1. Administration, Recording and Saving audios (Clinicians)

1. Administration, Recording and Saving audios (Clinicians)

0. Stimuli and task administration

 

The order of administration of these two sets of stimuli will be pseudorandomized so that half of all participants with each PPA variant are administered the first set in their L1, and the other half in their L2. Administration of both sets will occur one week apart. Additionally, three identical tasks (QAB Important event discourse task, procedural discourse task Brushing teeth and Picnic scene) will also be administered in both languages. (grant p.6, updated 20240406)

Update: CONTROLES, lab decision 20250829, Administration of both sets will occur one week apart, but 7-10 days is acceptable.

Task administration by timepoint:          

The tasks are organized in this table by order of administration (first to last). The redcap instruments match this order

SET 1 "Spanish"

SET 2 "Catalan"

TIMEPOINT

EB

Pre/Obs1

Mid

Post

6m

12m/Obs2

CatRescue - Picture Story Description ​

BirthdayParty – Picture Story Description​

ImportantEvent – QAB Discourse​ (S3)

Sunday - Personal Narrative​

Home – Personal Narrative​

Brushing teeth - Procedural Discourse ​(S3)

CatRescue - Recall​

BirthdayPartyRecall - Recall​

MAINDog – Story Generation​

MAINCat - Story Generation​

Frog, Where are you? - Wordless Picture Book Story Narration​

One Frog Too Many - Wordless Picture Book Story Narration​

PicnicScene – Picture Description​ (S3)

Bridge Video

Kindness Video 

Grandma (always in Spanish) - Picture Story Description​

MAINGoat (always in Spanish)  - Story Generation

Grandpa (always in Catalan)- Picture Story Description​

MAINBird (always in Catalan) - Story Generation​

Timepoint guide:

There are different timepoints depending on the branch of the study Clinical Trial, Observational Study and Screenings and Basic Evaluations

Evaluacion Basica

Clinical Trial

Observational/Controls

EB1

(Average duration 6min)

Pre

(average duration 40min)

Obs1

(average duration 40min)

Mid

(average duration 6min)

Post

(average duration 30min)

6m

(average duration 30min)

EB2

(Average duration 6min)

12m

(average duration 40min)

Obs2

(average duration 40min)

For

  • speechFTLD A: Cookie theft, picnic scene, brushing teeth, 10 min

  • SpeechFTLD B: picnic scene, brushing teeth, 6m

  • For Clinical Trial and Observational study participants:

The administration instructions can be found the Redcap BACC instruments:

Connected Speech S1 And S3 Span
 Connected Speech S1 And S3 Cat
Connected Speech S2 And S3 Span
Connected Speech S2 and S3 Cat

  • For extended screening/evaluación básica/non-R01 Obs participants:

  • Se rellena instrumento CS set 1 para los dos idiomas: Connected Speech S1 And S3 Span, Connected Speech S2 And S3 Span

  • Luego una vez este dentro de estudio R01 cuando se hace sesión de CS se mueven los datos al instrumento que corresponda y se borran los datos de ese instrumento.

! If order of administration is changed after a CS session please remember to fill out the Redcap instrument adding the order of administration by changing the numbers in the comment boxes of each task and adding the dates (do not do this during the session because it interferes with the audio recording).

  • Stimuli

The stimuli can be found here in Box: Connected Speech and SpeechFTLD stimuli

EXCEPTIONS of administration (applicable to all branches of the study (R01 clinical trial, R01 Obs, evaluación básica, etc.,) :

Administration exceptions for Severe Patients

When patients have a more severe difficulties and it's probable that they will not be able to complete all the CS tasks, please priorize the Post-treatment CS tasks.

Administer those in the predetermined order (shown in the scheme above and in the CS Redcap instrument)

After completing the CS Post-tx tasks take a small break and if the patient and if the patient isn't showing any moderate-severe signs of stress, tiredness continue with the administration of the CS tasks following the predetermined order of the battery (shown in the scheme above and in the CS Redcap instrument)

After a CS session please remember to fill out the Redcap instrument marking a variation in the order of the task administration and adding the order of administration by changing the numbers in the comment boxes.

