, director of research at the lab.
Now an app can do the job in less than 10 minutes.
The motion-capture app, created by Dr. Uhlrich and fellow bioengineers at Stanford, could help clinicians design better interventions to ward off mobility problems and speed recovery. It could also help researchers fill huge knowledge gaps about human mobility.
It’s currently available free for research and educational use. Model Health, a startup affiliated with the Stanford researchers, provides licenses for commercial use and clinical practice.
Here’s how it works. Footage of human movement, recorded by two smartphones, gets uploaded to the cloud, where an algorithm identifies a set of points on the body. The app relies on computer vision algorithms, a form of AI that trains computers to “understand” visual data – in this case, a person’s pose.
Next, the app quantifies how the body is moving through three-dimensional space. Musculoskeletal system models reveal insights into that movement, such as the angle of a joint, the stretch in a tendon, or the force being transferred through the joints.
“These are the quantities that relate to injuries and disease,” said Dr. Uhlrich, co-author of a study introducing the app. “We need to get to those quantities to be able to inform medical research and eventually clinical practice.”
The conventional approach to getting this kind of analysis requires special expertise and costs $150,000. By contrast, the app is free and easy to use.
It “democratizes” human movement analysis, said senior study author Scott Delp, PhD, professor of bioengineering and mechanical engineering at Stanford. The researchers hope this will “improve outcomes for patients across the world.”
‘Endless opportunities’
A lot about human mobility remains mysterious.
In aging adults, researchers can’t say when balance starts to degrade or by how much every year. They’re also still unraveling how sports injuries occur and how degenerative joint diseases like arthritis progress.
“We don’t really understand the onset of a lot of things, because we’ve just never measured it,” Dr. Uhlrich said.
OpenCap could help change that in a big way. Although biomechanics studies tend to be small – just 14 participants, on average – the app could allow for much larger studies, thanks to its lower cost and ease of use. In the study, the app collected movement data on 100 participants in less than 10 hours and computed results in 31 hours – an effort that would otherwise have taken a year.
“Studies of hundreds will be common, and thousands will be feasible, especially if assessments are integrated into clinic visits,” Dr. Uhlrich said.
About 2,600 researchers around the world are already using the app, according to Dr. Uhlrich. Many had never created a dynamic simulation before.
“The opportunities here are endless,” said Eni Halilaj, PhD, an assistant professor of mechanical engineering at Carnegie Mellon, Pittsburgh, who was not involved in creating the app. That’s especially true for “highly heterogeneous conditions that we have not been able to fully characterize through traditional studies with limited patients.”
In one case, researcher Reed Gurchiek, a former Stanford postdoc and current professor at Clemson (S.C.) University, used the app to study hamstring strain injuries during sprinting and found that these muscles lengthen faster during acceleration, compared with running at a constant speed.
“This aligns with the higher observed injury rates when athletes are accelerating,” Dr. Uhlrich explained. “Varied-speed sprinting studies are not possible in the lab, so this was really enabled by OpenCap’s portability.”
Movement as a biomarker
The researchers are already using the app to build new tools, including metrics to identify risk for anterior cruciate ligament injury in young athletes and to measure balance.
Someday, the technology could augment annual physicals, establishing movement as a biomarker. By having patients perform a few movements, like walking or standing up, clinicians could assess their disease risk and progression or their risk of falling.
Excessive loading in the knee joint puts patients at higher risk of developing osteoarthritis, for instance, but clinicians can’t easily access this information. The disease is typically diagnosed after symptoms appear, even though intervention could happen much earlier.
“Prevention is still not as embraced as it should be,” said Pamela Toto, PhD, professor of occupational therapy at the University of Pittsburgh, who also was not involved in making the app. “If we could tie the technology to intervention down the road, that could be valuable.”
A version of this article first appeared on Medscape.com.