Ambulatory speech monitoring to assess function and transfer of treatment changes beyond the clinical setting - Movement Disorder ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Ambulatory speech monitoring to assess function and
transfer of treatment changes beyond the clinical setting
Ellika Schalling, Associate Professor, SLP
Division of Speech and Language Pathology
Department of Clinical Science, Intervention and Technology
Karolinska Institutet
Elika Schalling July 1, 2021 1
Greetings from Stockholm Elika Schalling July 1, 2021 2
Karolinska Institutet
(teaching and research)
Division of Speech
and Language Pathology,
Department of Clinical Science,
Intervention and Technology
Karolinska University Hospital
Huddinge (clinical work)
Medical Unit Speech
and Language Pathology
Elika Schalling July 1, 2021 3
Agenda: ▪ Assessment of speech and communication disorders ▪ Ambulatory phonation monitors – what are they? ▪ Exampels of studies using APMs with people with PD ▪ Challenges using APMs Elika Schalling July 1, 2021 4
Speech and communication changes are very common in people with movement disorders Communication problems in 70-90% of people with Parkinson’s disease 92.5% reported at least one symptom related to speech and communication in a survey in Sweden (Schalling et al., 2017) Elika Schalling July 1, 2021 5
Symptoms PD: Changes in voice, articulation and prosody - Reduced intensity - Breathy, harsh voice quality - Reduced articulatory precision - Variable speech rate, monoloudness, monopitch HD: Involuntary movements affect respiratory, phonatory, resonatory and articulatory levels of speech Elika Schalling July 1, 2021 6
Communication impairment has negative consequences Ten percent of people with PD reported that worsening of speech was their greatest problem caused by PD (38% - one of four worse problems) (Miller et al., 2007) Reduced participation in life situations (work, leisure, family, friends) self- reported by 30-40% in a survey of people with PD (Schalling et al., 2017) Elika Schalling July 1, 2021 7
Assessment of communication changes in
motor speech disorders?
▪ Clinical tests (dysarthria tests)
→ Oral motor function, respiration,
phonation, articulation, prosody, intelligibility
Questionnaires or interviews to assess
communicative participation
Elika Schalling July 1, 2021 8
Assessment of communication changes
▪ Speech recordings
Sound-proof booth, high quality microphone (headset)
Standardized recording protocol for acoustic and perceptual analyses
E g: Measures of voice intensity,
fundamental frequency, speech
rate…
Elika Schalling July 1, 2021 9
▪ Is the data representative for voice use in real life?
→ Outside the clinic other factors, such as stress, background noise, air quality,
movement etc can affect the voice.
Elika Schalling July 1, 2021 10Documentation of speech and voice outside the clinic?
Ambulatory monitoring of voice/speech
▪ Devices that allow documentation of vocal use/vocal behaviour in
different situations outside the clinic (work and leisure)
▪ Voice accumulators, phonation monitors, voice dosimeters….
▪ Not a speech recording (confidentiality not an issue)
INFORMATION ABOUT VOCAL BEHAVIOUR
IN EVERYDAY LIFE…
Elika Schalling July 1, 2021 11Documentation of voice outside the clinic Elika Schalling July 1, 2021 12
Long-term monitoring of voice
- variation in terminology and methods…
Voice accumulators (e g Buekers, 2001; Ohlsson, 1988; Szabo Leroy, 2004)
Voice accumulator with microphone (Airo, et al., 2000)
DAT-recorders (Rantala et al., 1994; 2002)
Binaural DAT-recorders (McAllister et al., 2009; Södersten, et al., 2002)
Voice dosimeter (Titze et al., 2003)
Ambulatory Phonation Monitor APM (Cheyne et al., 2003; Hillman, et al., 2006).