Administration exceptions because of time contraint

Please make sure to plan the sessions ahead so there's enough time to administer the CS battery in complete (45 min moderate patients, 1h severe patients. Please read next page "preparation before CS sessions" to avoid this situation happening when possible.

Please priorize the Post-treatment CS tasks.

Administer those in the predetermined order (shown in the scheme above and in the CS Redcap instrument)

After completing the CS Post-tx tasks take a small break and if there's still time continue with the administration of the CS tasks following the predetermined order of the battery (shown in the scheme above and in the CS Redcap instrument).

After a CS session please remember to fill out the Redcap instrument marking a variation in the order of the task administration and adding the order of administration by changing the numbers in the comment boxes.

This tasks were created by Dr. Miguel Ángel Santos-Santos, each of the questions relate to different cognitive domains, the 2 tasks are balanced in terms of questions (Bridge Video, Kindness Video)

 Cognitive domain

Pregunta

PROPS SMK - Possible quantitative punctuation

Anomia 

1.En el video había varios personajes, puedes nombrarlos por favor? 

Language and speech analysis from transcription

Visual imaginary 

2. La historia del video ocurrió en la naturaleza, por favor describe con el mayor detalle el entorno, el lugar donde ocurrió. 

Language and speech analysis from transcription

 

Action words – Embodied cognition 

  1. En el vídeo, habia un perro y un pajaro que hicieron varias acciones, por ejemplo el perro estuvo corriendo, por favor describe con el mayor detalle todas las acciones que hizo el perro 

  2. También habia un pájaro, por favor describe con el mayor detalle todas las acciones que hizo el pajaro. 

Language and speech analysis from transcription

PROPS quantitative scoring: How many verbs/actions are mentioned?  

Theory of mind 

  1. Aunque el perro y el pájaro no pueden hablar, seguro que pensaron muchas cosas. Que crees que pensaba el perro cuando vio al pájaro por primera vez? ...  

  2. Que crees que pensaba cuando vio a los pajaros bebes al final 

  3. Que crees que pensaba el pájaro? 

Language and speech analysis from transcription

 

TOM-emotions 

  1. Cual era la actitud del perro al principio y al final de la historia? ... Por favor describe las emociones que sintio el perro. 

  2. Cual crees que es la moraleja de esta historia? 

Language and speech analysis from transcription

 

Variability in Administration

  • 20250929 R01 meeting decision Video task: Issue 1: Differences in whether clinicians stop participants who give extra information beyond the question. Decision: In the video part of the test, participants should be instructed to answer concretely and directly to each question. Issue 2: Variability in asking/skipping questions: Decision : All questions of the video CS part must be asked, even if the participant already answered within another response.


0.1 Preparation before a connected speech session (Clinicians)

 

  1. Mirar que todos los materiales esten en la bolsa, llevar todo (Recording hardware setup and session materials- Updated 20251215) y Revisar que este cargado.

Hospital Sant Pau: Clinitian 1

✓/✕

Equipment

Use

 

Laptop 1 (Lenovo ThinkPad L15 Gen3 - 21C30015SP for the sessions)

Show stimuli, record zoom session, play audios

 

Laptop 2 (clinicians lenovo Lenovo ThinkPad L15 Gen3 - 21C30015SP)

Record session in audacity, administer and collect the tests in redcap

 

Ipad 1 Apple iPad Pro 11"(a42731)

For carepartners to fill out the PASS and syntom questionnaire

 

Ipad 2 Apple iPad Pro 11" (A42727)

For clinitian to administer the NIH toolbox

 

Soundcard Behringer U-Phoria UM2 USB Audio Interface

Soundcard to connect with laptop 2

 

Closed back headphones AKG K72

For speechFTLD

 

Frog story books (3)

For conn.speech

 

Shure WH20XLR Dynamic Headset Microphone

To connect with soundcard

 

Ruler (4cm marked)

To make sure distance with mic is correct

 

Straws

For the patient to drink water while recording

 

Anagramas for Sentence production

In a folder or binder

Hospital Sant Pau Clinitian 2 (if the session is happening at the same time, less ideal equipment)

✓/✕

Equipment

Use

 

Laptop 3 (Lenovo ThinkPad L15 Gen3 - 21C30015SP )

