The Desired Sensation Level bone conduction device prescriptive formula: new adaptations and applications for skin-driven bone conduction fittings

Abstract

This manuscript describes the computational workflow and normative data used within the newly released Version 2 of the Desired Sensation Level (DSL) Bone Conduction Device (BCD) Prescription. The previous version 1.1 of the DSL BCD prescription was an innovation in the field, allowing for the prescription of output levels from direct drive BCDs worn percutaneously on an osseointegrated abutment. Verification of these prescriptions has been possible using a clinically-available skull simulator. Version 2 of the DSL BCD prescription extends upon version 1.1 and now allows for the fitting and verification of some skin-drive BCDs, which is the primary coupling method used for infants and young children. DSL BCD 2 introduces a new age-specific audiometric-to-dial-level (ADL) correction to predict in-situ BCD thresholds from audiometric bone conduction thresholds, as well as other transforms that allow target generation for skin-worn devices. This implementation provides the option to base the output prescription on audiometric thresholds for patients who may not yet be capable of performing in-situ BCD audiometry, computational support for a wider range of device types and couplings, and prescriptive targets for use with either skull simulators or verification of on-head force levels.

1 Introduction

People who experience conductive or mixed hearing losses may be candidates for bone conduction hearing devices (BCDs; Håkansson et al., 2019; Snik et al., 2005). BCDs provide incoming sound to the user by converting sound input into skull vibrations to bypass some or all of the conductive mechanisms of the outer and middle ear, and directly stimulating the cochlea. The level and spectral shape of these devices are programmed in clinical practice to provide speech audibility to the patients who will wear the devices. The purpose of this paper is to describe enhancements to a prescriptive algorithm that aims to compute recommended output levels for BCDs.

Currently, there are several categories of BCDs available, including: (a) direct-drive with percutaneous vibration, (b) non-invasive transcutaneous devices also known as skin-drive devices, and (c) direct-drive transcutaneous induction devices (Håkansson et al., 2019). Direct-drive percutaneous devices have a connection that is surgically implanted into the mastoid, through the skin. Energy is sent from an external sound processor through an abutment to an osseointegrated implant. In contrast, transcutaneous devices send either a transduced signal or vibrations across the skin and soft tissue. Non-invasive skin-drive transcutaneous BCDs transmit the signal though vibration, which passes through skin and soft tissue to the bone, and thereby to the cochlea. Transcutaneous induction devices pass a signal to an implanted transducer, which transmits vibration to bone through either electromagnetic or piezoelectric transduction. These add energy to the signal transmission and bypass skin and soft tissue attenuation. Skin-worn devices have an external processor that is held in place over the intact skin either through a magnet that is surgically-implanted beneath the skin, or using a device-specific soft headband, a device-specific hard headband (e.g., Cochlear SoundArc^®) or an adhesive adapter (e.g., ADHEAR^®).

Infants and young children are often fitted with non-surgical devices, and therefore are more likely to use non-invasive skin-drive BCDs, worn with child-friendly couplings, such as soft headbands or adhesive adapters. This application allows for device usage and access to sound until the child meets the skull thickness requirements to become candidates for surgically-implanted BCD options (Bagatto et al., 2022; Gordey and Bagatto, 2021; Ricci et al., 2011; Skarżyński et al., 2019; Snik et al., 2005; Willenborg et al., 2023).

Historically, BCD fitting approaches have often relied on the measurement of aided sound field thresholds. However, the limitations of using behavioral measures for hearing device fitting are well-documented and include limited test-retest reliability, noise floor effects, the inability to test at moderate or high levels, and the requirement that the device user be able to provide valid responses over a sustained period of testing (Hawkins, 2004; Stelmachowicz et al., 2002). This patient-driven feedback loop is not feasible for use in infants or young children, and some well-established BCD research suggest that moving to objective metrics of gain or objective measurement of aided brain activity may improve fitting consistency (Snik et al., 2019). Overall, the drawbacks of behavioral fitting methods created a need to develop a more objective, instrument-driven approach to BCD prescription and verification, similar to evidence-based clinical guidelines used with air conduction hearing aids (e.g., Ching et al., 2013; Galster et al., 2021, 2022; Wiesner et al., 2012). The conceptual purpose and benefits of hearing aid prescription and verification are well-established for both children and adults who use air conduction hearing aids (Amlani et al., 2017; Hodgetts et al., 2011; McCreery et al., 2013, 2015, 2017; Seewald et al., 2005; Valente et al., 2006; Amri et al., 2022; Marriage et al., 2018; Valente et al., 2006, 2018; Wiesner et al., 2012).

In 2017, Hodgetts and Scollie introduced an output prescription algorithm for clinical use with BCD verification based on version 5 of the Desired Sensation Level (DSL) Method, a validated prescription and verification algorithm originally developed for use with air conduction hearing aids (Scollie et al., 2005). The resulting DSL BCD Version 1.1 (v1.1) calculates output targets in decibels force level (dB FL) for use with direct drive percutaneous BCDs (Hodgetts and Scollie, 2017). This was later implemented within clinically available hearing aid test or programming systems. The targets are based on thresholds obtained using the patient's own BCD device worn on the surgically-implanted abutment (i.e., in-situ thresholds). The in-situ thresholds are used to calculate prescriptive targets for both speech and maximum output, and the prescription algorithm allows adaptation for use with multichannel compression and different target listening levels for use with adults and children, and in quiet and noisy environments. To verify the output levels of the device, the output of the BCD is measured on a skull simulator within a hearing aid test chamber and fine-tuned to match targets. By using a common reference point of dB FL at the percutaneous abutment, the measured BCD output can be mapped together with the patient's dynamic range and the prescribed targets onto a single visualization known as the force-level-gram (FLogram). This FLogram strategy mimics the commonly-used SPLogram visualization that is widely adopted for fitting air conduction hearing aids. This combination of prescriptive targets and commercially-available verification tools allows clinicians to measure and adjust the BCD output using procedures that are similar to those used in air conduction device fitting. Evaluations of the force levels prescribed by an earlier version of the DSL BCD prescription showed improved aided speech recognition outcomes compared to when devices were set at patients' preferred listening levels (Hodgetts et al., 2011). In current practice, BCD fittings to DSL BCD v1.1 targets support fine tuning of the device to provide output levels that are similar to the prescribed targets, unless device limitations are reached (Gascon et al., 2025). This highlights the importance of using fitting approaches and device selection procedures that are focused on maximizing audibility on an individual basis. The DSL BCD prescription aims to facilitate this process.

The DSL BCD v1.1 prescriptive algorithm is limited in that it cannot generate targets for non-invasive skin-drive BCDs fitted on soft headbands or direct-drive transcutaneous induction hearing aids. DSL BCD v1.1 was intended for use with adults and children fitted with direct-drive percutaneous BCDs. The computational workflow in v1.1 assumes that the patient is able to be tested using in-situ BCD audiometry, and therefore corrections from audiometric HL to prescription are not included. In summary, DSL BCD v1.1 does not support a wide range of BCD fitting types, nor does it facilitate prescription from conventional audiometric thresholds. With the increasing availability of surgical and non-surgical skin-drive BCD options, the limitations of DSL BCD v1.1 have become more relevant, particularly for infants and young children who require amplification to support spoken language development while waiting to meet surgical candidacy criteria (Håkansson et al., 2019; Hol et al., 2005; Wang et al., 2018). Some groups have begun to investigate the use of DSL BCD v1.1 with skin-drive devices (Griffin et al., 2023; Sun et al., 2024), which indicates interest in having a purpose-developed solution for other device categories. In order to meet this need, the purpose of this paper is to propose specific modifications to DSL BCD v1.1 to accommodate prescription and verification for non-invasive skin-worn BCD devices.

2 Method

2.1 Software development

Software development followed a standard iterative process (Royce, 2021). This cyclical yet stepwise approach is appropriate for use when a software product is iterated across multiple rounds of testing and feedback, often using a trial-and-error methodology that includes critical evaluation of prototype versions to check the impact and/or accuracy of modifications. This process begins by defining requirements both for system and software, followed by analysis and coding, with subsequent testing and revision until a fully operational version is achieved. Documentation development occurs in parallel with software iteration, and documentation is intended to support usage by personnel other than the developers.

