Prescriptive Fitting Targets (NAL-NL2 vs. DSL m[i/o]) & Dynamic Range

Prescriptive Fitting Targets (NAL-NL2 vs. DSL m[i/o]) & Dynamic Range

The fundamental goal of hearing aid electroacoustic fitting is to restore speech audibility across the audible frequency spectrum while maintaining auditory comfort and preventing acoustic trauma. Achieving this objective requires careful mapping of sound into the patient's restricted residual auditory dynamic range using scientifically validated prescriptive algorithms.

Dynamic Range of Hearing & Sensorineural Cochlear Loss

In healthy, normal-hearing human cochleae, the auditory dynamic range (DR) spans approximately $100$ to $120\,\text{dB}$ across speech frequencies, extending from the threshold of audibility ($0\,\text{dB HL}$) to the uncomfortable loudness level (UCL / LDL, typically $100-110\,\text{dB HL}$).

$\text{Dynamic Range (dB)} = \text{UCL (dB HL)} - \text{Hearing Threshold (dB HL)}$

Outer Hair Cell Damage & Loudness Recruitment

Sensorineural hearing loss (SNHL) primarily originates from structural damage to or loss of outer hair cells (OHCs) within the organ of Corti. Outer hair cells provide electromechanical motility (mediated by the motor protein prestin), acting as an active non-linear cochlear amplifier that boosts low-level sounds ($< 50-60\,\text{dB SPL}$) by up to $40-50\,\text{dB}$ while sharpening basilar membrane tuning curves.

When OHCs are damaged:

1. Threshold Elevation: Soft sounds are no longer amplified by the active cochlear mechanism, causing pure-tone threshold elevation ($30-70\,\text{dB HL}$). Soft speech ($45-55\,\text{dB SPL}$) becomes inaudible.
2. Abnormal Loudness Growth (Recruitment): As input levels rise to high intensities ($> 85-90\,\text{dB SPL}$), passive basilar membrane mechanics dominate. Passive fluid displacement stimulates remaining inner hair cells (IHCs) normally. Consequently, loud sounds are perceived just as loudly by an individual with SNHL as by an individual with normal hearing.
3. Dynamic Range Compression: Because the hearing threshold is elevated to $65\,\text{dB HL}$ while the UCL remains relatively unchanged at $100\,\text{dB HL}$, the patient's dynamic range collapses from $100\,\text{dB}$ down to only $35\,\text{dB}$.

Clinical UCL / LDL Measurement Protocol

• The Fallacy of Assumed UCLs: Fitting software frequently defaults to assumed population average UCLs ($100\,\text{dB HL}$ across all frequencies). Empirical measurements demonstrate that individual UCLs vary widely, ranging from $80\,\text{dB HL}$ to $> 115\,\text{dB HL}$. Relying on software defaults risks either under-amplifying loud speech inputs or causing acoustic pain and device rejection.
• Standard Clinical Measurement Protocol:
  • Test stimuli: Frequency-specific pulsed pure tones ($500, 1000, 2000, 4000\,\text{Hz}$) or speech-spectrum noise bursts delivered under calibrated insert earphones.
  • Rating scale: The Cox Contour Test for Loudness (Category 1: Very Soft; Category 2: Soft; Category 3: Comfortable, But Slightly Soft; Category 4: Comfortable; Category 5: Comfortable, But Slightly Loud; Category 6: Loud, But OK; Category 7: Uncomfortably Loud).
  • Procedure: Ascend in $5\,\text{dB}$ steps starting at comfortable listening levels ($65-70\,\text{dB HL}$) until the patient signals Category 7: Uncomfortably Loud. Re-test to verify reliability within $\pm 5\,\text{dB}$. The hearing aid Maximum Power Output (MPO / OSPL90) is programmed directly below this empirical boundary.

