Is It Possible to Differentiate the Impact of Pediatric Monophasic Demyelinating Disorders and Multiple Sclerosis After a First Episode of Demyelination? - Longitudinal Registration and Shape Modeling - page 42

Table 1. z-scores computation for the different groups of subjects: normal controls (NC
group), patients with multiple sclerosis (MS group), with ADEM and with other monophasic
diseases (monophasic group) - z-scores significantly different from 0 are in bold (one sample
t-test, p<0.01, corrected for multiple comparisons)

Metrics NC MS ADEM Mono

Phasic

Brain volume z-score -0.35

(±0.83)

-0.41
(±1.05)

-0.20
(±0.87)

-0.29
(±1.24)

Brain growth rate z-score -1.07

(±3.76)

-2.27
(±1.16)

-1.95
(±1.66)

-1.81
(±1.42)

Normalized Thalamus z-score 0.08

(±1.06)

-0.38
(±0.95)

-1.16
(±2.20)

-0.008
(±1.19)

Normalized thalamus growth rate z-score -0.75

(±18.08)

-10.17
(±9.98)

-11.72
(±5.74)

-3.44
(±4.14)

As the NC subjects had just two visits, the growth trajectories were less accurate.
That is why the standard deviations for the growth rate z-scores were higher. Howev-
er, we did not find any significant differences between the two healthy control groups
(z-scores of the NC group were not significantly different from zero), indicating that
there was no significant bias associated with the HSC scanner and permitting us to
confidently utilize the NIHPD as our healthy comparator group to the patient cohort.
The model-predicted brain volume and the normalized thalamus at the time of the
first attack did not differentiate the patient groups (cf. first and third rows of Table 1)
with either the NIHPD normal controls (the z-score magnitudes for all groups were
small and not significantly different from zero) or the patient groups from each other
(2 samples t-test, not significant p-value). That could mean that the disease has not yet
measurably impacted the brain or the thalamus volume at this stage.
The model-predicted growth rate z-scores were significantly different from zero for
the three patient groups (cf. second and fourth rows of Table 1). A negative z-score
indicates that the growth rate is smaller for the patients than for the normal controls.
For the brain, the mean model-predicted growth rate z-scores were all negative but
not significantly different between the three patients groups. For the thalamus, the
mean growth rate z-scores were all negative but significantly smaller for MS than
MONO and for ADEM than MONO. The difference between MS and ADEM was not
significant.
The standard deviation for the MS normalized thalamus z-scores is very large
( ±9.98 ) demonstrating that the MS group is very heterogeneous at this stage of the
disease. When we used the normalized thalamus growth rates for classification using
a linear discriminant analysis with cross-validation to differentiate MS vs monophasic
illness [18], 93% of monophasic subjects were correctly classified (monophasic or
ADEM), but only 13% of the MS subjects were correctly classified.