This is my PhD defense (2017-01-05) about structural and functional features of the auditory cortex in musicians with absolute pitch using MRI data.

1. PhD examination
Myeloarchitecture and Resting State Functional
Connectivity of Auditory Cortex in Musicians with
Absolute Pitch
Seung-Goo (“SG”) Kim
Max Planck Institute for Human Cognitive and Brain Sciences
5 Jan 2017
Institute for Psychology
Faculty of Biosciences, Pharmacy, and Psychology
University of Leipzig
1

2. Overview
• Introduction: Absolute pitch (AP)
• Study #1: In-vivo cortical myelin mapping
• Study #2: Resting-state functional connectivity
• Discussion: Distributed processing of AP
2

3. Introduction
Absolute pitch
3

4. What is absolute pitch (AP)?
• Rare ability to recognize the name pitch without an external
reference [1].
• Related to early musical experience (4~7 year-old [2]) and a
certain training method (“fixed-Do” than “movable-Do”) [3].
• Genetic factor: more APs in AP’s family [2];
Europeans (7%) vs. East-Asian (30%) in musicians [4].
• Sometimes make mistakes in octave recognition [1,4]:
good at pitch chroma rather than pitch height.
[1] Miyazaki, 2004. [2] Baharloo et al., 1998. [3] Willson et al., 2012. [4] Miyazaki et al., 2012.
4

5. What is pitch chroma?
• “C3”
• Pitch chroma:
Alphabetical label of 12
divisions of an octave
(C, C#, D, D#, …)
• Pitch height:
Octave index
5
B2 C3 D3 E3 F3 G3 A3 B3 C4
C#3 D#3 F#3 G#3 A#3 C#4A#2
GarageBand, (C) Apple, Inc.
The 3rd octave The 4thThe 2nd

6. AP-test confusion matrix
Kim & Knösche, 2016. 6

7. • Perceptual process
Categorization of pitch chroma
• Associative process
Verbal/non-verbal labeling
Dichotomy of AP recognition
Levitin & Rogers, 2005. 7
Categorization
Labeling

8. • Temporal lobe (perceptual proc.)
leftward asymmetry of planum temporale (PT)
area [1]; higher activation in the left PT [2,3].
• Frontal lobe (associative proc.)
thicker inferior frontal gyrus (IFG) [4]; higher
activation in right dorsolateral prefrontal cortex
(DLPFC) [3-5].
• Frontotemporal connectivity
phase synchrony [6]; fractional anisotropy of
superior longitudinal fasiculus (SLF) [7].
Previous neuroimaging findings
DLPFC
HG
PT
http://clipartbest.com/
[1] Keenan et al., 2001. [2] Ohnishi et al., 2001. [3] Willson et al., 2009. [4] Dohn et al., 2015.
[5] Zatorre et al., 1998. [6] Elmer et al., 2015. [7] Oechslin et al., 2009.
8
HG: Heschl’s gyrus

9. But how does the AP work?
• How does the cortical regions and the connectivity
between them contribute to the AP recognition?
> Multimodal imaging describing microscopic
architecture might give us better ideas to construct
a framework to explain how AP works.
9

10. Myeloarchitecture of auditory
cortex in musicians with AP
Study #1: Relaxometry at 7-T
10Kim & Knösche, 2016, Human Brain Mapping.

11. Intracortical myelination
• Increases information transfer speed
(but not very much for short-range
connections in gray matter)
• May reduce cross-talk between nearby
axons, thus increasing specificity
• Prevents neuroplasticity after critical
period of neurodevelopment (McGee
et al., 2005)
Vogt, 1903
Cyto-
architecture
Myelo-
architecture
11

12. Research questions
• Does the degree of cortical myelination in auditory
cortex (related to the perceptual subprocess of AP)
correlate with the acuity of AP?
• Is myelo-correlate of AP differentiable from that of
the other psychoacoustic characteristics (e.g.,
frequency discrimination threshold)?
12

13. Participants & imaging
• Participants
• 8 AP musicians & 9 non-AP musicians categorized
based on behavioral test of AP
• Cortical myelin mapping (7-T)
• MP2RAGE sequence for quantitative mapping of
longitudinal relaxation rate (qR1), which strongly
correlates with myelin content in the cortex
13

14. In-vivo cortical myelin mapping
Kim & Knösche, 2016. 14

15. General linear model
• x was either:
• Absolute pitch group index
• Absolute pitch score (APS) = 1 - scaled ([0,1]) error,
• Frequency discrimination threshold (FDT) = -log10(Fm/F0)
15
qR1
= β0
+ageβ1
+ sexβ2
+ethnicityβ3
+ xβ4
+ε

16. Absolute pitch score
HG
PT
STS
PP
STG
Frequency
discrimination threshold
STG
HG
PT
STS
PP
Kim & Knösche, 2016. 16
Concave
Convex

17. Pitch chroma vs. pitch height
Warren et al., 2003.17
Planum
polare
Planum
temporale
Primary auditory
cortex (overlap)
Chroma
Height
Kim & Knösche, 2016.

