The human voice contains complex acoustic markers which have been linked to important health conditions including dementia, mood disorders, and cancer. When viewed as a biomarker, voice is a promising characteristic to measure as it is simple to collect, cost-effective, and has broad clinical utility. Recent advances in artificial intelligence have provided techniques to extract previously unknown prognostically useful information from dense data elements such as images. The Bridge2AI-Voice project seeks to create an ethically sourced flagship dataset to enable future research in artificial intelligence and support critical insights into the use of voice as a biomarker of health. Here we present Bridge2AI-Voice, a comprehensive collection of data derived from voice recordings with corresponding clinical information. Bridge2AI-Voice v2.0 contains data for 19,271 recordings collected from 442 participants across five sites in North America. Participants were selected based on known conditions which manifest within the voice waveform including voice disorders, neurological disorders, mood disorders, and respiratory disorders. The release contains data considered low risk, including derivations such as spectrograms but not the original voice recordings. Detailed demographic, clinical, and validated questionnaire data are also made available. Audio recordings are included on a companion release on PhysioNet with the title "Bridge2AI-Voice: An ethically-sourced, diverse voice dataset linked to health information (Audio Included)". Please see that project for details to request access.

In our recent study, we used Llama-3.1-70B-Instruct to generate synthetic training examples resembling clinical trial eligibility criteria. We manually reviewed 1000 of these examples and release them here.

The 2024 Voice AI Symposium presented by the Bridge2AI-Voice Consortium, was a 2-day event which took place May 1st-May 2nd in Tampa, FL. The event included four interactive panel sessions, which are summarized here. All four interactive panels featured an innovative format, designed to maximize engagement and facilitate deep discussions. Each panel began with a 45 min segment where moderators posed targeted questions to expert panelists, delving into complex topics within the field of voice AI. This was followed by a 45 min “stakeholder forum,” during which audience members asked questions and engaged in live interactive polls. Interactive polls stimulated meaningful conversation between panelists and attendees, and brought to light diverse viewpoints. Workshops were audio recorded and transcripts were assembled with assistance from generative A.I tools including Whisper Version 7.13.1 for audio transcription and ChatGPT version 4.0 for content summation. Content was then reviewed and edited by authors.