Case study

Patient W.H.

• male; 70; manager; multilingual:
• German (native)
• Spanish (30 years Venezuela)
• English (10 years USA)

Symptoms

• auditory comprehension and repetition of heard speech greatly limited
• audiometric tests - normal
• comprehension of non-verbal auditory stimuli - normal
• the ability to speak, read and write - unaffected
• tendency to (unconciously) use lip-reading as an aid in auditory comprehension (careful observation of the face of the examiner)
• auditory comprehension is variant and context-related: it is superior when the patient introduces the subject of conversation himself, and it drops drastically when the examiner suddenly changes the topic
• isolated words are less easily identified than words embedded in sentences

Speech samples

Die folgenden Audioaufnahmen sind nur für registrierte Benutzer zugänglich.

• spontaneous speech, german [wav, 103.01 sec]
• spontaneous speech, spanish [wav, 27.52 sec]
• repetition, german [wav, 40.15 sec]
• repetition, english [wav, 34.24 sec]
• repetition, spanish [wav, 25.46 sec]
• reading, german [wav, 17.8 sec]
• reading, spanish [wav, 12.33 sec]

Tests

The following tests were applied 1 month, 6 months, 8 months and 12 months post onset:

1. Phoneme discrimination test (forty item CV test)
   • 20 identical pairs (e.g. /ta/ /ta/)
   • 10 phon. distant pairs (3 feature difference)
   • 10 phon. close pairs (1 feature difference)

   W.H.'s performance (69% - 73% correct) well below lower end of the normal range for his age group (89%). The results were stable over all 4 tests.

2. Phoneme restoration test

   14 short stories with an appropriate and an inappropriate conclusion expressed by a minimal pair (e.g. klettern - kleckern). Word frequency, word length, phoneme class, and the position for the restored phoneme were varied. Signal editing: whole phoneme deleted, burst deleted, transition deleted, only the transient present.

   Controls follow auditory cues (if any), W.H. follows only the contextual cues. The results were stable over all 4 tests.

3. Bimodal perception test (McGurk effect)
   • 5 synthetic auditory syllables (/ba/.../da/) are combined with 5 synthetic visual syllables (/ba/.../da/) = 25 stimuli
   • In addition audible and visible speech is presented alone = 10 stimuli
   • 35 stimuli in six random sequences

   W.H. relies on visible speech, controls rely on both types of cues. The results were stable over all 4 tests.

4. Intonation (Lx) perception test

   F0 cues (Lx signal) for male voice, female voice, assertion, question, verification, topic change, aspect change, agreement, dissonance, etc., were presented in 6 random sequences.

   W.H. identifies the intonational cues to discourse structure much better than the controls do (e.g. topic change 67% W.H. - 24% C.). The results were stable over all 4 tests.