Show stimuli, record zoom session

 

Laptop 4 (clinicians Lenovo ThinkPad L15 Gen3 - 21C30015SP)

Record session in audacity, administer and collect the tests in redcap

 

Tablet china Miguel

For carepartners to fill out the PASS and syntom questionnaire

 

Ipad 3 Apple iPad Pro 11" (a42729)

For clinitian to administer the NIH toolbox

 

Soundcard Behringer U-Phoria UM2 USB Audio Interface

Soundcard to connect with laptop 4

 

Ultrasone headphones

For speechFTLD

 

Frog story books (3)

For conn.speech

 

MicroMic the original C520L

To connect with soundcard focusrite Scarlett

 

Ruler (4cm marked)

To make sure distance with mic is correct

 

Straws

For the patient to drink water while recording

 

Anagramas for Sentence production

In a folder or binder

BClinic

✓/✕

Equipment

Use

 

Laptop 3 (Núrias laptop)

Show stimuli, record zoom session

 

Laptop 4 (Lenovo ThinkPad L15 Gen3 - 21C30015SP )

Record session in audacity, administer and collect the tests in redcap

 

Ipad 4 Apple iPad Pro 11" (a42730)

For carepartners to fill out the PASS and syntom questionnaire

For clinitian to administer the NIH toolbox

 

Soundcard Behringer U-Phoria UM2 USB Audio Interface

Soundcard to connect with laptop 4

 

AKG Pro Audio K240 STUDIO Over-Ear, Semi-Open,

For speechFTLD

 

Frog story books (3)

For conn.speech

 

MicroMic the original C520L

To connect with soundcard focusrite Scarlett

 

Ruler (4cm marked)

To make sure distance with mic is correct

 

Straws

For the patient to drink water while recording

 

Anagramas for Sentence production

In a folder or binder

Here you can find the checklist to print out: https://utexas.box.com/s/97b8u0j0c7wavms43gfnj152733cpl01

  1. Revisar que set se administra al paciente mirando el assessment timeline, redcap y consultando con co-clinico o equipo si es necesario.

Recommendations:

  • Preparation before the session: 

    • Make sure you have all the materials ready for the session, prepare set up 

    • Have time in between IP sessions, sessions should be scheduled with a 30min separation


1. Recording procedures (Clinicians)

1.1 Setting up Audacity for recording

If administering remotely: request remote control and download audacity 

Always keep audacity open, sometimes it'll drop out like an audacity will sort of lose the identification of the device.

1.1




Sampling Rate: 44.1k (> 2x 20,000)

1.2

Monochannel recording (not stereo)

1.3

Select microphone (USB Audio Codec)
And make sure that the level of recording is 50%

  1. updated 20240209 after checking it with Rosemary, before we had it in 100%






 

1.2. Mic set up

make sure this 48 volts switch is on in the back.

Mouth to mic distance (4cm, or 4 cm-poof thickness)

Level with the mouth and 45 degrees off axis

Remember to check that mic remains in this same position throughout the sample and ask patient to not touch the mic

!! Please check in the lenovos and other laptops that the mics are 100%, Volumen should be 100%

1.3 Adjust the gain (mic and audacity)

On sound card,

Set the first knob all the way to the right 

Se pone a la derecha al máximo (volumen máximo) y se reduce el mínimo posible (se mueve a la izquierda)

 

Adjust audacity nivel de grabación (new):

3.1

Ask patient to produce a comfortable /a/

If patient's comfortable /a/ is close to the upper range, lower the gain

 

  1. antes del cambio de 50% revisábamos que estuviera a -12 a -24, ahora lo que hay que confirmar es que este en VERDE, In consultation with Rosemary Lester Smith, we decided as long as it does not appear red, it is not saturated

 

check if we have to regulate the nivel de reproducción.
and how we decide to adjust the gain

adjust in audacity (ideally always the same)

adjust in soundcardLOUDER A:

Si sale ROJO, regular (Incorrecto! Sonido muy saturado):

3.2

Ask patient to say a louder /a/ to make sure it's still within range.