In this project, the focus was on the iteration stage, because the goal was to expand the functional range of the DSL BCD v1.1 software. Flowcharting of the computational workflow and tables of correction values were developed in Microsoft Word and Excel, for review and discussion by the interdisciplinary team of audiologists and software engineers in order to develop consensus on the scope and nature of the iteration, as well as to determine available evidence that would support specific changes. The Excel sheet went through five rounds of iteration as the team incorporated devices and data for the specific corrections described below (Authors and contributors Scollie, Bagatto, Hodgetts, El-Naji). A technical description document was originally developed in 2016–2017 for DSL BCD (Authors Scollie, Beaulac) and updated with the specific iterations through 2024 (Authors Scollie, Bagatto, Hodgetts). Two audiologist team members with experience in both BCD fitting and software development (Authors Beh, Banerjee) provided critical review of the technical description to ensure completeness and clarity. Software development iterated the dynamically-linked library (DLL) in the C programming language (Authors Beaulac, Holden), with version tracking as a branch on a self-hosted git service (Gitea), with associated documentation in HTML format. Because the DLL does not provide a user interface, a web-hosted test application was developed (Authors Holden, Beaulac) using Rust and Typescript (Svelte) to allow team members to enter test cases with graphical visualization of both the assessment data and prescriptive targets. The functionality of the test application was reviewed for accuracy and workflow by audiologists familiar with the technical documentation and with clinical BCD fitting (Authors Scollie, Brown, Hodgetts). Integration of the iterated workflow into related clinical protocols by team members with clinical experience in training and BCD fitting (Authors and contributors Beh, Brown, Bagatto) provided further opportunities to review for completion, clarity, and consensus. The final technical specification is described below under Algorithm Development. The primary construct was inclusion of workflows for non-invasive skin-drive BCDs, based on evidence review of key factors to consider.

2.2 Evidence review and key factors

Non-invasive skin-drive bone-conducted sound transmission is influenced by a multitude of patient- and device-specific factors that differ from those of direct-drive percutaneous BCD fittings. These factors include individual skin thickness differences, transducer characteristics, and coupling type (see review by Gascon et al., 2022). Studies have characterized the attenuation of bone-conducted sound when skin and soft tissue are added to the transmission pathway by comparing BCD output or audiometric thresholds in percutaneous vs. transcutaneous coupling conditions (Mattingly et al., 2015; Håkansson et al., 1985; Mylanus et al., 1998; Gascon et al., 2022). Sound transmission is attenuated when transmitted through skin and soft tissue, particularly in the higher frequencies where reductions of up to 28 dB have been observed (Håkansson et al., 1985). This suggests that device output needs to be higher for skin-drive BCD fittings, compared to percutaneous fittings, in order to overcome the skin transmission loss associated with these coupling conditions.

As one example, bone conduction thresholds are affected by the skin-drive properties of non-invasive transcutaneous BCDs and these devices are transducer-dependent. Three recent studies (Gascon et al., 2022; El-Naji et al., 2024, 2026) compared skin-driven thresholds measured using an audiometric bone conduction oscillator, to in-situ thresholds measured using skin-drive BCDs on soft headbands in adult listeners. Intra-subject comparisons within each study demonstrated that in-situ BCD thresholds were elevated (worse) compared to audiometric BC thresholds, and this was device-specific, despite the skin properties remaining constant within each subject across test conditions. These audiometric-to-in-situ differences were frequency-dependent but were greater for higher frequencies where threshold shifts of up to 40 dB were noted (Gascon et al., 2022; El-Naji et al., 2024). Similar observations have been made in the pediatric population as well (Bagatto et al., in review). Possible reasons for these differences were discussed by (Gascon et al. 2022), and included various transducer-specific characteristics, such as contact force, contact area, and the reference equivalent threshold force level (RETFL).

DSL BCD v1.1 computes targets only based on in-situ thresholds for abutment-worn BCDs. Adaptation of DSL BCD v1.1 to include a greater range of use cases could consider broadening the range of BC audiometric transducers (i.e., audiometric or in-situ) used to obtain thresholds. This would provide the clinical option of basing the prescription on audiometric data, particularly for young infants and children where assessment approaches can vary depending on developmental status, where in-situ thresholds using the BCD may not be possible in the early stages of assessment and intervention. Ideally, BC thresholds should be measured in-situ as a way to capture the patient's own individual skin transmission loss and device-specific characteristics and coupling with skin-drive BCDs (Bagatto et al., 2022). However, for cases where in-situ thresholds using the skin-drive BCD are unavailable, clinical alternatives are required to support accurate BCD prescription based on audiometric BC thresholds alone in addition to incorporating the loss of energy through skin and soft tissue.

One clinical scenario may be expected to have audiometric BC thresholds without available in-situ thresholds: the testing of infants and young children who require testing either by electrophysiology or by conditioned behavioral response procedures. It is well-established that several differences are expected in this developmental range. First, larger step sizes are used in bracketing threshold. Second, fewer test frequencies are typically measured due to constraints on test time and child responsivity, leading to a partial audiogram. Third, the lower limit of tested levels may be limited to the child's Minimum Response Level (MRL), which due to developmental factors may be at a slightly suprathreshold level (Norrix, 2015; Sabo et al., 2003). Fourth, results of threshold estimation from electrophysiological responses normally use corrections to account for the differences between nHL and HL (Bagatto et al., 2024). This difference is normally accounted for in air conduction prescriptions for use with infants (Bagatto et al., 2023), and similar considerations may apply for BCD prescriptions. Taken together, these factors may indicate that different computational pathways are warranted for corrections applied to either adult or pediatric audiometric thresholds, when predicting in-situ thresholds for the purposes of skin-worn BCD prescription.

Such considerations were key design goals that drove iteration to produce version 2 of the DSL BCD prescriptive formula, described in this paper. In the sections below, the computational pathways of corrections to support the prescription of output from non-abutment-worn BCDs, which are primarily used with infants and young children, are described.

3 Algorithm development

3.1 Technical description of DSL BCD v2

The DSL BCD v2 prescription is an extension of the DSL BCD v1.1 algorithm, in that it uses many of the same key principles for fitting percutaneous devices, but with adaptations that may enable use with non-invasive skin-drive fittings and direct bone transcutaneous induction fittings. Key concepts and features of DSL BCD v1.1 that are maintained in DSL BCD v2 will be described in brief here; additional details can be found in (Hodgetts and Scollie 2017). Added features that are specific to DSL BCD v2 will be described below, with specific adaptations for non-invasive skin drive BCDs. Summaries of key variables and computational workflow are displayed in Table 1 and Figure 1.

Figure 1

3.1.1 Display type and the auditory area

3.1.1.1 Display type

In DSL BCD v1.1, we proposed a targets and verification display format that emulated visualizations used when fitting air conduction hearing aids (Hodgetts and Scollie, 2017). Briefly, the commonly-used SPLogram display was modified for use in dB FL, generating the “FLogram” plot that places bone conduction thresholds, targets, upper limits of comfort, and device output on one display (see Figure 2 for examples). This display type is now available in clinical fitting equipment, and is used in research studies that investigate verification equipment (Nie et al., 2022; Persson et al., 2023, 2025).

Figure 2

The variables on the FLogram were originally conceptualized with reference to dB FL at the percutaneous abutment (Hodgetts and Scollie, 2017). Transforms were defined to move key variables to and from the abutment, allowing verification to take place on a skull simulator and incorporation of normative hearing detection values. One key advantage of defining target levels on the percutaneous abutment is reference to suprathreshold loudness characteristics for BCDs, which have mainly been developed with percutaneous BCD users (e.g., Kurz et al., 2014; Stenfelt and Håkansson, 2002; Stenfelt and Zeitooni, 2013). In considering modification of this strategy for use with skin-drive BCDs, we maintain the strategy of computing in dB FL on the abutment, using transforms that allow target development and verification in other locations, as described below.

