Foundations and Collaborations
UKRI Frontier AI and AIRR-ISAMBARD grants in support of our measurement platform, plus a structured internship programme with the University of Cambridge.
Read articleMethods from TimeTrace Labs
Lacuna: Missingness as Signal
Understanding and handling informative missingness in clinical data — where the absence itself carries systematic information about the outcome.
Read article
Prism: Spectral Self-Supervised Learning for Physiological Time Series
A time-aware regulariser for physiological time series, developed with NVIDIA, that preserves temporal structure where pooled isotropy objectives can lose it.
Read article
Sonata: A World Model of Human Motion
The first public TimeTrace Labs work, with NVIDIA — latent-state prediction for motion under real clinical constraints.
Read article
Forma: Subgroup Identification in Population Psychiatry
With Fondation FondaMental — latent class growth modelling on a five-year psychosis cohort resolves four stable severity subgroups robust to structured missingness.
Read article
Meridian: Measuring Biology As It Changes
A mathematical framework for measuring biological change from any physiological signal — subgroup identification, earlier endpoints, and smaller trials across colorectal cancer, ALS, and Parkinson's disease.
Read article
The Measurement Stack
Building the measurement infrastructure of healthcare — from stochastic processes and statistical inference through to world models on multimodal physiological data.
Read article
Ground Truth
Clinical scores and diagnostic categories are inherited as ground truth — but continuous biology is compressed into conventions that models learn to reproduce with ever greater precision.
Read article
Additional articles will appear here as we publish.
No posts match these filters.