  • visual cue: flat upper limit on the waveform

1.4 Measuring SNR (Signal to Noise Ratio, Audacity and Praat)

4.1

Sample 3 cycles of silence-softest /a/

Ask the patient to be silent for 5 seconds and then say /a/ for 5 seconds in their softest voice, complete 3 repetition of silent-softest /a/.

Some patients might need to be cued to say the /a/ softer during the cycle. That's ok!

Añadir pista en audacity para grabar


2. SNR - Nombrar pista correctamente (ej. BISE000_BACC1_calib_20240130) y pulsar botón rojo 🔴 para grabar y cuadrado ◼️ para parar.



4.2

  • Exporta la calibración como WAV.

  • This file should be saved as BISE000_BACC1_calib_20240130

1.5 Measuring intensity with Praat

5.1

Open File on Praat




5.2

Zoom in until the view shows the spectogram, seleccionar parte de audio, pulsar " IN "



5.3

From the silence-softest /a/ trials:

Select a 1 second segment of silence, put it in Praat and get intensity.

 

 



5.4

From the silence-softest /a/ trials:

Select a 1 second segment of softest /a/ and get intensity

5.5

Calculate SNR by subtracting the silence intensity from the softest /a/ intensity
(Softest /a/)dB – (silence)dB = SNR

Calculate it in the Redcap Connected Speech instrument.
Ideally SNR is between 20db and 40db.

If SNR is within 20dB-25dB, the session should not be completed. **(Look below at solutions)

  1. Decision 20250716 decir los numeros en voz alta, antes de esta fecha solo se hacia si la co-clinica estaba a distancia y tomaba nota de los numeros en redcap




Example:softest /a/ 60.34dB - silence 29.98dB = 30.36dB

1.6 Record Connected Speech track (NAMING CONVENTION)

Audios for all the sessions

6.1

Crear pista y nombrarla para hacer connected speech- y pulsar botón rojo 🔴 para grabar y cuadrado ◼️ para parar.

See naming convention below:


Para todos los participantes guardamos las muestras de connected Speech dentro de Data Raw, Connected Speech Box folder.

This table summarizes when to administer the battery, for the Clinical trial, Observational study and for Non-R01, Extended screening visits.

​CLINICAL TRIAL (therapy) 

CODE001_BACC1_CS1_Spa_Pre_20230914

​OBSERVATIONAL

CODE001_BACC1_CS1_Spa_Obs_20230914

Non-R01 Obs/Evaluación Básica/Extended screening 

CODE001_BACC1_CS3_Spa_EB1_20230914

! 20250516 confirmation- m4a has better audio quality than the mp4 videos converted to m4a audios, this is why we want to save m4a audios

​Pre​

​Obs1 ​

EB1

Mid ​

_ ​

-

​Post ​

​_ ​

-

​6m​

_​

-

​12m ​

​Obs2 ​

EB2

  • Audio Calibracion (ex. CODE001_BACC1_calib_20230914)​ – Recomendation: naming it before the session

  • Audio Connected speech: MADRlabID_BACCid_CS (Connected Speech) 1 o 2 language (span/cat)_TIMEPOINT fecha (YYYYMMDD)
     (ex. CODE001_BACC1_CS1_Spa_Pre_20230914) ​

  • Audacity project (ex. CODE001_BACC1_20230914)

Audio en varias tomas

Para Connected Speech si el audio se graba en más de una toma por favor poner "Part1" "Part2" "Part3" después del punto de tiempo, ejemplo:

Esto ayudará a los estudiantes que tienen que hacer clipping (dividir los audios de la sesión en audios de las pruebas correspondientes) y encontrar las pruebas ya que se deben de administrar en el mismo orden y también ayudará a que no borremos audios por error pensando que son los mismos (casi borro uno de estos por ejemplo). 