3.1.1.2 Threshold definition

The lower limit of the auditory area is used as a key variable in computation of prescriptive targets (Figure 1). This lower limit on the FLogram is graphed as the detection threshold per frequency, for the device and coupling type that is used by the wearer. In DSL BCD v1.1, this lower limit was defined only for percutaneous BCDs and required thresholds to be obtained using an in-situ approach rather than via conventional audiometry. This assumes the patient is developmentally capable of completing an in-situ audiogram, where manufacturer-specific fitting software is used to deliver frequency-specific stimuli through the BCD while coupled to the patient's abutment. The use of in-situ threshold measurement remains the preferred approach, for reasons of accuracy (Bagatto et al., 2022). However, because many users of skin-drive BCDs are infants and young children, one goal for DSL BCD v2 was to expand this threshold definition to incorporate input data from conventional audiometry, in order to provide as many assessment options as possible. Corrections to support this strategy will be summarized below.

For skin-worn BCDs, version 2 allows for threshold data entry for the following audiometric bone oscillator test procedures to be used in threshold entry: standard audiometry (HL Standard), pediatric audiometry, or auditory brainstem response (eHL or nHL ABR). Thresholds entered using these audiometric procedures completed with standardized transducers receive an audiometric-to-dial level (ADL) correction. The ADL is applied to audiometric thresholds from any standard audiogram to derive a prediction of the in-situ dial level thresholds (Equation 1). The ADL values are stored in a lookup table, which contains unique values for each of the audiometric data types, and can use device-specific values when these have been determined from normative studies. Similarly, the nHL to eHL correction is stored in a lookup table as a separate correction, and used only if the assessment was performed in nHL:

ADL values for use with the adults and children were derived from separate studies, each of which are briefly summarized below. Separate ADL values were used for adults, children, and across different device types for two reasons. First, the evidence behind each correction was study-specific, with different data sources informing only specific populations and/or devices. Second, the clinical application of any ADL must work with the type of audiometric data that is collected, and we anticipated that audiometric thresholds from very young children would be impacted by nHL to eHL factors and/or MRL factors, consistent with the well-established literature on children's audiometric thresholds, while adult thresholds would not. We therefore aligned correction procedures with the population type that generated each subtype of ADL.

3.1.1.2.1 Pediatric corrections

ADLs for the pediatric population were derived from a structured retrospective file review (Bagatto et al., in review). In this review, files were collected from three clinics who followed a common protocol for in-situ audiometry with young children who used one of six different brands/models of pressure-attached skin-worn BCDs (Bagatto et al., 2022). Their standard audiometric thresholds were obtained either through conditioned procedures including Visual Reinforcement Audiometry (VRA) or Conditioned Play Audiometry (CPA), or through standard audiometry. Approximately half of the 29 ears were tested using some form of conditioning, while the others were tested with standard audiometry. These clinical thresholds were compared to their later in-situ bone conduction audiograms, and the average differences were used to develop corrections that convert audiometric values to estimated in-situ values. The in-situ and standard audiometry values were tested, on average, within 6 months of one another, but varied in how closely in time they were measured. The differences between audiometric and in-situ thresholds varied about the mean difference by about 10 dB across frequencies, but the mean trend of differences across frequency was observed for most cases, including a larger correction required at 4,000 Hz than at other test frequencies. The average differences were rounded to the nearest 5 dB for use as pediatric ADL corrections (Table 2). Further details of this study are available in a companion paper (Bagatto et al., in review). Importantly, these corrections were obtained with audiometers that were calibrated with Reference Equivalent Threshold Force Levels (RETFLs) used with ANSI standards from 2004 through 2018 (ANSI/ASA, 2018). Subsequent changes to ANSI standards for RETFLs (ANSI/ASA, 2025; Margolis et al., 2024) also need to be considered since restandardization occurred in 2025, as discussed later in this paper.

3.1.1.2.2 Adult corrections

For adults tested with standard audiometry, ADL corrections were derived from a recent study of Equivalent Force Level Thresholds (ETFL, El-Naji et al., 2026). The El-Naji sample used three BCDs worn by 20 young adults on a pressure-fitted headband attachment. A second sample was also used to quantify ADLs for adhesive devices. These were derived from a recent study with adhesive-worn BCDs fitted to 36 young adults with normal hearing (Webster et al., 2026). The threshold differences were averaged across the participants from each study. The mean differences were rounded to the nearest 5 dB and then interpolated and extrapolated to other frequencies to produce the ADL corrections used in DSL BCD v2 (Table 2). Some of these ADL corrections are specific to a particular device (i.e., adhesive-worn devices) while others are averages across device categories (i.e., headband-worn devices that are also abutment-compatible). Rounding to the nearest 5 dB was done to align with common audiometric step sizes, which is the data collection resolution underlying the source dataset of the ADLs. Similar to the pediatric corrections described above, these ADL values were derived with pre-2025 ANSI values for RETFL, so should be modified if used with audiometric data derived from audiometers calibrated using ANSI S3.6-2025 values.

Modifications for either adult or pediatric ADL use for the new 2025 audiometric calibration standard have also been developed, to take into account the RETFL differences introduced in the 2025 RETFL changes. The differences between the pre- and post-2025 RETFL values for Mastoid placement were computed as a level-independent correction, and summed with the ADL values to create values appropriate for use with either the 2018 or 2025 S3.6 standards. These updated ADL corrections were limited to zero and rounded to the nearest 5 dB step (Table 2). This strategy is intended as a temporary approach while clinics transition between the two ANSI standards, as children's audiometric records over time will contain thresholds measured either before or after the update to calibration. The differences omit computation of the 250 Hz threshold, and will increase measured audiometric thresholds in the high frequencies, with noticeable changes at 4,000 Hz and above.

The resulting in-situ thresholds, whether estimated or measured directly, include attenuation due to transmission via skin-drive. As such, we expect that in-situ thresholds measured via skin-drive will incorporate the skin and soft tissue loss, and provide a personalized adjustment to the lower limit of the auditory area. The higher thresholds obtained through either skin-drive in-situ audiometry or the use of ADLs are the primary means of accounting for skin-drive in DSL BCD v2. A simulated pediatric case illustrating this effect is visualized in Figure 2. This figure shows the same case, tested using in-situ audiometry through a percutaneous abutment, vs. when tested through skin drive and at the developmental stage of using VRA for audiometry. The pediatric ADL has been incorporated into the right panel of Figure 2, causing an increase in threshold curves in Force Level as would be expected when attenuation through skin is incorporated. The targets are mapped to take this lower limit of audibility into consideration, resulting in higher prescribed target output levels.

Because of the skin and soft tissue loss differences between percutaneous and skin-drive audiometry, we have added a variable to DSL BCD v2 to specify the BCD Coupling Type, to tag whether the patient will be wearing the BCD on an abutment, via skin-drive with pressure (e.g., Baha^® or Ponto^® devices worn on headbands), or via skin-drive with adhesive attachment (e.g., ADHEAR). This tag supports characterization in clinical documentation. For example, it would allow differentiation of patient records across the transition from skin-drive to percutaneous device use, even when the same external processor is used in both configurations.

