Comparing outcomes with the cloud-based and portable versions of an open-source hearing aid

Abstract

The open-source speech processing platform (OSP) is an open-source hearing aid (HA), developed to overcome proprietary restrictions of commercial HAs and pave the way for innovative HA technology through rigorous and reproducible research. This study evaluates portable and cloud-based platforms of the OSP that will enable a wide range of lab and field applications. The two platforms share common software applications for HA signal processing and objective outcome assessments (HO2). The cloud-based platform is hosted on a web server whereas the portable platform is accessed on a browser via a WiFi-enabled portable communication device. Listeners with mild to moderately severe hearing loss or normal hearing completed a word identification task (Modified Rhyme Test) in a 6-alternate forced choice paradigm using the HO2 app. Words were mixed with different levels of speech-shaped noise; presented via calibrated headphones or behind-the-ear receivers-in-the-canal for cloud-based and portable platforms, respectively. Signals were processed with individualized gain and compression ratios based on a clinically validated prescriptive formula. Results-to-date from four listeners show comparable range of word identification scores between platforms. The full dataset will also compare phoneme-level scores and response time measures generated from the HO2 app. Clinical research applications will be discussed. [Work supported by NIH-R44DC020406.]