TY - CHAP
T1 - Statistical learning in rooms under transcranial magnetic stimulation
AU - Hernandez-Perez, Heivet
AU - McAlpine, David
AU - Monaghan, Jessica
PY - 2020
Y1 - 2020
N2 - In the auditory system, adaptation to sound intensity can
operate over multiple time scales and contributes to statistical learning (Dean
et al., 2005, 2008; Simpson et al., 2014). Moreover, efferent feedback from the
cortex to the inferior colliculus has been directly linked to learning the
longer-term statistical structure of sound environments so that this
information can be used when a familiar environment is re-encountered (Robinson
et al., 2016). Brandewie and Zahorik (2010) described a psychoacoustic task
that demonstrated speech perception advantages of re-exposure to familiar environments.
They found that prior exposure to room statistics (Reverberation Times; RTs) in
the form of a “carrier phrase” spoken in a simulated room significantly improved
participants’ masked speech understanding for speech spoken in the same room.
Here, we aimed to understand: 1) how real room RTs contribute to the statistical
learning of room acoustics and 2) how this learning may be affected when the
inferior frontal gyrus (IFG), involved in the expression of auditory
statistical learning (Karuza et al., 2013), is temporarily impaired using
repetitive transcranial magnetic stimulation (rTMS). Following the methods of
Bradewie & Zahorik (2010; 2013), phrases from the Coordinated Response Measurement
(CRM) corpus were presented in noise to ten participants with normal hearing.
The CRM phrases had different durations: 0, 0.2357, 0.5715 and 0.8485 s (e.g.,
“Green Five now”, “Go to Green Five Now”, “Baron Go to Green Five now” and
“Ready Baron Go to Green Five now”, respectively). Carrier phrases were
convolved with impulse responses from 3 real rooms (Room 1: RT60 = 0.46
s; Room 2: RT60 = 0.96 s; Room 3: RT60 = 2.42 s) and presented in free
sound-field conditions in an anechoic chamber using a ring of 41 loudspeakers. Similar
to Brandewie & Zahorik (2013) participants perceived target phrases e.g.,
‘’Green Eight”, more accurately when the carrier phrases had longer duration (containing
more statistical information from the rooms presented). It was also observed
that rTMS disrupts this learning across all carrier phrases and rooms tested (p<
0.05). In addition, we found that the carrier length effect is only observed
when Room 1 is presented in the mixture of rooms. Our results revealed that IFG
is involved in learning room statistics, which may reflect adaptation through
feedback mechanisms from IFG to the auditory pathway. Moreover, our data
suggest that ecologically relevant RTs such as that of Room 1 enable a
perceptual anchoring that allows learning room statistics in less commonly
encountered environments.
AB - In the auditory system, adaptation to sound intensity can
operate over multiple time scales and contributes to statistical learning (Dean
et al., 2005, 2008; Simpson et al., 2014). Moreover, efferent feedback from the
cortex to the inferior colliculus has been directly linked to learning the
longer-term statistical structure of sound environments so that this
information can be used when a familiar environment is re-encountered (Robinson
et al., 2016). Brandewie and Zahorik (2010) described a psychoacoustic task
that demonstrated speech perception advantages of re-exposure to familiar environments.
They found that prior exposure to room statistics (Reverberation Times; RTs) in
the form of a “carrier phrase” spoken in a simulated room significantly improved
participants’ masked speech understanding for speech spoken in the same room.
Here, we aimed to understand: 1) how real room RTs contribute to the statistical
learning of room acoustics and 2) how this learning may be affected when the
inferior frontal gyrus (IFG), involved in the expression of auditory
statistical learning (Karuza et al., 2013), is temporarily impaired using
repetitive transcranial magnetic stimulation (rTMS). Following the methods of
Bradewie & Zahorik (2010; 2013), phrases from the Coordinated Response Measurement
(CRM) corpus were presented in noise to ten participants with normal hearing.
The CRM phrases had different durations: 0, 0.2357, 0.5715 and 0.8485 s (e.g.,
“Green Five now”, “Go to Green Five Now”, “Baron Go to Green Five now” and
“Ready Baron Go to Green Five now”, respectively). Carrier phrases were
convolved with impulse responses from 3 real rooms (Room 1: RT60 = 0.46
s; Room 2: RT60 = 0.96 s; Room 3: RT60 = 2.42 s) and presented in free
sound-field conditions in an anechoic chamber using a ring of 41 loudspeakers. Similar
to Brandewie & Zahorik (2013) participants perceived target phrases e.g.,
‘’Green Eight”, more accurately when the carrier phrases had longer duration (containing
more statistical information from the rooms presented). It was also observed
that rTMS disrupts this learning across all carrier phrases and rooms tested (p<
0.05). In addition, we found that the carrier length effect is only observed
when Room 1 is presented in the mixture of rooms. Our results revealed that IFG
is involved in learning room statistics, which may reflect adaptation through
feedback mechanisms from IFG to the auditory pathway. Moreover, our data
suggest that ecologically relevant RTs such as that of Room 1 enable a
perceptual anchoring that allows learning room statistics in less commonly
encountered environments.
UR - https://aro.org/wp-content/uploads/2020/02/2020-Abstracts_1-21-20-Web.pdf
UR - https://aro.org/meetings/aro-midwinter-meeting-archives/
M3 - Conference abstract
T3 - The Abstracts of the Association for Research in Otolaryngology
SP - 479
EP - 480
BT - Conference Program of the 43rd Annual MidWinter Meeting
PB - Association for Research in Otolaryngology (ARO)
CY - Nashville, Tennessee
T2 - Annual ARO MidWinter Meeting (43rd : 2020)
Y2 - 25 January 2020 through 29 January 2020
ER -