Solutions
Pricing
Log in
Teams
Get started
Log in
Get started
A Classifier for the Categorization of Thumb Carpometacarpal Osteoarthritis
AM
by
Amy Morton
·
1 view
·
Updated Jul 15, 2020
Share
Embed
Create your own
Audio Transcript
Auto-generated
00:01 - 00:03
This is Amy Morton On behalf of my co authors,
00:04 - 00:07
Brown Medical School, Rhode Island Hospital in Stanford School of
00:07 - 00:09
Medicine would like to present a classifier for the categorization
00:10 - 00:12
of thumb carpal metacarpal osteoarthritis.
00:14 - 00:16
Some seems you A is a disease of bone cartilage.
00:17 - 00:20
Symptoms include pain and impairment in the base of the
00:20 - 00:24
thumb. Using CT scans, we can quantify bone pathology and
00:25 - 00:26
look at relative cinematics of the joint.
00:29 - 00:29
Seems you A.
00:29 - 00:34
Is a disease that progresses and has highly individualized rates
00:34 - 00:40
of progression, one of the metrics associated with away Austro
00:41 - 00:41
fight volume.
00:42 - 00:44
Recently, we have come up with a method to quantify
00:45 - 00:48
Oster fight volume, using that as a surrogate for a
00:48 - 00:49
metric of disease progression.
00:52 - 00:56
Another factor associated with away is joint model alignment or
00:56 - 01:02
subluxation. Using articular surface curvature based coordinate systems, we can
01:02 - 01:05
compute relative positions of the metacarpal from the trapeze IAM
01:06 - 01:08
and analyzed these in three D or in anatomical directions.
01:18 - 01:21
Using longitudinal data captured over a course of six years,
01:22 - 01:24
with patients presenting with osteoarthritis as well as an age
01:25 - 01:30
matched cohort, we can investigate and classify away using these
01:30 - 01:35
metrics. Competing Austria fight balling for every individual at every
01:35 - 01:35
time point.
01:37 - 01:39
We can then use absolute magnitude and the rate of
01:40 - 01:43
progression as specified by linear regression within a subject.
01:45 - 01:49
To then classify our away individuals into an advanced or
01:49 - 01:56
stable cohort using linear mixed models with random interception slope,
01:57 - 02:02
weaken. Then analyze the regression of the advanced away category
02:03 - 02:06
versus the that of stable away and compare that as
02:06 - 02:07
well to the control covert.
02:09 - 02:14
Note the significance in both starting baseline value of our
02:14 - 02:14
study as well.
02:15 - 02:18
A ce the slope between advanced away as well, a
02:18 - 02:22
stable and control using the same metrics.
02:22 - 02:26
We can then investigate the differences as observed in dorsal
02:26 - 02:27
subluxation over time.
02:30 - 02:35
Advanced category endorses subluxation as measured in the flexion position
02:35 - 02:38
had a very significant difference in both.
02:39 - 02:44
Intercept in both are starting value of dorsal subluxation and
02:45 - 02:46
the rate of dorsal subluxation.
02:47 - 02:51
Over time, similar trends were observed in six of 11
02:51 - 02:55
of R C T positions, grasps, grass, loaded, pinch loaded
02:55 - 02:56
induction and abduction.
03:00 - 03:05
This classifications metrics and our initial findings can be important.
03:06 - 03:09
Stepping stones in a heterogeneous point patient population.
03:11 - 03:11
Thank you
See full transcript