Speech domains, mapped by indication

A few minutes of speech carry three domains: cognitive, motor‑speech, and verbal output and content. The map shows which ones each indication uses.

Why speech

Where speech comes from

Speaking draws on the brain, the breath, the voice and the muscles of the mouth at once. Each leaves its own trace in the recording.

Brain

Plans what is said, finds the words and sets the timing.

Breath

Drives each phrase; changes show in phrase length and pauses.

Vocal folds

Produce the voice; changes show in pitch and voice quality.

Lips, tongue and jaw

Shape each sound; changes show in articulation and speech rate.

Three domains

Cognitive, motor‑speech, and verbal output and content

One pipeline extracts the speech features in every indication; each reading sits beside the scale that indication’s trials use.

Our pipeline

Cognitive

How the words are found and what they mean: memory, word-finding and semantic content. Read from picture description, narrative recall, verbal fluency, reading aloud and conversation.

  • Memory
  • Word-finding
  • Verbal fluency
  • Semantic content
  • Reading speed

Motor-speech

How the sounds are made: breath, voice, articulation and timing. Read from free speech, reading aloud, sustained vowels and syllable repetition, as the protocol sets.

  • Articulation
  • Prosody
  • Speech timing
  • Voice quality
  • Speech rate
  • Phonation
  • Respiratory support

Verbal output and content

How much is said, how quickly replies come and which words are chosen. Read from conversation, recorded rating interviews and short home recordings.

  • Verbal output
  • Response latency
  • Lexical content
Domain map

The domains read in each indication

Each row is one indication: the scale its readings sit beside and the domains they use. Recorded rating interviews also get flags per rater and per site. Trial QC ›

Scopes and deliverables
Speech domains by indication
IndicationCognitiveMotor-speechVerbal output and content
Alzheimer’s diseaseMMSE or MoCA; rater checks on CDR
  • Memory
  • Word-finding
  • Verbal fluency
  • Semantic content
  • Speech timing
Not used
Parkinson’s diseaseMDS‑UPDRS item 2.1 (Speech)Not used
  • Articulation
  • Prosody
  • Speech timing
  • Voice quality
Not used
Huntington’s diseaseStroop Word Reading (cUHDRS)
  • Reading speed
  • Articulation
  • Speech timing
Not used
ALSALSFRS‑R speech itemNot used
  • Articulation
  • Speech rate
  • Phonation
Not used
DM1 and neuromuscularMDHI communication subscale, FVC or bulbar itemsNot used
  • Articulation
  • Phonation
  • Respiratory support
Not used
Major depressive disorderMADRS or HAM‑D range, for pre‑screening; rater checks on bothNot used
  • Prosody
  • Speech rate
  • Verbal output
  • Lexical content
SchizophreniaPANSS negative symptom items; rater checks on PANSSNot used
  • Speech timing
  • Verbal output
  • Response latency
  • Lexical content
Agitation and psychosis in dementiaNPI caregiver interviewsEach recorded NPI interview is checked, with flags per rater and per site. Trial QC ›

Let’s talk about your program

Share your indication and the scale your trial uses; we map the domains to it.

Request

We use your work email to send you the document and to follow up about it and related work. You can opt out at any time. Privacy policy

Thank you. We will send it to your work email.

Something went wrong. Please write to contact@cephalgo.com.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.