The in-situ thresholds are expressed in decibels Dial Level. These are converted into dB FL, in one of two locations. First, the Dial-to-Force-Level (DFL) transform can be used to convert the thresholds to dB FL on a skull simulator. This DFL correction uses the same published DFL transform for all devices (Carlsson and Håkansson, 1997). This DFL value is within about one audiometric step size of other published DFLs (Flynn and Hillbratt, 2012). Next, the Real-Head-To-Coupler-Difference (RHCD) can be added to convert thresholds to dB FL on the percutaneous abutment, which is sometimes characterized as on head, real head, or real bone (Ebrahimian et al., 2026; Hodgetts et al., 2018; Persson et al., 2023) values. Together, these two types of FL values parallel the on-coupler and on-ear values that are used for air conduction hearing aids; this distinction was an intentional design feature of DSL BCD aimed at facilitating objective device verification, and has been discussed previously (Hodgetts and Scollie, 2017). We have used the terms Real Head Force Level (RHFL) and Skull Simulator Force Level (SSFL) to label them in software development. For example, Equation 2 computes the RHFL at Threshold, based on the Dial Level from in-situ audiometry:

Typically, the RHCD values are device-specific, and are therefore defined for each percutaneous BCD. In DSL BCD v1.1, it was not possible to transform from skull simulator to abutment force levels for devices that were worn without an abutment. In DSL BCD v2, for devices that are not coupled to an abutment, a generic RHCD was developed (Figure 3) to facilitate verification of devices at the on-abutment measurement location even when a device-specific RHCD is unavailable. To develop this generic RHCD, we compiled brand-specific, model-specific RHCDs from 17 BCDs produced by both Cochlear and Oticon Medical. The RHCD reference values for each make and model of BCD were provided by the manufacturers as reference values for use in prescription and in verification, and were referenced for use with the Verifit Skull Simulator. Brand-specific median values were derived, and the midpoint between the two brands was computed for use as a generic RHCD transform. This RHCD can be used as a default value in Equation 2, when a measured value is not available. This allows transformation of targets and/or verification curves to the RHFL measurement scale, and aligns with the calibration point of other measurement devices such as the Real Bone Measurement system described elsewhere in this issue (Ebrahimian et al., 2026).

Figure 3

3.1.1.3 Upper limit

Variables defining the upper limit of the auditory area in DSL BCD v2 are computed using procedures developed for DSL BCD v1.1. Briefly, the algorithm continues to select the lower of either: 1) the patient's loudness discomfort levels (LDL), or 2) the device-specific maximum force output (MFO), to be the upper limit of the patient's dynamic range. Clinicians can enter measured patient LDLs if available. If LDLs are not available, the DSL-calculated estimate of the LDL will be based on patient thresholds, predicted via non-linear equations from DSL v5 (Scollie et al., 2005). The resulting dynamic range is then further compressed to account for the expected loudness growth with BCDs (Stenfelt and Håkansson, 2002; Stenfelt and Zeitooni, 2013). Manufacturer-provided MFOs are stored in a lookup table, and accessed based on the make and model chosen for each fitting. If the device-specific MFO is lower than the predicted LDL, the device's dynamic range will be the main factor that limits the upper limit of the auditory area, leading to clinical recommendations to select devices with high MFOs whenever possible, to minimize limitations on prescribed target output (Bagatto et al., 2022; Hodgetts and Scollie, 2017). The impact of choosing a higher-MFO device on target mapping is illustrated in Figure 4. This figure visualizes two sets of prescribed output for the same set of thresholds: the device labeled “a” is an older-generation device with a lower MFO, while the newer device “b” has a higher MFO response. Both devices have MFOs that are lower than the predicted LDL for this case, and therefore the targets for soft, conversational, and loud speech have been mapped using higher compression ratios for the device with the smaller output dynamic range. These differences are minimal for soft speech (3.5 dB averaged across 500–4,000 Hz), and increase as the input level increases to a 7 dB difference between devices for loud speech. Sound access is present for both fittings, mainly because the MFO accommodation in the prescription is driven by the maximum output, while the threshold characteristic remains constant.

Figure 4

From a loudness compensation perspective, this approach uses full compensation (100%) for skin loss at threshold, while the compressive aspect of the non-linear equations use a smaller ratio as the signal level increases, to about 80% at LDL for a listener with a normal detection threshold by bone. No further skin loss correction is incorporated into the LDL prediction, so the compressive nature of these targets may provide a more compressive approach if they under-predict the listener's LDL. Given the large range of error associated with predicting average skin loss vs. individual skin loss, this conservative computational strategy produces a lower output at MFO. Because the revisions are, in part, aimed for use in pediatric audiology, this conservative strategy was selected for DSL BCD v2.

3.1.2 Mapping across the auditory area

Once the auditory area is defined for the patient, target generation in DSL BCD v2 follows the same logic as DSL BCD v1.1. Briefly, the Target Type must be specified as being for either adults, assuming that the hearing loss was acquired post-lingually, or for children, assuming that the hearing loss was acquired during the years of language development. The children's prescription has most often been used for children with congenital hearing loss, and prescribes the higher listening levels associated with children's requirements for both performance and preference (see review by Scollie et al., 2005, see outcomes described in Scollie, 2008; Scollie et al., 2010a,b; McCreery et al., 2015, 2017). The Program Type allows either the Adult or Child target to be adapted for listening in noisy environments, where a higher compression threshold and an adjusted frequency-importance weighting are used to reduce listening levels somewhat, improving comfort for listeners when in background noise (Crukley and Scollie, 2012; Glista et al., 2021). Fittings for adults can be further adapted for monaural or binaural listening, with lower listening levels prescribed if the fitting is bilateral. As with DSL v5 for air conduction, bilateral adult targets for speech are reduced by 3 dB across frequencies.

3.1.3 Displaying targets for use in clinical verification systems

As with any DSL prescription, once all parameters are specified, a compressive algorithm is applied to produce a set of input-output curves for each frequency (Scollie et al., 2005). The input-output curves in DSL BCD prescriptions use a loudness growth function that is adapted for hearing by bone conduction, as described above. Output targets for speech inputs can be derived from the input-output curves; separate input-output curves are used to generate targets for high-level tones so that clinical verification of maximum output can be completed. The prescribed target outputs can be mapped onto the FLogram in the context of the patient's auditory area for various input levels, including soft, average, and loud running speech. Software adaptations are available to customize the compression threshold, crossover frequencies for compression channels, and to use custom-defined input spectra. All speech targets are expressed for one-third octave bands of speech.

Depending on the type of BCD and BCD test equipment, targets may be required for use with a skull simulator or for on-abutment verification. Equation 3 describes the relationship between on-abutment targets and skull simulator targets. This equation can be used to convert between skull simulator and on-abutment locations for target displays on the FLogram. The BCD output can then be verified and fine-tuned either on-abutment or on a skull simulator and plotted on the FLogram alongside targets. Conversion of values for verification displays should also apply Equation 3 as needed to align with the displayed target type. Using this strategy, a common reference point of dB FL allows comparison of the fitting to the targets, facilitating clinical decision-making about whether the device has adequate output and informing fine-tuning of device settings.

4 Results

The transforms, normative data, and adaptations described above have been programmed into a Dynamically-Linked Library (DLL) that can be embedded into the software systems used to both program and verify hearing aids. The clinical application of these software developments are illustrated using a series of case examples in which the DSL BCD v2 software has been integrated into a clinical verification system and used to simulate fittings that demonstrate common use cases, with a focus on pediatric practice. Each case below was tested using the Speechmap feature of the commercially-available Audioscan Verifit 2 (Software Version 2024.1.3) hearing aid test system, which has implemented the DSL BCD v2 prescriptive formula, and the Skull Simulator (Ebrahimian et al., 2026). Case fittings were completed using a Baha 6 Max (Cochlear Limited) and a Ponto 5 SP (Oticon Medical) sound processor, both of which are the most recent BCDs available from each manufacturer at the time of this writing. Any gain adjustments needed to fit to prescriptive targets were made using each device's respective fitting software (Baha Fitting Software 6.1; Genie Medical 2023.1). Case examples are not based on real patients.

4.1 Case examples

4.1.1 Case 1 – Conductive hearing loss

Case 1 is a 12-month-old infant who was diagnosed with bilateral atresia at birth. She was fitted with two Baha 6 Max devices at 2 months of age using her available ABR bone conduction thresholds and is now a consistent user of her devices. At 12 months of age, an in-situ BCD audiogram was completed using VRA assessment methods to obtain thresholds for device programming (Table 3 shows assessment results for the left ear only). The appropriate parameters are selected in the Verifit 2 to perform electroacoustic verification (Table 4). The device was re-fitted to DSL BCD v2 Child Targets. Figure 5 shows the resulting fitting for the left ear only.

