System and method for image data analysis for objective analysis of joint motion data
The disclosure related to methods and systems for image data analysis for objectively analyzing joint motion data. In one embodiment, the methods and system disclosed herein use automated methods to capture image data, extract motion data, and determine an objective score for one or more joints in one or more motions. The system and methods disclosed herein comprise a motion capture system with one or more image capture devices to obtain motion data for a subject and using one or more regression models and analysis techniques, various objective scores can be generated related to the subject's joint motion.
1 . A motion analysis system for image data analysis and objective evaluation of joint motion, the system comprising:
a motion capture system;
one or more image data capture devices designed to capture image data;
a database storing joint measurements associated with a subset of a population, the joint measurements including first measurements associated with a first type of movement associated with a first type of joint and second measurements associated with a second type of movement associated with the first type of joint;
a memory storing programmable instructions;
at least one processor communicatively connected to the memory and designed to execute the programmable instructions to cause the motion analysis system to:
determine that a first type of regression model best fits the first measurements associated with the first type of movement associated with the first type of joint;
determine that a second type of regression model best fits the second measurements associated with the second type of movement associated with the first type of joint, the first type of regression model different than the second type of regression model;
extract motion data from the image data of the subject, the motion data indicating the first type of movement associated with the first type of joint of the subject;
select, based in part on the first type of movement, the first type of regression model for analyzing the motion data using a regression model selection module; and
calculate a score based on an output of the first type of regression model compared to the motion data of a subject using a scoring module.
2 . The motion analysis system of claim 1 wherein the motion capture system can be provided in the form of a markerless motion capture system.
3 . The motion analysis system of claim 1 wherein the motion data can include kinematic data, kinetic data, or a combination thereof.
4 . The motion analysis system of claim 1 further comprising a user interface, wherein an output score is generated from the scoring module and generated on the user interface.
5 . The motion analysis system of claim 1 wherein the first type of regression model is at least one of a linear, multi-linear, logistic, polynomial, quantile, principle components, partial least squares, support vector, Poisson, quasi, Cox, Tobit Lasso, Beta, Burr, log-normal, Gamma, Logarithmic, or Ridge regression model.
6 . The motion analysis system of claim 1 , wherein the first type of joint is a knee, the first type of movement is a squat, and the second type of movement is a jump.
7 . The motion analysis system of claim 1 , wherein the first type of regression model is used to create a regression formula derived from a slope of the first measurements.
8 . The motion analysis system of claim 1 wherein the output of the scoring module includes a visual representation of an output score for the subject.
9 . A motion analysis method for analyzing image data and scoring joint motion, the method comprising the steps of:
capturing image data;
extracting motion data of a subject performing a specific movement from the image data;
defining a population against which to compare the subject;
selecting a regression formula that corresponds to a variance in measurements of the specific movement among members of the population; and
generating, based on an output of the regression formula, a score for the subject that indicates performance of the specific movement by the subject relative to the population.
10 . The motion analysis method of claim 9 further comprising the step of:
generating a first variable ratio for a first measurement of a first joint of the subject for the specific movement.
11 . The motion analysis method of claim 10 , wherein the first variable ratio is generated using at least one of a linear, multi-linear, logistic, polynomial, quantile, principle components, partial least squares, support vector, Poisson, quasi, Cox, Tobit Lasso, Beta, Burr, log-normal, Gamma, Logarithmic, or Ridge regression model.
12 . The motion analysis method of claim 10 , wherein the first variable ratio is generated in accordance with the following:
Variable
Dependent
↘
Y
i
=
β
0
↘
Population
Y
intercept
+
β
1
↘
Population
Slope
Coefficient
X
i
↓
independent
variable
︸
Linear
component
+
ε
i
↙
Random
Error
term
︸
Random
Error
component
.
13 . The motion analysis method of claim 10 , wherein the first variable ratio is generated in accordance with the following:
[
c
1
c
2
c
3
]
=
[
∑
k
=
1
n
f
1
2
(
x
k
)
∑
k
=
1
n
f
1
(
x
k
)
f
2
(
x
k
)
∑
k
=
1
n
f
1
(
x
k
)
f
3
(
x
k
)
∑
k
=
1
n
f
2
(
x
k
)
f
1
(
x
k
)
∑
k
=
1
n
f
2
2
(
x
k
)
∑
k
=
1
n
f
2
(
x
k
)
f
3
(
x
k
)
∑
k
=
1
n
f
3
(
x
k
)
f
1
(
x
k
)
∑
k
=
1
n
f
3
(
x
k
)
f
2
(
x
k
)
∑
k
=
1
n
f
3
2
(
x
k
)
]
-
1
·
[
∑
k
=
1
n
f
1
(
x
k
)
·
y
k
∑
k
=
1
n
f
2
(
x
k
)
·
y
k
∑
k
=
1
n
f
3
(
x
k
)
·
y
k
]
.
14 . The motion analysis method of claim 10 further comprising the step of:
generating a second variable ratio for a second measurement of the first joint of the subject for the specific movement.
15 . The motion analysis method of claim 13 further comprising the step of:
combining the first variable ratio with at least the second variable ratio to generate a first composite score for the first joint of the subject for the specific movement.
16 . The motion analysis method of claim 14 further comprising the step of:
combining the first composite score with at least a second composite score to generate an overall score for the first joint of the subject.
17 . The motion analysis method of claim 15 further comprising the step of:
combining the overall score for the first joint of the subject with an overall score for at least a second joint of the subject to generate an overall subject score.
18 . The motion analysis method of claim 9 wherein the motion data can include kinematic data, kinetic data, or a combination thereof.
19 . A motion analysis system for image data analysis and objective evaluation of joint motion, the system comprising:
a markerless motion capture system;
one or more image data capture devices designed to capture image data of a subject;
a first database storing first joint measurements associated with a subset of a population, the first joint measurements associated with a first type of movement associated with a first type of joint;
a second database storing second joint measurements associated with the subset of the population, the second joint measurements associated with a second type of movement associated with the first type of joint;
a memory storing programmable instructions;
at least one processor communicatively connected to the memory and designed to execute the programmable instructions to cause the motion analysis system to:
extract first motion data from the image data, the first motion data indicating the first type of movement associated with the first type of joint of the subject;
select, based on a first variance in the first joint measurements, a first regression model for analyzing the first motion data using a regression model selection module, wherein a first regression model output is generated from the first regression model;
calculate a first score associated with a subject using a scoring module, wherein the first score is calculated based on a comparison of the first motion data relative to the first regression model output;
extract second motion data from the image data, the second motion data indicating the second type of movement associated with the first type of joint of the subject;
select, based on a second variance in the second joint measurements, a second regression model for analyzing the second motion data using the regression model selection module, wherein a second regression model output is generated from the second regression model; and
calculate a second score associated with a subject using the scoring module, wherein the second score is calculated based on a comparison of the second motion data relative to the second regression model output.
20 . The system of claim 19 , wherein the programmable instructions further cause the motion analysis system to determine an overall score for the first type of joint of the subject based on the first score and the second score.