18. Dual pathway hypothesis (Rauschecker et al., 1995)
• Ventral (“what”): vs. dorsal (“where”) auditory pathways
• Spatially separated sensitivity to non-spatial (i.e. pitch) vs.
spatial (i.e. location) information
Barret & Hall, 2006. 18

19. • Absolute recognition of pitch
chroma (rather than height)
• Cortical myelin can increase
specificity and suppress
neuroplasticity
• Dual pathway hypothesis: non-
spatial information (e.g.
phoneme) via ventral pathway.
• Conjecture: Categorization of
pitch chroma along the ventral
pathway?
Categorical processing & ventral pathway
DLPFC
PT
HG
http://clipartbest.com/
19
PP
VLPFC

20. Summary of Study #1
• We found: greater cortical myelin in the right
planum polare (PP) in musicians with than without
absolute pitch.
• We think: the right PP and its connection (ventral
pathway) may be important in chromatic
categorization.
20

21. Resting-state functional connectivity of
the ventral pathway in musicians with AP
Study #2: Resting-state fMRI at 3-T
21Kim & Knösche, 2016, In preparation.

22. Spontaneity of AP
• Unintended recognition of pitch chroma of
environmental sound (e.g., cluttering of glasses, siren
of ambulance) [1]
• Uncontrollable and instant perception (e.g., Stroop-
like effect for incongruent pitch and pitch name [2,3])
• Spontaneous & automatic processing of AP could be
reflected in BOLD signal during resting
[1] Miyazaki, 2004. [2] Itoh et al., 2005. [3] Schultz et al., 2013.
22

23. Research questions
• Is the planum polare (PP) strongly
coupled with other cortical regions
that belong to the ventral pathway
in musicians with AP?
• What about the planum temporale
(PT) and the dorsal pathway?
23
HG
DLPFC
PT
http://clipartbest.com/
VLPFC
PP
HG: Heschl’s gyrus

24. Participants & imaging
• Same participants: 8 AP musicians & 9 non-AP musicians
• Whole-brain functional images at 3-T
• Echo-planer imaging; 2.3-mm isovoxel; TR = 1.4 s; for
9.8-min resting
• Anatomical image at 3-T
• MPRAGE; 1-mm isovoxel
24

25. Zero-lag cross-correlation
Kim & Knösche, In prep.25

26. Kim & Knösche, In prep.26
• Frequency-
dependent
connectivity
(regardless of
time lag)
• Averaged into
9 frequency bins
with 50% overlap:
[0, 20],
[10, 30],
[20, 40], …,
[90, 110] mHz
Cross-coherence

27. General linear model
• FC was either:
• cross-correlation
• cross-coherence
• x was either:
• AP-group index (AP)
• AP score (APS)
27
FC = β0
+ motionβ1
+ ethnicityβ2
+ xβ3
+ ε

28. AP/APS effect in correlation
28
Kim&Knösche,inprep.

29. AP/APS effect in coherence
Kim&Knösche,inprep.
29

30. Greater RSFC of the right PP
• Ipsilateral STG, contralateral PP/STG:
adjacent/homologous regions via bilateral
ventral streams [1]
• Bilateral STS: specialized auditory object
recognition; e.g. voice recognition [2]
• Left IFG: verbal labeling of pitch chroma [3]
• Bilateral ACC, medial SFG: default-mode
network [4]
[1] Upadhyay et al., 2008 [2] Belin et al., 2004. [3] Wengenroth et al., 2013. [4] Smith et al., 2009.
30

31. What about the left dorsal pathway?
31
• Leftward asymmetry of planum temporale (PT) [1]
• Greater rCBF/BOLD in the left PT during passive
listening [2,3]
• Phase synchrony between PT and dorsal lateral
prefrontal cortex (DLPFC) [4], superior
longitudinal fasiculus [5]
★But not conclusive: the other studies also found
volume/thickness/FA increase in the right STG/
ILF [6,7].
[1] Schlaug et al., 1995. [2] Ohnishi et al., 2001. [3] Wilson et al., 2009. [4] Elmer et al., 2015.
[5] Oechslin et al., 2009. [6] Wengenroth et al., 2013. [7] Dohn et al., 2015.

32. -log10minP
RSFC seeding from planum temporale
Kim & Knösche, In prep.32

33. Summary of Study #2
• We found: greater resting-state functional
connectivity (RSFC) of the the right planum polare
(PP) with the anteroventral auditory network and
the anteromedial network in musicians with
absolute pitch (AP).
• We think: the ventral auditory pathway via the right
PP may be crucial in AP processing (presumably in
chromatic categorization).
33

34. Discussion
Distributed processing of AP
34

35. Main findings
• Greater cortical myelin in the right planum
polare (PP) in musicians with absolute pitch
• Greater RSFC of the right PP in
• the right STG, the left PP/STG,
• the bilateral STS,
• the left IFG, and
• the bilateral ACC/SFG
• No difference in the left planum temporale.
35

36. Distributed processing of AP
Left dorsal pathway?
(spatial processing?)
Kim & Knösche, In prep.36

37. Future directions
• Topological organization for pitch chroma in APs
• Model-based encoding fMRI analysis
• Dynamics of frontotemporal network
• Effective connectivity in multimodal data
(fMRI+MEG)
37

38. Thank you for attention!
38