Figure 5

4.1.2 Case 2 – mixed hearing loss

Case 2 is a 24-month-old child who was diagnosed at birth with a mild bilateral high frequency sensorineural hearing loss through ABR. He has had ongoing middle ear dysfunction for the past 8-months that eventually led to a perforated eardrum on the right side. Given that his ear canals were determined to be too small to attempt a tympanoplasty procedure, he has a mixed hearing loss. His parents have opted to pursue a BCD fitting on a soft headband to ensure he has optimal access to speech until he is a surgical candidate. An in-situ BCD audiogram was attempted, however, Case 2 did not comply with testing using the softband during the appointment. Audiometric thresholds obtained from a recent VRA assessment were therefore used to program and verify the Ponto 5 SP device for use until in-situ thresholds could be measured (Table 5). The audiometric thresholds are entered into the Verifit 2 for electroacoustic verification along with the appropriate fitting parameters (Table 6). The resulting fitting to DSL BCD v2 Child target is shown in Figure 6 and indicates a match to targets between 500 and 3,000 Hz to within 5 dB RMSE. However, due to device output limitations, targets could not be met above 3,000 Hz.

Figure 6

5 Discussion

This paper provides a technical description of the modifications made to the DSL BCD prescriptive algorithm to accommodate prescription-based fitting of transcutaneous BCDs. With DSL BCD v2, clinicians can specify both pressure and adhesive coupling types for skin-worn BCDs in addition to the abutment coupling that was supported in DSL BCD v1.1. Clinicians can also prescribe force level outputs based on a wider range of assessment methods used to obtain threshold information for the fitting, because prescription based on audiometric thresholds is now supported. Specifically, audiometric-to-dial-level (ADL) corrections were incorporated into DSL BCD v2 to account for known threshold differences between audiometric BC and

in-situ BCD methods, including response differences between young children and adults. The resulting pediatric ADL correction is based on a sample of children who varied in age and test type, and as such is a generic pediatric correction that was based on in-situ thresholds for children who were, on average, seven years old. The differences between audiometric and in-situ thresholds could be influenced by headband tension, skull maturation, audiometric step size, conditioning and reinforcement procedures, child state and developmental status, and adult-child differences in any of these factors. One primary developmental difference is that younger children, particularly those who are tested with conditioned procedures, may test to a Minimum Response Level (MRL) rather than to audiometric threshold. Future studies could attempt to re-quantify this correction to be more specific to the MRL properties of testing in early childhood (Norrix, 2015; Sabo et al., 2003). Adult-derived ADL corrections are also available, with separate corrections for use with adhesive vs. pressure-worn devices, given the different couplings that have been tested to date. Based on the age ranges tested in ADL studies to date, the pediatric correction may be appropriate for children up to age 16, and the adult corrections may be appropriate for people aged 18 and over. The transition between the two is not directly informed by evidence at this time, and future study is warranted.

In the time since the primary data for the ADLs summarized in this paper were collected, the audiometric standard for bone conduction calibration has also been updated (ANSI-ASA, 2025; Margolis et al., 2024). We have suggested ADL adaptations in Table 2. It may be appropriate to use the updated corrections in clinics that have had audiometers calibrated to the 2025 standard. This means that clinicians must know which calibration standard has been used on an audiogram-by-audiogram basis to ensure that the correct ADL is chosen. Clear software labeling of ADL type across calibration type will help to ensure appropriate selection of an ADL type that is aligned with calibration, device type, and patient age. Furthermore, a new audiometric bone oscillator, the B81, was recently released commercially as an improvement over the B71 with better design elements to reduce distortion (Rao et al., 2020). It is possible that this new B81 oscillator may have an ADL transform that differs from the currently implemented transform that is based on audiometric thresholds from a B71 oscillator. However, work by (Rao et al. 2020) found RETFL differences between the B71 and B81 to be non-significant, suggesting that the ADL transform may be generalizable across the two transducers, nor does the ANSI 2025 standard include separate calibration for this new transducer. Continued evaluation of ADL accuracy and generalizability is recommended. Use of in-situ thresholds rather than ADL corrections allows prescription of targets without the need to align audiometric calibration type with ADL correction type, and therefore may be an important strategy in cases where the background calibration is unknown.

5.1 Limitations and future directions

Throughout the computational workflow described in this paper for DSL BCD v2 and for its predecessor v1.1, correction factors and normative data are used. In all cases, these represent the average correction or curve derived from a sample. In all cases, the raw data in each sample varies from the average. This variation may be influenced by the step sizes used in audiometry, by the variability in the loss of signal transmission across skin and connective tissue to bone, and by maturational or individualized response factors. This means that there is error around each correction, which also means that the errors may either sum or cancel depending on the individual. Several sources of uncertainty contribute to this uncertainty, including variability in audiometric threshold measurement, estimation of in-situ thresholds through ADL corrections, coupling-related effects, and RHCD transforms. These sources of uncertainty are expected to accumulate to some extent and contribute residual variability between predicted and actual force levels in individual patients.

Several strategies were employed to permit the development of a prescriptive strategy despite these sources of error. First, a generic Real Head to Coupler Difference (RHCD) was proposed to allow the same computational workflow to apply across a wide range of device types. Second, and as with previous versions of the DSL BCD prescription, DSL BCD v2 incorporates device-specific maximum output into the definition of the output auditory area, so that the targets are mapped to the available combined dynamic range of the listener and the device. This employs a pragmatic strategy to create targets that can be used with clinically-available devices; an “ideal” target would not be limited by poor dynamic range. This limitation, however, has become less impactful as device engineering has improved dynamic range over time, as illustrated by the differences between cases in Figure 4. Clinical protocols recognize the importance of choosing devices with higher maximum output, and because the DSL BCD software can incorporate new MFOs as they evolve, the targets will be adapted in turn to use a more open dynamic range.

5.1.1 Audiometric-to-dial level corrections

The main aim of the addition of ADLs in DSL BCD v2 is the prescription of skin-drive devices, rather than percutaneous abutment fittings. For percutaneous abutment fittings, DSL BCD v1.1 and v2 continue to use directly measured in-situ thresholds as the preferred input. Because the original workflow was developed around in-situ measurements obtained through the patient's own device, an ADL correction is not required when valid in-situ thresholds are available. The ADL corrections used in DSL BCD v2 to compute predicted in-situ BCD thresholds from audiometric BC thresholds were derived in part from two studies in adults, which were used to develop device-independent corrections for both pressure-worn (El-Naji et al., 2026) and adhesive-worn devices (Webster et al., 2026) and one study from children (Bagatto et al., in review). The pediatric ADL corrections were derived from children tested in several real clinical environments, following a clinical consensus protocol, on children who used a variety of device brands and models. The ADLs generated from these data are likely to generalize to clinical practice, because they were collected in clinical practice.

ADL corrections are frequency-specific, and a greater correction is needed for higher frequencies, compared to low and mid frequencies, because BCD output is known to be more limited in these regions relative to audiometric bone oscillator transducers. While the average skin transmission loss will be accounted for in the predicted in-situ BCD thresholds, the ADL corrections in use with DSL BCD v2 are mean values generated from small sample sizes with great variability across individuals. Unfortunately, adequate data to support all BCDs to have a device-specific ADL are not available at this time. Additional studies could develop device-specific strategies, which could enhance the range of supported devices with only minor revision to the software workflow described in this paper.

In light of these factors, in-situ BCD thresholds should be obtained whenever possible in order to individualize the resulting BCD fitting, while ADL corrections should be used whenever in-situ audiometry is not possible, and the prescriptive targets should be used clinically with an awareness of these sources of error. Similar limitations exist for the prescription of air conduction hearing aids, in which individual differences in ear canal transforms can contribute error to the prescriptive target if not measured and incorporated. In BCD prescription, particularly for skin-drive devices, the use of an ADL-corrected audiometric bone conduction threshold leaves open the possibility that significant individual variability is not incorporated. The error associated with assuming an average ADL ranges from 10 to 20 dB, depending on test frequency and population. Because this error is large, we elected to apply ADL corrections at threshold, but not at LDL, in order to avoid inadvertent prescription of higher-than-needed outputs for high-level signals. The use of this ADL strategy was included, despite this variation, in order to support early intervention for infants and young children who require BCDs, by supporting prescription from a wide range of assessment types. Continued evaluation of ADL accuracy, usage, and expansion of normative data across a wider range of device and patient types may improve the application of this strategy over time. Concern for this drives clinical recommendation to measure in-situ, device-specific thresholds for use in prescription, whenever possible (Bagatto et al., 2022), and the use of verification and outcome measurement as best practices to monitor fittings and outcomes at the individual level and over time (Bagatto et al., 2022; Gascon et al., 2025; Hill-Feltham et al., 2021).