VoxLog (Lindström et al., 2009; 2010; 2011)
Elika Schalling July 1, 2021 13Long-term monitoring of speech and voice use
outside the clinic
▪ Studies of long-term monitoring of voice use…
→ Individuals with high vocal loads such as singers or teachers (Hunter & Titze, 2010;
Lehto et al., 2009; Laukkanen et al., 2008; Rantala et al., 2002; Lyberg-Åhlander, 2014; Toles et al., 2020; 2021; Van
Stan et al., 2021)
→ Few studies of neurogenic voice disorders, such as people with PD (Bulukin Wilén &
Gustafsson, 2011, Schalling et al., 2013)
→ Adults with cochlear implants (Mozzanica et al., 2019)
▪ Published reference values are limited
Elika Schalling July 1, 2021 14▪ Commercially available voice accumulators:
→ Ambulatory phonation monitor, APM
(KayPentax, New Jersey, USA)
→ Vocalog
(Griffin Laboratories, California, USA)
(VoxLog)
(Sonvox AM, Umeå, Sweden)
Elika Schalling July 1, 2021 15VoxLog - a portabel voice accumulator developed in Sweden
(Lindström et al., 2009; 2010; 2011)
▪ Accelerometer - registers
fundamental frequency and
phonation time)
▪ Microphone registers voice
intensity AND background noise
▪ Waist-pack for data storage
Elika Schalling July 1, 2021 16Ambulatory Phonation VoxLog
Monitor, APM
Elika Schalling July 1, 2021 17Vocalog - ”vocal activity monitor”
Griffin laboratories (www.vocalog.com)
▪ Registration of voice intensity and phonation time
▪ Feedback on voice intensity
Elika Schalling July 1, 2021 18Monitoring parameters: Elika Schalling July 1, 2021 19
Comparison and description of commercially available
PVA:s with focus on clinical applicability
(Van Stan, Gustafsson, Schalling, & Hillman, 2014)
Ambulatory Phonation Monitor (APM)
Voxlog
Vocalog
(smart-phone based system – reference)
Simultanous registrations during a 90-minute
lecture
Elika Schalling July 1, 2021 20The VoxLog
The VoxLog can register:
▪ Pitch (Hz)
▪ Voice intensity (dB SPL)
▪ Phonation time (%)
▪ Level of background noise (dB SPL)
Biofeedback function (pitch and voice intensity)
▪ Registration of voice data for up to one week
▪ No calibration - users can put on the neck-collar themselves
▪ Charge during the night
Elika Schalling July 1, 2021 21Biofeedback function
▪ Tactile feedback
(vibration signal from waist-pack)
▪ Feedback re either fundamental frequency
(Hz) or voice intensity (dB SPL)
▪ Administered below or above threshold
determined by the SLP
For PD – feedback on voice intensity
Elika Schalling July 1, 2021 22VoxLog Connect software for processing and presentation of data from the VoxLog Elika Schalling July 1, 2021 23
Label and process different sections Elika Schalling July 1, 2021 24
EXAMPLES FROM OUR STUDIES: - Is speech in (PD) in the clinic representative of voice use in daily life - How does variations in background noise impact voice use (in PD) - Do treatment effects carry over to speech outside the clinic - Can voice intensity change following biofeedback in PD Joakim Körner Gustafsson Collaborators/co-supervisors: Professor Maria Södersten, KI Professor Sten Ternström, KTH Elika Schalling July 1, 2021 25
Voice use outside the clinic in different levels of background noise in pwPD compared to matched healthy controls? Elika Schalling July 1, 2021 26
Participants
▪ 21 participants (11 ♂, 10 ♀)
→ PD, hypophonia, subjective speech/voice problem, hearing ok
▪ 21 healthy speakers matched for:
→ Gender
→ Age
→ Work (working or retired)
→ Living (alone/with partner)
no speech/voice problem, no neurological disease, no severe hearing impariment
Elika Schalling July 1, 2021 27Procedure
▪ Voice-data was registered with the VoxLog outside the clinic for seven
consecutive days.
▪ Three studio recordings of monologue speech were performed.
