Mantis Biotech is turning fragmented health data into usable digital twins
Mantis Biotech combines disparate sources of clinical, imaging, sensor and behavioral data to generate synthetic datasets that feed into so-called "digital twins" — detailed virtual representations of human anatomy, physiology and behavior. By synthesizing data rather than relying solely on scarce, siloed patient records, Mantis aims to give researchers and developers a richer, privacy-preserving foundation for building medical models and running simulations.
The promise is practical and immediate: synthetic datasets can be used to train machine learning models, run virtual cohorts for early-stage testing, and explore how interventions might play out across diverse patient types without exposing real patient records. Because the twins capture multiple layers — from anatomy to day-to-day behavior — they enable more holistic simulations than traditional single-modality datasets.
Key benefits include:
- Improved data availability: Synthetic data fills gaps where clinical data are limited or fragmented.
- Privacy preservation: Generated datasets reduce the need to share identifiable patient records.
- Accelerated R&D: Digital twins allow safer, faster iteration of models and virtual experiments before moving to costly real-world trials.
While synthetic twins are not a substitute for real-world validation, they can meaningfully reduce barriers to innovation by democratizing access to realistic medical data and enabling earlier, lower-risk testing. If adopted widely, this approach could speed development of diagnostics, therapeutics and personalized care pathways, helping researchers and clinicians bring better treatments to patients more quickly.