5.1.2 Verification and validation of prescriptive targets

Signal transmission transforms for verification in DSL BCD v2 include the microphone location effect (MLE) correction factor and the Real Head to Coupler Difference (RHCD). The MLE was previously developed for high mastoid placement; future work could determine the MLE for forehead placement of non-surgical transcutaneous BCDs, or develop individually-measured estimates of the MLE. Device MLEs are used to convert measured dB FL on the skull simulator into dB FL on-abutment outputs that can be visualized on the FLogram. The current MLE transform was custom-developed for BCDs positioned on the patient's mastoid (Hodgetts and Scollie, 2017). While best practices do recommend BCD positioning over the mastoid process to optimize signal quality and to better encourage localization skill development, there may be some rare instances where a forehead placement of the BCD may be used (Bagatto et al., 2022). For example, early fittings completed for younger infants may find difficulty positioning the BCD over the mastoid process while the child's head is still small and developing. Moreover, at these early stages of development when muscle strength may still be lacking, the infant may often be in a prone position, increasing the tendency for audible feedback from a BCD that would be placed over the mastoid. As such, a forehead placement may be warranted in these early stages and an MLE transform specific to forehead BCD placement should be implemented into the verification to better reflect the accuracy of the dB FL calculations. Future versions of DSL BCD could look to incorporating a custom stored forehead MLE for clinicians to select, depending on their fitting decisions.

Similarly, the current transforms for the RHCD are for device-specific and device-independent averages, and clinical systems for measurement of individual RHCDs do not yet exist. However, the computational pathways within DSL BCD v2 could incorporate such data if measurement of them became available. At the time of this publication, a tool for measuring individual RHCDs, such as with the Surface Microphone first presented by (Hodgetts et al. 2018), with later iterations described by (Persson et al. 2023, 2025), is an emerging technology (Ebrahimian et al., 2026).

At this time, commercial verification of BCDs is available through the Audioscan real bone measurement system (Ebrahimian et al., 2026) and the generic RHCD developed for use with DSL BCD v2.0 is specifically referenced to the Verifit Skull Simulator. The generalizability to other measurement devices is not known. This generic RHCD was derived from manufacturer-provided reference values that likely represent signal transmission on an average adult head. Because skull mass and mechanical impedance change with development and growth, it is possible that age-specific RHCD values and/or individually-measured RHCD values may provide a more appropriate prescription. At present, age-specific RHCD averages are not available. However, it is possible that by incorporating the RHCD into this prescriptive workflow, and through the development of new clinical verification systems such as the real bone measurement system, future measurement of pediatric RHCDs may become possible. This, in turn, could generate new normative data that could offer improvement over and above the transforms presented in this paper.

The development of DSL BCD v2 allows for the objective electroacoustic prescription of BCD output for a broader range of device and coupling types, expanding on the previously-developed targets for direct-drive percutaneous BCDs, but several future directions remain. First, validation of the targets is a required next step to ensure that they provide comfort, good speech intelligibility, and appropriate levels of aided loudness, as well as acceptable subjective outcomes in real-world use. Similar data are available for well-established prescriptions for air conduction hearing aids, but have received fewer investigations for bone conduction prescriptive targets, mainly because they have been developed more recently. Emerging consensus on outcome measurement for BCD fittings may help enable this research (Hill-Feltham et al., 2021). Second, the impact of available prescription and verification equipment may be to improve practice consistency, over the current variability, by increasing adoption of objective device measurements that use high-quality assessment procedures and speech-based verification (Bagatto et al., 2022; Snik et al., 2019). Third, the device category of transcutaneous induction BCDs may also be considered for development of a prescriptive strategy, as several early studies have already considered the role of dynamic range and adaptation of prescriptive targets for use with transcutaneous induction BCDs (Jenkins et al., 2004; Rahne et al., 2022). Overall, the continued development of a prescription for use with BCDs may support consistent, objective fitting, but verification and clinical judgment remain essential.

References

• 1

  AmlaniA. M.PumfordJ.GesslingE. (2017). Real-ear measurement and its impact on aided audibility and patient loyalty. Hearing Rev.24, 12–21.

  • Google Scholar
  • View reference in article

• 2

  AmriN. A.QuarT. K.ChongF. Y.BagattoM. (2022). Optimising hearing aid output to paediatric prescriptive targets improves outcomes in children. Int. J. Audiol.61, 924–931. doi: 10.1080/14992027.2021.2005832

  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  ANSI/ASA (2018). Specification for Audiometers (Standard S3.6-2018). Available online at: https://webstore.ansi.org/standards/asa/ansiasas32018?srsltid=AfmBOoq8i91O_RqqRvQFAfpcXPx-8ZXS0Ri356neFx0VDPGObMNPeknr

  • Google Scholar
  • View reference in article

• 4

  ANSI/ASA (2025). Specification for Audiometers (Standard S3.6-2025). Available online at: https://webstore.ansi.org/standards/asa/asaansis32025

  • Google Scholar
  • View reference in article

• 5

  BagattoM.El-NajiR.BrownC.BehK.ScollieS. (in review). Clinical implementation of DSL v2.0 in the fitting of non-surgical transcutaneous bone conduction devices for children. Front. Audiol. Otology.

  • Google Scholar
  • View reference in article

• 6

  BagattoM.El-NajiR.PurcellD.EaswarV.PigeonM.WitteJ.et al. (2024). Correction factor evaluation and between-system comparison of behavioral threshold predictions from auditory brainstem response measures in infants. Am. J. Audiol.33, 1104–1117. doi: 10.1044/2024_AJA-24-00018

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  BagattoM.GordeyD.BrewsterL.BrownC.ComeauM.DouglasC.et al. (2022). Clinical consensus document for fitting non-surgical transcutaneous bone conduction hearing devices to children. Int. J. Audiol.61, 531–538. doi: 10.1080/14992027.2021.1939449

  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  BagattoM.ScollieS.MoodieS. T. F.SeewaldR.HydeM.El-NajiR.et al. (2023). Protocol for the Provision of Amplification v2023.01. London, ON: Ontario Infant Hearing Program.

  • Google Scholar
  • View reference in article

• 9

  CarlssonP. U.HåkanssonB. (1997). The bone-anchored hearing aid: reference quantities and functional gain. Ear Hear.18, 34–41. doi: 10.1097/00003446-199702000-00004

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  ChingT.GalsterJ.GrimesA.JohnsonC.LewisD.McCreeryR.et al. (2013). Clinical Practice Guidelines: Pediatric Amplification. Reston, VA: American Academy of Audiology.

  • Google Scholar
  • View reference in article

• 11

  CrukleyJ.ScollieS. D. (2012). Children's speech recognition and loudness perception with the desired sensation level v5 quiet and noise prescriptions. Am. J. Audiol.21, 149–162. doi: 10.1044/1059-0889(2012/12-0002)

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  EbrahimianE.PietrobonJ.AlizotiN.LucasM.McInerneyC.PumfordJ. (2026). Comparing real-bone measurement microphone and skull simulator measurements for soft headband bone conduction devices in adults. Front. Audiol. Otol. 4:1768441. doi: 10.3389/fauot.2026.1768441

  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  El-NajiR.BagattoM.ScollieS. (2026). Equivalent threshold force levels for skin drive bone conduction hearing devices. Front. Audiol. Otol.3:1714106. doi: 10.3389/fauot.2025.1714106

  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  El-NajiR.ScollieS.BagattoM. (2024). Examining force level output of skin-drive bone conduction hearing devices in adults with simulated conductive hearing loss. Am. J. Audiol.33, 695–704. doi: 10.1044/2024_AJA-23-00258

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  FlynnM. C.HillbrattM. (2012). Improving the accuracy of baha^® fittings through measures of direct bone conduction. Clin. Exp. Otorhinolaryngol.5, S43–S47. doi: 10.3342/ceo.2012.5.S1.S43

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  GalsterJ.HoganC.JorgensenL.McCreeryR.PhelanJ.PicouE.et al. (2022). Hearing Aid Fitting Standard for Pediatric Patients (Version S2.2). Audiology Practice Standards Organization (APSO).

