Full metadata record

DC Field Value Language
dc.contributor.authorLee, Gyuhwan-
dc.contributor.authorJeong, Iljoo-
dc.contributor.authorPark, Keonhyeok-
dc.contributor.authorLee, Kangmin-
dc.contributor.authorLee, Seungchul-
dc.contributor.authorChoi, Ji Hyun-
dc.contributor.authorPark, Choonsu-
dc.date.accessioned2024-01-12T02:44:34Z-
dc.date.available2024-01-12T02:44:34Z-
dc.date.created2023-10-19-
dc.date.issued2023-10-25-
dc.identifier.urihttps://pubs.kist.re.kr/handle/201004/76355-
dc.description.abstractTo understand the neural mechanisms underlying social behaviors, pinpointing the speaker within a group is essential. In this study, we employ the mouse, a popular animal model in neuroscience, to identify vocalizations within group conditions. A primary challenge in localizing mouse vocalizations is the sensor inter-distance requirement―less than half of 1.9 cm―to satisfy the Nyquist condition. However, reducing the sensor dimensions inadvertently diminishes sensitivity. To address this, we designed an optimal sparse sensor arrangement and adopted the functional beamforming method. Two parameters were our primary focus: the maximum side-lobe level (MSL) for estimating the image distortion and the half-power bandwidth (HPB) ofthe main lobe, indicating the spatial resolution. First, we optimized a 2D arrangement fortwelve sensors through a genetic algorithm, targeting a minimized MSL whilst preserving the main lobe's HPB. Through simulations and experiments with various source locations and frequencies, we observed that source localization improves with increasing frequency and decreasing target distance. Intriguingly, frequency and distance had a negligible impact on the MSL stemming from an identical main lobe. Secondly, we enhanced the image resolution by adjusting the eigenvalues linked to the beam pattern's side-lobe distribution via eigenvalue analysis, termed functional beamforming. This approach, compared with conventional beamforming, yielded a 30% enhancement in HPB and a decrease of 10 dB in MSL, thus increasing the spatial resolution for a given sensor array system. Lastly, we presented a proof-of-principle by applying this methodology to socially engaged mice. A novel CBRAIN technique was employed to enable concurrent recording of neural and vocal signals from freely interacting mouse groups. We expect that these advancements will contribute to unveil the intricate neural mechanisms underpinning vocalizations, furthering our understanding ofthe neural origins giving rise to social behavior.-
dc.languageKorean-
dc.publisher한국물리학회-
dc.titleTracking mouse USV origins with functional beamforming on sparse arrays-
dc.typeConference-
dc.description.journalClass2-
dc.identifier.bibliographicCitation2023 KPS Fall Meeting-
dc.citation.title2023 KPS Fall Meeting-
dc.citation.conferencePlaceKO-
dc.citation.conferencePlaceChangwon Exhibition Convention Center-
dc.citation.conferenceDate2023-10-24-
dc.relation.isPartOf2023 KPS Fall Meeting-
Appears in Collections:
KIST Conference Paper > 2023
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML

qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE