Dynamic current focusing for loudness encoding in cochlear implants: a take-home trial

Abstract

Objective: This study aimed to evaluate a more energy-efficient dynamic current focussing (DCF) speechprocessing strategy after long-term listening experience. In DCF, tripolar stimulation is used near the threshold and loudness is controlled by the compensation coefficient r. A recent acute pilot study showed improved spectral-temporally modulated ripple test (SMRT) scores at low loudness levels, but battery life was reduced to 1.5-4 hours. Design: Within-subject comparisons were made for the clinical versus. DCF strategy after 5 weeks of athome usage. Speech intelligibility in noise, spectral ripple discrimination, temporal modulation detection, loudness growth, and subjective ratings were assessed. Study sample: Twenty HiRes90K (Advanced Bionics, Valencia, USA) cochlear implant (CI) users. Results: Average battery life was 9 hours with the newly implemented DCF compared to 13.4 hours with the clinical strategy. Compared with measurements made at the beginning of the study, SMRT-scores and speech intelligibility in noise were significantly improved with DCF. However, both measures suffered from unexpected learning effects over time. The improvement disappeared and speech intelligibility in noise declined significantly relative to the final control measurement with the clinical strategy. Conclusion: Most CI users can adapt to the DCF strategy in a take-home setting. Although DCF has the potential to improve performance on the SMRT test, learning effects complicate the interpretation of the current results.

Investigating the Electrical Properties of Different Cochlear Implants

“…There is evidence for better speech discrimination with partial tripolar stimulation compared with monopolar stimulation (11), but another study finds no difference in pitch discrimination between tripolar and monopolar stimulation (12). More recently, a modified form of tripolar stimulation termed dynamic current focusing is purported to give better spectral resolution under certain conditions (13). Another proposed strategy is phased array stimulation, which aims to produce focused electrical stimulation at a particular site by sending a complex phase-shifted array of inverted currents as the stimulation signal to all other sites that tries to cancel out the effects of current spread (14).…”

Current Focusing to Reduce Channel Interaction for Distant Electrodes in Cochlear Implant Programs

“…Their results showed an improvement in noise, but no improvement under quiet conditions. Another similar study found that when using dynamic focusing spectral resolution improved, but this did not extend to speech performance ( De Jong, Briaire, & Frijns, 2018 ). These studies show that while results are mixed, channel focusing has the potential to improve speech perception, particularly in noise.…”

Cochlear implant research and development in the 21st century: A critical update

“…However, in that study the order of testing was not fully counterbalanced between the strategies, such that ACE was never tested first, and so one cannot rule out the influence of practice effects (cf. de Jong et al, 2019). Psarros et al (2002) reported significant improvements on word identification in quiet when seven children were converted from the SPEAK to the ACE strategy, and, importantly, that performance deteriorated again when they were switched back to SPEAK.…”

“…Unfortunately, this design is susceptible to practice effects, such that the improvement might for example be due to increased familiarity with the test materials (Psarros et al, 2002;Vermeire et al, 2010;de Jong et al, 2019). Indeed, the differences between test sessions observed with a smaller number of tested ears, whose implants were not reprogrammed, overlapped substantially with those having deactivated electrodes.…”

“…A recent innovation is to dynamically vary  as a function of input level so that the mode Carlyon & Goehring -21 st century CI 20 of stimulation is more similar to full tripolar at low levels and more similar to monopolar at high levels (Arenberg et al, 2018;de Jong et al, 2019). An unblinded study by Arenberg et al (2018) found that this method improved the perception of vowels in fourtalker babble (but not the identification of vowels in quiet or spondees in babble) compared both to MP and pTP strategies.…”