Wide Dynamic Range Compression (WDRC) Mechanics

Linear amplification applies a fixed, identical gain factor across all input levels. In a collapsed dynamic range, a linear hearing aid providing $35\,\text{dB}$ of gain to make soft speech ($50\,\text{dB SPL}$) audible would amplify loud conversational peaks ($85\,\text{dB SPL}$) to $120\,\text{dB SPL}$, causing loudness discomfort and potential acoustic trauma. Conversely, if gain is lowered to keep loud sounds comfortable, soft speech becomes inaudible. Wide Dynamic Range Compression (WDRC) resolves this trade-off by dynamically altering gain as a function of incoming signal intensity.

Compression Architecture & Parameters

Output (dB SPL)
   ^
   |                    Output Limiting (CR >= 8:1)
   |                     /--------------------------  MPO / OSPL90
   |                    / (High TK ~ 85-90 dB SPL)
   |                   /
   |      WDRC        /
   |   (CR 1.5:1-3:1)/
   |                /
   |  Linear       / 
   |  (CR = 1:1)  /
   |             /  <-- Low TK (~40-50 dB SPL)
   +------------+------------------------------------> Input (dB SPL)

1. Compression Threshold / Kneepoint (TK): The input sound pressure level at which the amplifier transitions from linear amplification to compressed gain reduction.
   • WDRC Systems: Utilize low threshold kneepoints ($\text{TK} = 40-50\,\text{dB SPL}$). This guarantees that the softest phonemes of speech receive maximum acoustic gain to clear the elevated hearing threshold.
2. Compression Ratio (CR): The ratio of decibel input change to decibel output change along the compression slope: $\text{CR} = \frac{\Delta \text{Input (dB)}}{\Delta \text{Output (dB)}}$
   • A linear circuit has a $\text{CR} = 1:1$ (a $10\,\text{dB}$ increase in input yields a $10\,\text{dB}$ increase in output).
   • A WDRC circuit typically uses mild-to-moderate compression ratios: $\text{CR} = 1.5:1$ to $3.0:1$. At $\text{CR} = 2:1$, a $10\,\text{dB}$ increase in input produces only a $5\,\text{dB}$ increase in output.
   • High compression ratios ($> 3.5:1$ to $4.0:1$) applied across wide bandwidths reduce speech contrast, flatten the temporal speech envelope, smear acoustic phoneme boundaries, and degrade speech recognition in noise.
3. Output Limiting Compression (Compression Limiting): Implemented at high input levels with a high threshold kneepoint ($\text{TK} \ge 85-90\,\text{dB SPL}$) and high compression ratio ($\text{CR} \ge 8:1$ to $12:1$). This strictly caps maximum output (OSPL90) below the patient's measured LDL while maintaining low harmonic distortion ($< 2\%$), replacing legacy peak clipping.

Temporal Compression Parameters: Attack & Release Times

• Attack Time ($t_a$): The time required for the compression circuit to reduce gain to within $2-3\,\text{dB}$ of its final steady-state value following an abrupt increase in input level ($> 10-20\,\text{dB}$). Standards mandate fast attack times ($5-20\,\text{ms}$) to prevent sudden acoustic transients (e.g., slamming doors, clattering dishes) from exceeding the patient's discomfort threshold.
• Release Time ($t_r$): The time required for the circuit to restore gain to within $2-3\,\text{dB}$ of its target value following a drop in input level.
  • Syllabic Compression (Fast Release: $50-100\,\text{ms}$): Gain recovers between syllables, amplifying weak consonants following loud vowels. Risk: Rapid gain changes amplify background noise during speech pauses, creating an audible breathing or "pumping" artifact.
  • Slow / Adaptive Compression (Release: $800\,\text{ms} - 2.0\,\text{s}$): Maintains stable gain across conversational sentences, preserving natural temporal envelope cues. Modern systems employ dual/adaptive release time algorithms: fast release for transient acoustic spikes, and slow release during continuous speech to maintain comfort and avoid pumping.

