Speech-in-Noise Testing and Audiometric Evaluation Vocabulary

Audibility, Distortion, and Sensorineural Hearing Loss

• Audibility (Attenuation): Refers to loss of signal intensity due to elevated pure-tone thresholds.
• Distortion ($D$ Component): Peripheral and neural factors causing speech degradation that cannot be predicted by pure-tone thresholds alone for mild-to-moderately-severe losses.
• Plomp ($1986$) Framework: Divided sensorineural hearing loss (SNHL) into attenuation and distortion components, demonstrating that peripheral distortion occurs before neural transmission to the auditory cortex.
• Carhart and Tillman ($1970$) and Fitzgerald: Advocated for speech-in-noise testing as a standard component of the basic audiologic evaluation for all patients.
• Low-Pass Filter Effect: High-frequency SNHL acts as a low-pass filter, impairing high-frequency consonant recognition and requiring higher auditory closure skills.

Recruitment and Dynamic Range

• Recruitment Definition: Abnormal growth of loudness in SNHL caused by outer hair cell damage and the compensatory recruitment of adjacent outer hair cells doing double duty.
• Differential Diagnosis:
  • Recruitment occurs exclusively in SNHL.
  • Hyperacusis (hypersensitivity to normal sound), phonophobia (fear of normal sound), and misophonia (dislike of specific sounds) can occur in both normal and impaired hearing systems.
• Dynamic Range Compression: Normal dynamic range is $0\,\text{dB}$ to $100\,\text{dB}$. SNHL compresses this range (e.g., a $60\,\text{dB}$ hearing loss reduces dynamic range to $40\,\text{dB}$).
• Pathological Variations:
  • Hyper-recruitment: Abnormally rapid loudness growth rate.
  • Decruitment: Abnormally slow loudness growth, associated with extremely high loudness discomfort levels (LDLs).

Speech-in-Noise Concepts and Protocols

• Signal-to-Noise Ratio (SNR) Loss Definition: The increase in SNR (in $\text{dB}$) required by a hearing-impaired listener to achieve $50\%$ word or sentence recognition compared to normal listeners.
• Normal SNR Advantage: Normal-hearing adults typically require a $2\,\text{dB}$ signal-to-noise ratio advantage over noise.
• Etymotic Research Study (Killion): Mild-to-moderate SNHL listeners exhibit SNR losses ranging from $2\,\text{dB}$ to $20\,\text{dB}$.
• Signal-to-Noise Test Categories:
  • Fixed SNR Tests: Speech Perception in Noise (SPIN), Connected Speech Test (CST).
  • Adaptive SNR Tests: QuickSIN, Hearing in Noise Test (HINT), BKB-SIN.
• Bottom-Up vs. Top-Down Processing:
  • Words-in-Noise (WIN): Monosyllabic words assessing bottom-up auditory processing.
  • HINT and BKB-SIN: Rich semantic context used for evaluating top-down processing and Cochlear Implant (CI) candidacy.

Speech-in-Noise Testing Procedures

• Original SIN Test (Killion, $1997$): Used Institute of Electrical and Electronics Engineers (IEEE) sentences (derived from Harvard WWII lists), female talker, $4$-talker babble at $15\,\text{dB}$, $10\,\text{dB}$, $5\,\text{dB}$, and $0\,\text{dB}$ SNR; abandoned due to inefficient scoring and limited equivalent lists ($2$ lists).
• QuickSIN Protocol:
  • Stimuli: IEEE sentences, female talker, $4$-talker babble across $6$ SNR levels ($25\,\text{dB}$ to $0\,\text{dB}$ in $5\,\text{dB}$ steps).
  • Standard Presentation Level: $70\,\text{dB HL}$ (adjusted if Pure Tone Average or SRT $\ge 45\,\text{dB HL}$).
  • Formula: $\text{SNR Loss} = 25.5 - \text{Total Correct Key Words}$.
  • Test Lists: $12$ clinically equivalent lists.
  • Instruction Delivery: Must be read verbatim to minimize variability (supported by House Institute findings).
  • SNR Loss Severity Categories:
  • $0\,\text{dB}$ to $3\,\text{dB}$: Normal / near normal.
  • $3\,\text{dB}$ to $7\,\text{dB}$: Mild SNR loss.
  • $7\,\text{dB}$ to $15\,\text{dB}$: Moderate SNR loss.
  • $> 15\,\text{dB}$: Severe SNR loss (requires assistive listening devices / FM systems).
• Binaural Interference:
  • Massoy and Metler Study (Iowa): Approximately $17\%$ of the population demonstrates poorer binaural speech performance than monaural performance.
  • Protocol Order: Test binaurally first, followed by right ear, then left ear. Over $90\%$ of individuals display a right ear advantage due to left hemisphere language processing.
• BKB-SIN Protocol:
  • Designed for children, Auditory Processing Disorder (APD) evaluations, older adults, and CI candidates.
  • Stimuli: Bamford-Kowal-Bench sentences, male talker, background noise increases in $3\,\text{dB}$ steps (range from $+21\,\text{dB}$ to $-6\,\text{dB}$ SNR).
  • Administration: Requires paired lists (e.g., List 1a and 1b) per condition.
  • Formula: $\text{SNR 50} = 23.5 - \text{Total Correct}$.

Questions & Discussion

• Question on BKB-SIN Negative SNR Steps: Why does the BKB-SIN test extend down to $-6\,\text{dB}$ SNR when normal listeners need a $2\,\text{dB}$ SNR difference?
  • Response: BKB-SIN sentence material is more predictable than isolated words or IEEE sentences, so test designers extended the noise level to create adequate task difficulty for listeners with higher performance ceilings.
• Question on Scoring Strictness: How strictly are keyword targets scored?
  • Response: Scoring is exact; minor variations such as changing a word from plural to singular or vice versa are marked as incorrect.
• Class Logistics & Scheduling:
  • The class agreed to meet in the Wilson Room on Wednesday at $02:45$ to $03:45$ (or starting at $03:45$ depending on prior lecture end times) to review lecture materials and clinical protocols.