VU Research Portal

(1)

VU Research Portal

Speech recognition in cochlear implant users

Kaandorp, M.W.

2018

document version

Publisher's PDF, also known as Version of record

Link to publication in VU Research Portal

citation for published version (APA)

Kaandorp, M. W. (2018). Speech recognition in cochlear implant users: Outcome measures and the role of

linguistic skills.

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal ? Take down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

E-mail address:

vuresearchportal.ub@vu.nl

(2)

516088-L-bw-Kaandorp 516088-L-bw-Kaandorp 516088-L-bw-Kaandorp 516088-L-bw-Kaandorp Processed on: 2-1-2018 Processed on: 2-1-2018 Processed on: 2-1-2018

(3)

Processed on: 2-1-2018 PDF page: 11PDF page: 11PDF page: 11PDF page: 11

(4)

(5)

General introduction

(6)

(7)

WKHVHOLQJXLVWLFPHDVXUHVFRPSDUHGWRWKHLQÁXHQFHRIWZRFRJQLWLYHPHDVXUHVRQVSHHFK UHFRJQLWLRQLQ&,XVHUV)LQDOO\ZHH[DPLQHPHDVXUHVRIOLQJXLVWLFVNLOOVLQWKHDXGLWRU\ GRPDLQDQGWKHUHODWLRQZLWKVSHHFKUHFRJQLWLRQLQ&,XVHUVEHFDXVHWKH\FDQSURYLGH more insight in the interaction between these linguistic skills and speech understanding as it occurs in daily life.

OUTCOME MEASURES OF SPEECH RECOGNITION IN CI USERS

Speech-recognition ability can be measured in several ways and it is important to understand advantages and disadvantages of those measures. Often phoneme scores RUZRUGVFRUHVRIPRQRV\OODEOHZRUGVLQTXLHWDUHPHDVXUHG,QDGGLWLRQDOVRVHQWHQFH UHFRJQLWLRQLQTXLHWLVRIWHQXVHG:KHUHZRUGUHFRJQLWLRQZLWKSKRQHPHVFRUHVPDLQO\ UHSUHVHQWVKRZZHOODOLVWHQHUSHUFHLYHVODQJXDJHVSHFLÀFGHWDLOVRIWKHLQSXWVLJQDO ERWWRPXSVHQWHQFHUHFRJQLWLRQLVDVVXPHGWRGHPDQGPRUHWRSGRZQSURFHVVLQJ DQGWKHXVHRIOLQJXLVWLFNQRZOHGJH:KHQVHQWHQFHVDUHSUHVHQWHGLQQRLVHWKHVHWRS GRZQGHPDQGVLQFUHDVHHYHQPRUHVHHHJ0DWW\VHWDO 7KHGLYHUVLW\LQPHDVXUHVRI&,RXWFRPHPDNHVLWGL඼FXOWWRFRPSDUHUHVXOWVIURPGLරHUHQW VWXGLHVRUFHQWHUV0XOWLFHQWHUVWXGLHVIUHTXHQWO\XVHGSHUFHQWLOHUDQNVIRUSDUWLFLSDQWV SHUFHQWHUWRFRPSDUHSDUWLFLSDQWVEHFDXVHFHQWHUVXVHGGLරHUHQWVSHHFKUHFRJQLWLRQWHVWV DWGLරHUHQWSUHVHQWDWLRQOHYHOVHJ%ODPH\HWDO5HFRJQLWLRQRISKRQHPHVRUZRUGV LQTXLHWLVRIWHQXVHGWRGHÀQH&,RXWFRPHLQVWHDGRIWKHPRUHUHOHYDQWVSHHFKUHFRJQLWLRQ abilities in noise, to determine auditory functioning in daily life listening situations. ,Q'XWFKFOLQLFDOSUDFWLFHVSHHFKUHFRJQLWLRQSHUIRUPDQFHLVSULPDULO\DVVHVVHGZLWK PRQRV\OODEOHV RI WKH FRQVRQDQWYRZHOFRQVRQDQW W\SH &9& LQ TXLHW H[SUHVVHG LQ WKH SHUFHQWDJH RI FRUUHFWO\ UHSURGXFHG SKRQHPHV ,Q DGGLWLRQ VSHHFKUHFRJQLWLRQ performance in adverse conditions is usually assessed with the standard Dutch speech-LQQRLVH WHVW WKDW XVHV VKRUW PHDQLQJIXO VHQWHQFHV VHQWHQFHspeech-LQQRLVH WHVW 6,1 WHVW 3ORPSDQG0LPSHQ7KLVWHVWLQYROYHVDQDGDSWLYHSURFHGXUHWRHVWLPDWHWKHVSHHFK UHFHSWLRQWKUHVKROG657GHÀQHGDVWKHVLJQDOWRQRLVHUDWLR615ZKHUHRQDYHUDJH RIWKHVHQWHQFHVLVUHSHDWHGFRUUHFWO\6LPLODUVHQWHQFHLQQRLVHWHVWVH[LVWLQRWKHU ODQJXDJHVHJWKH+,17WHVW1LOVVRQ6ROLDQG6XOOLYDQIRUDUHYLHZVHH7KHXQLVVHQ 6ZDQHSRHODQG+DQHNRP$OWKRXJKWKH6,1WHVWKDVEHHQZLGHO\XVHGLQUHVHDUFK and clinic, severe hearing impairment or limited linguistic skills (as in children or second ODQJXDJHXVHUVOLPLWWKHDSSOLFDELOLW\RIWKLVWHVW7KH'XWFKGLJLWVLQQRLVH',1WHVW ZDVGHYHORSHGDVDQDOWHUQDWLYHIRUWKH6,1WHVWWKDWPHDVXUHVWKH657IRUGLJLWWULSOHWV