  • Google Scholar
  • View reference in article

• 17

  GalsterJ.HoganC.JorgensenL.MeullerG.McCreeryR.PicouE.et al. (2021). Hearing Aid Fitting Standard for Adult Geriatric Patients (Version S2.1). Audiology Practice Standards Organization (APSO).

  • Google Scholar
  • View reference in article

• 18

  GasconA.BagattoM.ScollieS.CowanC.HodgettsW. (2025). Clinical evaluation of DSL-BCD fittings: assessing output-to-targets for percutaneous bone conduction devices in practice. Front. Audiol. Otol. 3:1714575. doi: 10.3389/fauot.2025.1714575

  • CrossRef
  • Google Scholar
  • View reference in article

• 19

  GasconA.OstevikA. V.HuynhT.HodgettsW. E. (2022). Which threshold do we trust? A comparison of threshold measurements in adult bone-conduction device users and normal hearing adults. Hearing Res. 421:108491. doi: 10.1016/j.heares.2022.108491

  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  GlistaD.O'HaganR.Van EeckhoutteM.LaiY.ScollieS. (2021). The use of ecological momentary assessment to evaluate real-world aided outcomes with children. Int. J. Audiol.60, S68–S78. doi: 10.1080/14992027.2021.1881629

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  GordeyD.BagattoM. (2021). Fitting bone conduction hearing devices to children: audiological practices and challenges. Int. J. Audiol.60, 385–392. doi: 10.1080/14992027.2020.1814970

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  GriffinA. M.AtriA.LicameliG.StilesD. J. (2023). Effect of hearing device use on speech-in-noise performance in children with severe-to-profound unilateral hearing loss. Ear Hear.44:588. doi: 10.1097/AUD.0000000000001310

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  HåkanssonB.ReinfeldtS.PerssonA-. C.JanssonK-. J. F.RigatoC.HultcrantzM.et al. (2019). The bone conduction implant - a review and 1-year follow-up. Int. J. Audiol.58, 945–955. doi: 10.1080/14992027.2019.1657243

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  HåkanssonB.TjellströmA.RosenhallU. (1985). Acceleration levels at hearing threshold with direct bone conduction versus conventional bone conduction. Acta Oto-Laryngologica100, 240–252. doi: 10.3109/00016488509104786

  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  HawkinsD. (2004). Limitations and uses of the aided audiogram. Seminars Hear.25, 51–62. doi: 10.1055/s-2004-823047

  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  Hill-FelthamP. R.JohanssonM. L.HodgettsW. E.OstevikA. V.McKinnonB. J.MonksfieldP.et al. (2021). Hearing outcome measures for conductive and mixed hearing loss treatment in adults: a scoping review. Int. J. Audiol.60, 239–245. doi: 10.1080/14992027.2020.1820087

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  HodgettsW.ScottD.MaasP.WestoverL. (2018). Development of a novel bone conduction verification tool using a surface microphone: validation with percutaneous bone conduction users. Ear Hear.39, 1157–1164. doi: 10.1097/AUD.0000000000000572

  • CrossRef
  • Google Scholar
  • View reference in article

• 28

  HodgettsW. E.HaglerP.HakanssonB. E. V.SoliS. D. (2011). Technology-limited and patient-derived versus audibility-derived fittings in bone-anchored hearing aid users: a validation study. Ear Hear.32, 31–39. doi: 10.1097/AUD.0b013e3181f0b685

  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  HodgettsW. E.ScollieS. D. (2017). DSL prescriptive targets for bone conduction devices: adaptation and comparison to clinical fittings. Int. J. Audiol.56, 521–530. doi: 10.1080/14992027.2017.1302605

  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  HolM. K. S.CremersC. W. R. J.Coppens-SchellekensW.SnikA. F. M. (2005). The BAHA softband. Int. J. Pediatr. Otorhinolaryngol.69, 973–980. doi: 10.1016/j.ijporl.2005.02.010

  • CrossRef
  • Google Scholar
  • View reference in article

• 31

  JenkinsH. A.NiparkoJ. K.SlatteryW. H.Gail NeelyJ.FredricksonJ. M. (2004). Otologics middle ear transducerTM ossicular stimulator: performance results with varying degrees of sensorineural hearing loss. Acta Oto-Laryngologica124, 391–394. doi: 10.1080/00016480410016298

  • CrossRef
  • Google Scholar
  • View reference in article

• 32

  KurzA.FlynnM.GoodT.CaversaccioM.KompiM. (2014). Influence of loudness compression on hearing with bone anchored hearing implants. Int. J. Med. Pharm. Sci. Eng.8, 1–6.doi: 10.5281/zenodo.1336164

  • CrossRef
  • Google Scholar
  • View reference in article

• 33

  MargolisR. H.SanchezV.HunterL. L.RaoA.BoyleS.ZadehL. M.et al. (2024). Air-conduction and bone-conduction reference threshold levels-a multicenter study. J. Acoust. Soc. Am.156, 3545–3553. doi: 10.1121/10.0034456

  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  MarriageJ. E.VickersD. A.BaerT.GlasbergB. R.MooreB. C. J. (2018). Comparison of different hearing aid prescriptions for children. Ear Hear.39, 20–31. doi: 10.1097/AUD.0000000000000460

  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  MattinglyJ. K.GreeneN. T.JenkinsH. A.TollinD. J.EasterJ. R.CassS. P.et al. (2015). Effects of Skin Thickness on Cochlear Input Signal Using Transcutaneous Bone Conduction Implants. Otol. Neurotol. 36, 1403–1411. doi: 10.1097/MAO.0000000000000814

  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  McCreeryR. W.BentlerR. A.RoushP. A. (2013). Characteristics of hearing aid fittings in infants and young children. Ear Hear.34, 701–710. doi: 10.1097/AUD.0b013e31828f1033

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  McCreeryR. W.BrennanM.WalkerE. A.SpratfordM. (2017). Perceptual implications of level- and frequency-specific deviations from hearing aid prescription in children. J. Am. Acad. Audiol.28, 861–875. doi: 10.3766/jaaa.17014

  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  McCreeryR. W.WalkerE. A.SpratfordM.BentlerR.HolteL.RoushP.et al. (2015). Longitudinal predictors of aided speech audibility in infants and children. Ear Hear. 36, 24S–37S. doi: 10.1097/AUD.0000000000000211

  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  MylanusE. A.van der PouwK. C.SnikA. F.CremersC. W. (1998). Intraindividual comparison of the bone-anchored hearing aid and air-conduction hearing aids. Arch. Otolaryngol. Head Neck Surg.124, 271–276. doi: 10.1001/archotol.124.3.271

  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  NieY.SangJ.ZhengC.XuJ.ZhangF.LiX.et al. (2022). An objective bone conduction verification tool using a piezoelectric thin-film force transducer. Front. Neurosci.16:1068682. doi: 10.3389/fnins.2022.1068682

  • CrossRef
  • Google Scholar
  • View reference in article

• 41

  NorrixL. W. (2015). Hearing thresholds, minimum response levels, and cross-check measures in pediatric audiology. Am. J. Audiol.24, 137–144. doi: 10.1044/2015_AJA-14-0095

  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  PerssonA.-C.HåkanssonB.Fredén JanssonK.-J.ReinfeldtS.Eeg-OlofssonM. (2025). Objective verification of audibility in bone conduction devices. Int. J. Audiol.64, 217–223. doi: 10.1080/14992027.2024.2335511

  • CrossRef
  • Google Scholar
  • View reference in article

• 43

  PerssonA-. C.HåkanssonB.Caveramma MechandaM.HodgettsW.Fredén JanssonK.-J.Eeg-OlofssonM.et al. (2023). A novel method for objective in-situ measurement of audibility in bone conduction hearing devices - a pilot study using a skin drive BCD. Int. J. Audiol.62, 357–361. doi: 10.1080/14992027.2022.2041739