VoxLog day 1 + VoxLog day 4 + Voxlog day 7 +
VoxLog day VoxLog day VoxLog day VoxLog day
studio studio recording studio
2 3 5 6
recording 1 2 recording 3
▪ Participants logged their activities in a voice journal and rated
subjective voice use three times per day on VAS (100 mm)
→ Subjective rating of loudness, background noise and phonation time
Elika Schalling July 1, 2021 28Data analysis
▪ Data imported for analysis with the software VoxLog connect
▪ Visually inspected (faulty registrations removed, e g disconnection
between device and collar)
▪ MATLAB custom-script for analyses
→ calculation of mean voice sound level, noise sound level, and phonation ratio for
the whole registration period as well as means for three different environmental
noise ranges
▪ Calculation of self-to-others-ratio (SOR) – difference between
uncompensated voice sound level and background noise for the same
time-frame.
Elika Schalling July 1, 2021 29Results Voice sound level (dB) and environmental noise (dB) for the whole registration period for participants with PD and matched controls Elika Schalling July 1, 2021 30
Average voice sound level for monologue speech during a studio registration and in different levels of environmental noise during long-term registration in daily life for participants with PD and healthy controls Elika Schalling July 1, 2021 31
Self-to-others ratio (SOR) in different levels of environmental noise during long-
term registration in daily life for participants with PD and healthy controls
Elika Schalling July 1, 2021 32Phonation ratio in different levels of environmental noise during long-term registration in daily life for participants with PD and healthy controls Elika Schalling July 1, 2021 33
Conclusions
▪ Speakers with PD use lower voice sound levels then healthy controls during
studio recordings (we knew that…) – the difference is however even greater
during registrations in daily life
▪ Speakers with PD have a harder time adjusting their voice sound level to
background noise – only 15 dB above noise in low background noise levels
▪ Speakers with PD have lower phonation ratios, especially in noisy
environments (where communication often occurs…) – an objective
indication of communicative participation?
▪ Phonation monitoring give valuable, objective, more detailed information
about voice use outside the clinical setting
Elika Schalling July 1, 2021 34Do treatment effects carry over to speech in situations outside the clinic? Elika Schalling July 1, 2021 35
Effect of treatment, in daily life…
Participant with PD (4 years post onset), monozygotic twin-
brother as control (similar life/work situations)
Registrations with VoxLog for both participants
PD: base-line, four weeks during LSVT, post tx and
follow up (3, 6 and 12 months post tx)
Control: base-line and follow up (total 378 h)
Registrations of monologue speech in studio
environment
Elika Schalling July 1, 2021 36Treatment outcome (subjective)
▪ Increased loudness
▪ Speech not affected by general fatigue after the intervention
▪ Made himself heard at work (meetings)
▪ More energy, could speak in the evenings
▪ Increased fluency
▪ Follow up (12 months) – reported progression of symptoms (limbs,
balance), but stable voice function
Elika Schalling July 1, 2021 37Treatment outcomes (VoxLog registrations) Studio recording: 5.6 db SPL increase post tx, 3.8 dB increase at one year FU Elika Schalling July 1, 2021 38
Changes in loudness in different levels of background noise Elika Schalling July 1, 2021 39
Conclusions
▪ Case study…
▪ First attempt to use an APM to study carry over of treatment effects
following LSVT in PD
▪ Differences between lab-recording and registrations in daily life
▪ Voice use in relation to background noise provide us with interesting
information about the ability to adapt to the environment (changed over
time)
Elika Schalling July 1, 2021 40Can voice intensity change as a result of biofeedback in PD? Elika Schalling July 1, 2021 41
Aims
▪ Examine voice use in daily life during a period
of up to 3 weeks in subjects with PD.
▪ Evaluate the effects of tactile biofeedback,
administered with a portable voice
accumulator, on voice intensity in subjects
with PD.
▪ Obtain information from subjects about their
experiences using the VoxLog for a period of
up to 3 weeks.
