Automated, scalable speech-based measures relevant to symptom fluctuation, including anhedonia and psychomotor retardation. Precision psychiatry delivered through a short spoken task.
At a glance
100+
features extracted per recording
~2 min
task protocol duration
The MDD
Measurement Problem
280M
People affected by depression worldwide
High
Placebo response rates in clinical trials
Zero
Universally accepted objective biomarker for treatment response
Subjective
Inter-rater variability using traditional scales
Beyond recall bias
Traditional depression rating scales (like the HAM-D or MADRS) heavily rely on patient recall and clinician interpretation, clouded by recall bias or subjective scoring. Analyzing acoustic properties, like speech rate, pause duration, and pitch variability, during a simple task provides objective measures without heavy burden.
Quantifies what’s hardest to measure
Psychomotor retardation and blunted affect are core diagnostic criteria for MDD, yet incredibly difficult to quantify precisely. ki:elements extracts over 100 acoustic, prosodic, and linguistic features from a brief speech sample — paving the way for automated tracking of subtle treatment responses.