(8)

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රHUHQWSDUWVRIVSHHFK UHFRJQLWLRQHDFKGHPDQGLQJDGLරHUHQWDPRXQWRIWRSGRZQSURFHVVLQJ

MEASURES OF LINGUISTIC SKILLS

(9)

Lexical decision and word naming. ,QWKLVWKHVLVOH[LFDODFFHVVDELOLW\LVDVVHVVHGE\OH[LFDO decision tests (LDT, because these are easy to use and can be applied both in the auditory D/'7DQGYLVXDOPRGDOLW\Y/'77KH/'7LVHVVHQWLDOO\DGLVFULPLQDWLRQWDVNLQZKLFK SDUWLFLSDQWVKDYHWRGHFLGHTXLFNO\ZKHWKHUDWHVWLWHPLVDZRUGRUDSVHXGRZRUGE\ SUHVVLQJDEXWWRQWKDWFRUUHVSRQGVWRWKHFKRLFHPDGH5HDFWLRQWLPHV57VDQGDFFXUDF\ are recorded for both word- and pseudoword responses. Pseudowords are letter strings that obey the orthography and phonology of the test language, but lack meaning. Using well-formed pseudowords is important to ensure that the response is based on lexical access, and not purely on a perceptual process that assesses whether the stimulus looks QRUPDO'H*URRW%HVLGHVWKHWLPHQHHGHGIRUWKHSURFHVVRILQWHUHVWOH[LFDODFFHVV WKH /'7 W\SLFDOO\ DOVR HQFRPSDVVHV WKH WLPH QHHGHG IRU WDVNVSHFLÀF SURFHVVHV OLNH SUHVVLQJWKHEXWWRQ)RUGHWHUPLQLQJWKHUROHRIOH[LFDODFFHVVVNLOOVLQVSHHFKUHFRJQLWLRQ LWLVWKHUHIRUHSUHIHUDEOHWRREWDLQFRQYHUJLQJHYLGHQFHIURPPXOWLSOHWHVWVFI*UDLQJHU -DFREVVRWKDWDPRUHDFFXUDWHHVWLPDWHRIOH[LFDODFFHVVFDQEHREWDLQHGWKDQXVLQJ HLWKHUWHVWRQLWVRZQ:HWKHUHIRUHXVHDVHFRQGOH[LFDODFFHVVWDVNZRUGQDPLQJZKLFK ZHNHHSVLPSOHIRUSRWHQWLDOXVHLQWKHFOLQLF,QWKLVWDVNSDUWLFLSDQWVKDYHWRUHDGZRUGV DORXGDVTXLFNO\DVSRVVLEOHWKXVLWUHTXLUHVSURQRXQFLQJWKHZRUGDIWHUUHFRJQL]LQJLW and does also not provide a pure measure of lexical-access ability. Combining the results of WKHOH[LFDOGHFLVLRQOH[LFDODFFHVVGLVFULPLQDWLRQUHVSRQVHDQGZRUGQDPLQJOH[LFDO DFFHVVVSHHFKSURGXFWLRQWHVWLQWRDVLQJOHFRPSRVLWHPHDVXUHPLJKWSURYLGHDPRUH accurate estimate of lexical-access ability.