Prescriptive Target Philosophies: NAL-NL2 vs. DSL m[i/o]

Modern clinical practice relies on two primary prescriptive target paradigms. While both utilize multichannel WDRC, they are derived from fundamentally contrasting psychoacoustic philosophies.

1. NAL-NL2 (National Acoustic Laboratories - Non-Linear 2)

• Core Philosophy: Loudness Equalization & Intelligibility Maximization.
  • Developed by the Australian National Acoustic Laboratories (Dillon, Keidser, et al.).
  • NAL-NL2 calculates target gain across frequency bands such that each band contributes equally to overall perceived loudness. Instead of attempting to normalize loudness to match a healthy ear, NAL-NL2 balances loudness across bands to maximize the Speech Intelligibility Index (SII) while keeping overall perceived loudness at or below normal listening comfort.
• Desensitization Factor: NAL-NL2 incorporates empirical cochlear desensitization data: as hearing loss exceeds $65-70\,\text{dB HL}$, damaged inner hair cells and auditory nerve fibers lose the capacity to extract useful spectral and temporal information. Pumping high gain into severe hearing loss bands causes physiological distortion. NAL-NL2 deliberately compresses or limits target gain at frequencies with severe-to-profound loss, focusing amplification where residual hearing can extract intelligible speech cues.
• Empirical Patient Adjustments:
  • Bilateral Fitting Correction: Because central binaural loudness summation increases perceived loudness by $2-3\,\text{dB}$ at low-to-moderate levels and up to $5-6\,\text{dB}$ at high input levels, NAL-NL2 reduces prescribed gain by $\approx 2-3\,\text{dB}$ per ear when programmed for bilateral use.
  • Age Adjustment: Recognizes that adults over 60 years demonstrate reduced auditory processing and intolerance for excessive high-frequency loudness; prescribes slightly reduced gain for older adults.
  • Gender Correction: Prescribes approximately $1-2\,\text{dB}$ more gain for adult males than females based on empirical loudness preference differences.
  • Language Structure (Tonal vs. Non-Tonal): Prescribes slightly greater low-frequency gain ($< 1\,\text{kHz}$) for tonal language speakers (e.g., Mandarin, Cantonese) to preserve fundamental frequency pitch contours critical for lexical meaning.
  • Experience Level: Prescribes slightly reduced gain ($2-4\,\text{dB}$) for new, inexperienced hearing aid users, transitioning to full prescriptive targets via adaptive acclimatization algorithms.

2. DSL m[i/o] (Desired Sensation Level Multistate Input/Output)

• Core Philosophy: Loudness Normalization across Sensation Levels.
  • Developed at the University of Western Ontario (Seewald, Scollie, et al.).
  • DSL operates on a loudness normalization philosophy: it aims to make soft sounds sound soft, conversational sounds sound average, and loud sounds sound loud, recreating the normal loudness growth curve within each individual frequency band.
• Most Widely Used Pediatric Prescription:
  • DSL v5.0 is the most widely used prescription for infants and children in North America. The AAA (2013) pediatric amplification guideline calls for a validated, age-appropriate prescriptive method, such as DSL v5.0 or NAL-NL2.
  • Developmental Rationale: Unlike adults, who have established central phonemic representations, children cannot use top-down cognitive context to fill in missing acoustic cues. Children require access to the entire acoustic speech spectrum—including weak, high-frequency consonant fricatives ($/s/, /z/, /\theta/$)—to acquire spoken language, learn morphological markers (plurals, possessives, verb tenses), and monitor their own vocal feedback.
  • DSL usually prescribes higher overall gain than NAL-NL2 for the identical audiogram (often several decibels more, especially in the low and mid frequencies) across soft, average, and loud input levels.
  • Incorporates child-specific Real-Ear-to-Coupler Difference (RECD) corrections to account for the smaller acoustic volume and higher resonant frequency of the infant/child ear canal.