  • CrossRef
  • Google Scholar
  • View reference in article

• 44

  RahneT.FröhlichL.WagnerL.KroppM. H.MüllerA. (2022). Speech perception and hearing effort using a new active middle ear implant audio processor. Eur. Arch. Oto-Rhino-Laryngol.279, 4667–4675. doi: 10.1007/s00405-021-07207-4

  • CrossRef
  • Google Scholar
  • View reference in article

• 45

  RaoA.AltusB.SchroederR.MargolisR. H. (2020). Comparison of thresholds obtained with the Radioear B-71 and B-81 bone-conduction transducers. Ear Hear.41, 1775–1778. doi: 10.1097/AUD.0000000000000873

  • CrossRef
  • Google Scholar
  • View reference in article

• 46

  RicciG.VolpeA. D.FaralliM.LongariF.LancioneC.VarricchioA. M.et al. (2011). Bone-anchored hearing aids (Baha) in congenital aural atresia: personal experience. Int. J. Pediatr. Otorhinolaryngol.75, 342–346. doi: 10.1016/j.ijporl.2010.11.027

  • CrossRef
  • Google Scholar
  • View reference in article

• 47

  RoyceW. W. (2021). “Managing the development of large software systems (1970),” in Ideas That Created the Future, ed. LewisH. R. (Cambridge, MA: The MIT Press), 321–332. doi: 10.7551/mitpress/12274.003.0035

  • CrossRef
  • Google Scholar
  • View reference in article

• 48

  SaboD. L.ParadiseJ. L.Kurs-LaskyM.SmithC. G. (2003). Hearing levels in infants and young children in relation to testing technique, age group, and the presence or absence of middle-ear effusion. Ear Hear.24, 38–47. doi: 10.1097/01.AUD.0000051988.23117.91

  • CrossRef
  • Google Scholar
  • View reference in article

• 49

  ScollieS.ChingT. Y. C.SeewaldR.DillonH.BrittonL.SteinbergJ.et al. (2010a). Evaluation of the NAL-NL1 and DSL v4.1 prescriptions for children: preference in real world use. Int. J. Audiol.49, S49–S63. doi: 10.3109/14992020903148038

  • CrossRef
  • Google Scholar
  • View reference in article

• 50

  ScollieS.SeewaldR.CornelisseL.MoodieS.BagattoM.LaurnagarayD.et al. (2005). The desired sensation level multistage input/output algorithm. Trends Amplification9, 159–197. doi: 10.1177/108471380500900403

  • CrossRef
  • Google Scholar
  • View reference in article

• 51

  ScollieS. D. (2008). Children's speech recognition scores: the speech intelligibility index and proficiency factors for age and hearing level. Ear Hear.29, 543–556. doi: 10.1097/AUD.0b013e3181734a02

  • CrossRef
  • Google Scholar
  • View reference in article

• 52

  ScollieS. D.ChingT. Y. C.SeewaldR. C.DillonH.BrittonL.SteinbergJ.et al. (2010b). Children's speech perception and loudness ratings when fitted with hearing aids using the DSL v.4.1 and the NAL-NL1 prescriptions. Int. J. Audiol.49, S26–S34. doi: 10.3109/14992020903121159

  • CrossRef
  • Google Scholar
  • View reference in article

• 53

  SeewaldR.MoodieS.ScollieS.BagattoM. (2005). The DSL method for pediatric hearing instrument fitting: historical perspective and current issues. Trends Amplification9, 145–157. doi: 10.1177/108471380500900402

  • CrossRef
  • Google Scholar
  • View reference in article

• 54

  SkarżyńskiP. H.RatuszniakA.KrólB.KoziełM.OsińsskaK.CywkaK. B.et al. (2019). The Bonebridge in adults with mixed and conductive hearing loss: audiological and quality of life outcomes. Audiol. Neurotol.24, 90–99. doi: 10.1159/000499363

  • CrossRef
  • Google Scholar
  • View reference in article

• 55

  SnikA.MaierH.HodgettsB.KompisM.MertensG.Van De HeyningP.et al. (2019). Efficacy of auditory implants for patients with conductive and mixed hearing loss depends on implant center. Otol. Neurotol.40, 430–435. doi: 10.1097/MAO.0000000000002183

  • CrossRef
  • Google Scholar
  • View reference in article

• 56

  SnikF. M.MylanusE. A. M.ProopsD. W.WolfaardtJ. F.HodgettsW. E.SomersT.et al. (2005). Consensus statements on the BAHA system: where do we stand at present?Ann. Otol. Rhinol. Laryngol.114, 2–12. doi: 10.1177/0003489405114S1201

  • CrossRef
  • Google Scholar
  • View reference in article

• 57

  StelmachowiczP. G.HooverB.LewisD. E.BrennanM. (2002). Is functional gain really functional?Hear. J.55, 38–42. doi: 10.1097/01.HJ.0000324170.43736.78

  • CrossRef
  • Google Scholar
  • View reference in article

• 58

  StenfeltS.HåkanssonB. (2002). Air versus bone conduction: an equal loudness investigation. Hear. Res. 167, 1–12. doi: 10.1016/S0378-5955(01)00407-5

  • CrossRef
  • Google Scholar
  • View reference in article

• 59

  StenfeltS.ZeitooniM. (2013). Loudness functions with air and bone conduction stimulation in normal-hearing subjects using a categorical loudness scaling procedure. Hear. Res.301, 85–92. doi: 10.1016/j.heares.2013.03.010

  • CrossRef
  • Google Scholar
  • View reference in article

• 60

  SunL.PingL.FanX.WangJ.ChenX. (2024). Simulator verification is potentially beneficial for the fitting of softband bone conduction hearing devices in young children. Otol. Neurotol.45:e500. doi: 10.1097/MAO.0000000000004245

  • CrossRef
  • Google Scholar
  • View reference in article

• 61

  ValenteM.AbramsH.BensonD.ChisolmT.CitronD.HamptonD.et al. (2006). Guidelines for the Audiologic Management of Adult Hearing Impairment. Reston, VA: merican Academy of Audiology.

  • Google Scholar
  • View reference in article

• 62

  ValenteM.OedingK.BrockmeyerA.SmithS.KallogjeriD. (2018). Differences in word and phoneme recognition in quiet, sentence recognition in noise, and subjective outcomes between manufacturer first-fit and hearing aids programmed to NAL-NL2 using real-ear measures. J. Am. Acad. Audiol.29, 706–721. doi: 10.3766/jaaa.17005

  • CrossRef
  • Google Scholar
  • View reference in article

• 63

  WangY.FanX.WangP.FanY.ChenX. (2018). Hearing improvement with softband and implanted bone-anchored hearing devices and modified implantation surgery in patients with bilateral microtia-atresia. Int. J. Pediatr. Otorhinolaryngol.104, 120–125. doi: 10.1016/j.ijporl.2017.11.010

  • CrossRef
  • Google Scholar
  • View reference in article

• 64

  WebsterE.LochnerJ.FolkeardP.ScollieS.MojallalH.BagattoM.et al. (2026). Reference threshold measures for a non-surgical skin-drive bone conduction hearing device. Front. Audiol. Otol.4:1850851. doi: 10.3389/fauot.2026.1850851

  • CrossRef
  • Google Scholar
  • View reference in article

• 65

  WiesnerT.BohnertA.LimbergerA.MassingerC.NickischA. (2012). Konsenspapier der GPP zurHörgeräte-Versorgung bei Kindern, Vers. 3.5 (Germany). Bonn: KIND.

  • Google Scholar
  • View reference in article

• 66

  WillenborgK.LenarzT.BuschS. (2023). Surgical and audiological outcomes with a new transcutaneous bone conduction device with reduced transducer thickness in children. Eur. Arch. Oto-Rhino-Laryngol.280, 4381–4389. doi: 10.1007/s00405-023-07927-9

  • CrossRef
  • Google Scholar
  • View reference in article