,QRUGHUWRGHVFULEHVRPHRIRXUK\SRWKHVHVDQGLQWHUSUHWWKHUHVXOWVZHXVHDVLPSOH modality-independent model for lexical decision. This model was based on established PRGHOVRIOH[LFDODFFHVVLQVSRNHQZRUGUHFRJQLWLRQWKH&RKRUW0RGHOE\0DUVOHQ:LOVRQ 6KRUWOLVW%E\1RUULVDQG0F4XHHQWKH75$&(PRGHOE\0F&OHOODQGDQG(OPDQ HJWKH1HLJKERXUKRRG$FWLYDWLRQ0RGHOE\/XFHDQG3LVRQLDQGRIYLVXDO OH[LFDODFFHVVOLNHWKH0XOWLSOH5HDGRXW0RGHO*UDLQJHUDQG-DFREV,WFRQVLVWVRI WKHIROORZLQJVWDJHV  Activation:LWKWKHRQVHWRILQSXWRUWKRJUDSKLFDOO\VLPLODUOH[LFDOHOHPHQWVLQ YLVXDOOH[LFDOGHFLVLRQHJPRXVHKRXVHPRXWKPXVHPRUVH¬RUSKRQRORJLFDOO\ VLPLODUOH[LFDOHOHPHQWVLQDXGLWRU\OH[LFDOGHFLVLRQHJOLNHOLJKWOLPH¬DUH simultaneously activated and form a cohort.

(10)

 Cohort evaluation :KHQ WKH VWLPXOXV XQIROGV DQG PRUH LQIRUPDWLRQ EHFRPHV DYDLODEOHFRPSHWLWRUZRUGVLQWKHFRKRUWDUHGHDFWLYDWHGLQKLELWLRQ  Decision:KHQDZRUG·VOHYHORIDFWLYDWLRQUHDFKHVDFULWLFDOGLරHUHQFHIURPWKDWRI LWVFRPSHWLWRUV0DUVOHQ:LOVRQDZRUGUHVSRQVHIROORZVRULIDWLPHGHDGOLQH LVSDVVHGEHIRUHDPDWFKLVIRXQGDSVHXGRZRUGUHVSRQVHIROORZVHJ*UDLQJHU DQG-DFREV5DWFOLර*RPH]DQG0F.RRQ 0RGHOVRIDXGLWRU\DQGYLVXDOOH[LFDODFFHVVODUJHO\RYHUODS$QLPSRUWDQWGLරHUHQFHWR keep in mind, however, is that in the auditory domain the stimulus is not presented and LQWHUSUHWHGLQVWDQWDQHRXVO\DVLQWKHYLVXDOPRGDOLW\EXWVHTXHQWLDOO\

(11)