Asegúrese de actualizar los Smartsheets de Connected Speech antes de dar el proceso como acabado

1.7 Guardar pistas (calibración y connected speech y proyecto audacity)

1.7.1 Cómo guardar muestras de Connected Speech

Inmediatamente después de la sesión: Tutorial_savingCSdata_inBox_20240130

7.1

Guardar pistas (calibración y connected speech y proyecto audacity)

Seleccionar "múltiples archivos" y exportar todos los audios en escritorio
! las pistas tienen que estar azules y "silencio" no puede estar marcado


Y comprobar que se han exportado bien:

7.2

Guardar proyecto audacity en escritorio


Pulsar si y nombrar: MADRID_BACCID_20240130
Ejemplo:CODE001_BACC1_20240130

7.3

En carpeta de escritorio "Audacity" Subir a BOX todos los audios de la sesion y proyecto audacity

Los audios y el proyecto deben de estar guardados en el escritorio en carpeta llamada "audacityesta carpeta la crea cada clínico y se asegura de que los archivos de audacity se guardan aquí"





Subir a C:\\Box\SLHS_Grasso\International Participants - Recordings and Materials\MADR_HSP R01\4. Data_Raw\ALL audios IP sessions

Audio Calibracion (ej. CODE001_BACC1_CS1_Spa_Pre_20240130

Audio Connected speech (ej. CODE001_BACC1_CS1_Spa_Pre_20240130

Proyecto audacity (ej. CODE001_BACC1_ 20240130)





NOTA: mover a "already uploaded to box" después los archivos de la carpeta del escritorio

7.4

Copiar los audios en la carpeta de BOX correspondiente

Decision (MADRlab Meeting – 2025-09-26):
For BClinic, Núria will upload the audio files to ALL IP PARTICIPANT audios. The co-clinician will then copy each file to the corresponding folder and add a new row to the Smartsheet, completing all columns up to (but not including) the transcription column, given the time limitations during in-person sessions.

Copiar a C:\Users\IR_User\Box\SLHS_Grasso\International Participants - Recordings and Materials\MADR_HSP R01\4. Data_Raw\B--Connected Speech_Data a la carpeta correspondiente (ejemplo: CODE001_BACC1_CS1_Spa_Pre_20240130, paciente de Clinical trial en preterapia que ha hecho sesión en español)

Para el audio de Mid-treatment guardarlo en la carpeta de castellano como en la de catalan

Therapy trial o Observational



Spanish, Catalan or English



P

7.5

Updating Smartsheet

After saving the audios in Box the last step is going into the Data analysis smartsheet Spanish, Catalan or English depending on the language of the CS session:

Esto tiene que ser completado por el clínico en persona que hace la grabación, y el clínico principal tiene que revisar de que esté bien completado.

SPANISH Connected Speech Data analysis

CATALAN Connected Speech Data analysis

ENGLISH Connected Speech Data analysis

Create a row for the participant and fill out the information:

BACC ID, MADRID, enrolled in study, administered set, Audio uploaded and copied to Box, Recorded in (audio format). See first row "recording clinician" indicating that the column needs to be fill out.

  • latest comment column, añadir comentario de dificultades de grabación, variabilidad, producción limitada de paciente… en la fila del paciente es smartsheet

1.7.2 Midtx muestras, cómo guardar

Para el audio de Mid-treatment guardarlo en ALL IP session audios, copiarlo en la carpeta de castellano como en la de catalan (o castellano-inglés si son bilingües castellano inglés). Y la misma clínica que descarga los audios de la sesión de Zoom debe actualizar el smartsheet de CS para los dos idiomas:

20240117 Decision - not to download both audios, with and without clinician speech, just the one with clinician speech to make the process more manageable.

20250516 confirmation- m4a has better audio quality than the mp4 videos converted to m4a audios, this is why we want to save m4a audios

Esto lo hacemos así para que los estudiantes que hacen clipping puedan encontrar los audios fácilmente y porque el Connected Speech sucede en la misma sesión a diferencia de otros puntos de tiempos.

Revisar aquí NAMING CONVENTION de como guardar: 1. Administration, Recording and Saving audios (Clinicians)

1.7.3 DFT participants muestras, cómo guardar

El neurólogo de la Unidad de Memoria Jesús sube los audios aquí después de las sesiones. Las logopedas en Sant Pau se aseguran de que no haya audios en el ordenador que no se hayan subido.

C:\Box\SLHS_Grasso\BCN Participants - Recordings and Materials\MADR_HSP R01\5. Data_Raw\B--Connected Speech_Data\Screening, Non-R01 and Basic Evaluations\Z. DFT evaluations

Connected Speech SOP para DFT