Upload de-identified DICOM data to a dedicated instance of Nota, our web-based platform
Automatically quality-control and index meta- and pixel data
Query data and build cohorts in a secure, access-controlled environment
Automatically generate accurate, reproducible predictions for each baseline and follow-up scan
View predictions at the cohort, subject, and scan level
Download quantifications to incorporate into statistical analyses
Detect treatment effect with greater confidence across subpopulations
Inform trial design and go / no-go decisions to increase probability of late-stage success
Generate higher quality evidence from existing data to inform reimbursement strategy
▶
Leverage existing imaging data to its fullest analytical potential
▶
Instantaneously generate predictions for 1,000+ subjects
▶
Enrich analyses across past and current Ph 1-3 trials with comprehensive and reproducible quantifications
▶
Instead of relying on arbitrary thresholds, we aim to redefine what it means for clinical trial subjects to "respond" and "progress" by evaluating IPRO against meaningful outcomes
▶
Outcomes are what ultimately matter to patients, physicians, payors, and regulators
▶
Imaging data is 105 times richer than EMR data; a longitudinal 3D reconstruction of the entire human anatomy
▶
Training IPRO on this data allows predictions to be hyper precise and generalizable across diverse populations
▶
IPRO models can be applied across therapeutic areas, as imaging is used in screening, staging, and monitoring in most diseases.
▶
IPRO's "learned" features reveal their protective or detrimental nature with respect to outcomes, offering novel biological insights
▶
Explainable predictions help our field move beyond reductionist measures that fail to capture the full complexity of disease and toxicity
IPRO for
IPRO for