EFFECTS OF LINGUISTIC AND COGNITIVE ABILITIES ON SPEECH RECOGNITION

Relations between visual measures of linguistic and cognitive abilities and speech recognition. :HK\SRWKHVL]HWKDWLQY/'7ORQJHUZRUG57VDUHUHODWHGWRSRRUHUODQJXDJH SURÀFLHQF\DQGWKDWORQJHU57VLQJHQHUDODUHUHODWHGWRSRRUHUVSHHFKUHFRJQLWLRQLQQRLVH Accuracy is expected to be near ceiling in vLDT, when only highly familiar words are used. vLDT could be measured preoperatively to predict speech-recognition performance with a &,)RULQVWDQFH/\[HOOHWDOIRXQGLQDJURXSRI&,FDQGLGDWHVWKDWSUHRSHUDWLYHO\ measured verbal information processing speed and working-memory capacity were SUHGLFWRUVRI&,XVHUV·VXEMHFWLYHOHYHOVRIVSHHFKXQGHUVWDQGLQJ$EHWWHUXQGHUVWDQGLQJ RIWKHUHODWLRQEHWZHHQY/'7DQGVSHHFKUHFRJQLWLRQLQJURXSVRI1+OLVWHQHUVWKDWGLරHU LQODQJXDJHSURÀFLHQF\LVQHHGHGIRUWKHLQWHUSUHWDWLRQRIWKHVHUHVXOWVIRUOLVWHQHUVZLWK KHDULQJORVV7KHODWWHUSODXVLEO\GLරHUEHWZHHQWKHPLQOLQJXLVWLFDELOLWLHVZKHUHERWK auditory and non-auditory factors play a role.

Relations between auditory measures of linguistic abilities and recognition of degraded speech.$OVRLQWKHDXGLWRU\PRGDOLW\OH[LFDODFFHVVKDVEHHQVXEMHFWRILQYHVWLJDWLRQ IRUPDQ\\HDUVKRZHYHUPRVWO\ZLWKFOHDULQSXWVLJQDOV(඼FLHQWYLVXDOOH[LFDODFFHVV KRZHYHUGRHVQRWQHFHVVDULO\LPSO\H඼FLHQWDXGLWRU\OH[LFDODFFHVV7KLVKROGVHVSHFLDOO\ IRUVLWXDWLRQVZLWKGHJUDGHGLQSXWOLNH&,XVHLQHYHU\GD\FRPPXQLFDWLRQ$IHZUHFHQW studies used the visual world paradigm to investigate lexical access with degraded auditory stimuli by tracking eye-movements to pictures of target words and competitor words. They IRXQGWKDW1+OLVWHQHUVDGMXVWWKHLUVWUDWHJ\ZKHQWKHUHLVPRUHXQFHUWDLQW\LQWKHVLJQDO 0F4XHHQDQG+XHWWLJDQGWKHUHDUHLQGLFDWLRQVWKDW&,XVHUVDGDSWWKHLUDXGLWRU\ OH[LFDODFFHVVWRGHDOZLWKGHJUDGHGLQSXWHJ)DUULV7ULPEOHHWDO0F0XUUD\HW DO.QRZLQJLQZKDWZD\WKH\PLJKWKDYHGRQHVRFRXOGEHLPSRUWDQWIRUVSHHFK UHFRJQLWLRQRXWFRPHVRIDQLQGLYLGXDO&,XVHU

Based on the model, we hypothesize that for speech stimuli that are degraded by a noise YRFRGHUFRPSDUHGWRRULJLQDOVWLPXOLWKH57IRUUHDOZRUGVZLOOEHORQJHULQ1+OLVWHQHUV because the process of activation of candidates and evaluation of competitors takes longer HJ)DUULV7ULPEOHHWDOGXHWRXQFHUWDLQW\LQWKHVLJQDODQGSUREDEO\PRUHVSXULRXV DFWLYDWLRQRIOH[LFDOLWHPVLQWKHFRKRUW:HDOVRH[SHFWWKH57IRUSVHXGRZRUGVWREH longer in the vocoded condition because the response deadline is usually set higher for PRUHGL඼FXOWFRQGLWLRQVHJ5DWFOLර*RPH]DQG0F.RRQ)RU&,XVHUVZHH[SHFWD larger bias toward word responses, because they are used to guessing. These hypotheses DUHWHVWHGLQ&KDSWHU

(12)

OUTLINE OF THE THESIS

This thesis describes four studies that were conducted to achieve a better understanding of the role of linguistic skills, particularly of lexical-access ability, in speech-recognition performance of cochlear-implant users.