Elika Schalling July 1, 2021 42▪ Participants
Six individuals with idiopathic PD, speech/voice symptoms
(documented by SLP)
no other neurological disease, no previous SLP-intervention,
no severe hearing loss (participant or partner)
▪ Procedure
→ 1 week of baseline registrations (pre)
→ 1 week of biofeedback with threshold levels for biofeedback
based on baseline registration (feedback/intervention)
→ 1 week of post intervention registrations (post)
Elika Schalling July 1, 2021 43Results
Significant difference in
MEAN VOICE INTENSITY (dB SPL)
voice intensity between
Pre and feedback periods
No significant difference
Between pre and post
Elika Schalling July 1, 2021 44Participant’s experiences using the VoxLog ”It was easy to use the VoxLog” ”I would consider wearing the VoxLog again” 6 participants responded ”completely agree” to both statements 5/6 reported no discomfort caused by neck-collar (1 participant thought it was slightly too tight) One participant wanted to keep the device to use it all the time… Elika Schalling July 1, 2021 45
Biofeedback-study 2: Körner Gustafsson et al.
(under review)
▪ 9 participants with hypophonia in PD, continuous registration with
feedback during four weeks (one week base-line, four weeks with
feedback, one week post and FU at 3 months)
Baseline Intervention Post intervention 3 month follow-
(week 1) (week 2-5) (week 6) up (week 7)
• Field • Field • Field • Field
registration registration registration registration
with VoxLog with VoxLog with VoxLog with VoxLog
• 3 visits to the • 4 weeks of • 2 visits to the • 2 visits to the
clinic biofeedback clinic clinic
• 3 recordings intervention • 2 recordings • 2 recordings
of monologue • 1 visit to the of monologue of monologue
speech clinic each speech speech
• Dysarthria week to • QASD, VHI • QASD, VHI
assessment, upload data
MoCA, QASD, and adjust
VHI threshold level
if needed
Elika Schalling July 1, 2021 46Outcome measures:
▪ Self-to-other ratio (SOR)
- difference between the voice sound level and the level of
environmental noise (Szabo Portela et al., 2018)
▪ SOR in different levels of environmental noise
(low 70 db)
▪ Voice sound level in studio recordings (comparison)
→ Divided in two groups: Participants with MoCA-score above or below 26 points
Elika Schalling July 1, 2021 47Results
SOR in three levels of
background noise
(70 dB)
Left: MoCA above 26
Right: MoCA below 26
T1 (pre)
T2 (post)
T3 (FU)
Elika Schalling July 1, 2021 48Voice sound level, monologue speech (studio) at
base-line, post intervention and follow up
MoCA > 26 MoCA < 26
90 90
88
Mean Mean post: Mean FU: 88 Mean Mean post: Mean FU:
baseline: 79.4 dB 78.7 dB baseline: 78.9 dB 75.5 dB
86 86
Voice sound level (dB)
Voice sound level (dB)
76.3 dB 77.5 dB
84 84
82 82
80 80
78 78
76 76
74 74
72 72
70 70
ID1 ID2 ID7 ID8 ID9 ID3 ID4 ID5 ID6
Elika Schalling July 1, 2021 49Conclusions:
▪ Feedback regarding voice sound level administered in daily life can
help people with PD increase their voice sound level in relation to
background noise
▪ Positive effects of feedback regarding voice sound level in daily life for
participants with normal cognitive function (as screened with MoCA)
▪ Positive changes also found in studio-recordings (for participants with
MoCA-scores above 26)
▪ Biofeedback regarding voice sound level may be used as a
complement to traditional voice treatment (cognitive function should be
taken into account)
Elika Schalling July 1, 2021 50Challenges in data collection using APMs:
Data loss
▪ Insufficient data
▪ Declined/logistic problems
▪ Drop out during intervention
▪ Technical problems….
▪ In a study with 23 participants, only
complete data from six participants
Elika Schalling July 1, 2021 51Factors found to facilitate data collection:
▪ Clear instructions (verbal and written)
▪ Monitor data registration during registration period (clarify/repeat
instructions)
▪ Continued development of more user-friendly devices is needed
Elika Schalling July 1, 2021 52Summary:
▪ Speech (voice) monitoring in daily life provides information that may
complement data from controlled studio recordings
▪ Voice monitoring in daily life provides valuable information about ability
to regulate loudness in relation to background noise
▪ Feedback on voice sound level may improve loudness, however
cognitive dysfunction may impact results
▪ Data collection can be challenging - technical development of more
user-friendly devices is needed
Elika Schalling July 1, 2021 53You can also read
