The SB-M objectively detects declining speech intelligibility, while the SB-C tracks parallel shifts in cognitive performance. Together, a single brief speech protocol captures both motor and cognitive status to support trial inclusion, enrichment, and progression prediction.
0.86 AUC
<10
32
The ALS
Measurement Gap
8–12 weeks
50%
of people with ALS exhibit neuropsychological changes
30%
of people with ALS present with bulbar onset — the hardest subgroup to track
36%
Rate of attrition over a 12-month ALS study
The SB-M captures bulbar subtypes through a short, standardized speech task. AI-driven analysis extracts features to return an objective Intelligibility Score, which is important and meaningful to patients with ALS, as well as articulation and respiration subscores. The SB-C can be easily added to track and monitor cognition.
Assessment Protocol
01
Participant opens Mili
Any smartphone,
any location
Standard smartphone app or webpage. No special microphone, no wearable, no site visit required.
02
Standardized speech tasks administered
~10 minutes
Picture description, text reading, sustained phonation, and diadochokinesis are performed.
03
AI pipeline extracts intelligibility and acoustic features
Automated
The SIGMA pipeline automatically generates the SB-M Intelligibility Score, alongside articulation and respiration subscores, and single features, such as speech rate. No manual transcription required.
04
Score delivered via API or Mili dashboard
Same day
Structured output integrates into your EDC, CTMS, or study portal. Full audit trail. Longitudinal trajectory visualization available for site monitoring teams.
SB-M validation follows the Digital Medicine Society’s V3 Framework, covering verification, analytical validation and clinical validation.A full validation package is available for regulatory submissions.
Validated in German, English, Czech, Italian, Swedish, French, Spanish, and Colombian Spanish ALS cohorts. Supports multilingual, multinational clinical trial deployment.