Chapter 2 describes a study that examines the suitability of clinical outcome measures of VSHHFKUHFRJQLWLRQLQ&,XVHUV7KHIHDVLELOLW\UHOLDELOLW\DQGYDOLGLW\RIWKH',1WHVWIRU PHDVXULQJVSHHFKUHFRJQLWLRQLQ&,XVHUVDQGKHDULQJDLGXVHUVLVH[DPLQHGLQUHODWLRQ WRFRPPRQO\XVHGFOLQLFDOWHVWVRIVSHHFKUHFRJQLWLRQLQTXLHWDQGQRLVH7ZHOYH1+ SDUWLFLSDQWVKHDULQJDLGXVHUVDQG&,XVHUVSDUWLFLSDWHGLQWKHVWXG\

Chapter 3 examines the relations of visual lexical-access ability and vocabulary size with GLරHUHQWPHDVXUHVRIVSHHFKLQQRLVHUHFRJQLWLRQLQ\RXQJDGXOW1+OLVWHQHUVZLWKYDULRXV OHYHOVRIOLQJXLVWLFDELOLWLHV7RREWDLQDZLGHUDQJHRIOLQJXLVWLFVNLOOVWKUHHJURXSVRI SDUWLFLSDQWVHDFKZHUHLQFOXGHGKLJKHUHGXFDWHGQDWLYHOLVWHQHUVORZHUHGXFDWHGQDWLYH listeners, and higher-educated non-native listeners. The aim of the study is to evaluate suitable linguistic measures and examine their relation with speech recognition in noise. $GGLWLRQDOO\WKHHරHFWRINH\ZRUGVFRULQJLQWKH6,1WHVWLVH[DPLQHG

Chapter 4 SUHVHQWV D VWXG\ WKDW H[SORUHG WKH LQÁXHQFH RI YLVXDOO\ PHDVXUHG OH[LFDO DFFHVVDELOLW\FRPSDUHGWRRWKHUFRJQLWLYHIDFWRUVRQVSHHFKUHFRJQLWLRQDELOLW\RI SRVWOLQJXDOO\GHDIHQHG&,XVHUV7KHOH[LFDODFFHVVPHDVXUHVDQGYRFDEXODU\VL]HPHDVXUH RI&KDSWHUDUHXVHGLQDGGLWLRQWRWZRFRJQLWLYHPHDVXUHVZRUNLQJPHPRU\FDSDFLW\ 5HDGLQJ6SDQWHVWDQGDYLVXDODQDORJXHRIWKHVHQWHQFHLQQRLVHWHVW7H[WUHFHSWLRQ WKUHVKROGWHVW7KHUHODWLRQVRIWKHVHDELOLWLHVZLWKWKHFOLQLFDOVSHHFKUHFRJQLWLRQPHDVXUHV WKDWDUHDOVRXVHGLQ&KDSWHUVDQGDUHH[DPLQHG Chapter 5 GLVFXVVHVDVWXG\RQWKHHරHFWRIGHJUDGHGDXGLWRU\LQSXWRQDXGLWRU\OH[LFDO GHFLVLRQUHDFWLRQWLPHVDQGDFFXUDF\LQ([SHULPHQWIRU\RXQJ1+OLVWHQHUVOLVWHQLQJ WRQRLVHEDQGYRFRGHGVWLPXOLDQGLQ([SHULPHQWIRU&,XVHUVIRUZKRPWKH&,LWVHOI SURGXFHVDGHJUDGHGLQSXWVLJQDO$QDO\VHVVKRZKRZ&,XVHUVGLරHULQWKLVDELOLW\IURP 1+OLVWHQHUVOLVWHQLQJWRGHJUDGHGLQSXW,QDGGLWLRQWKHUHODWLRQRIWKHVHDXGLWRU\OH[LFDO GHFLVLRQPHDVXUHVZLWKVSHHFKUHFRJQLWLRQLQTXLHWDQGLQQRLVHDUHH[DPLQHG

(13)