The Need for
MultiModal Biomarkers
Biomarkers across a range of modalities have been shown to correlate with virtually every CNS indication. But unlike other therapeutic areas, such as oncology, individual biomarkers often fail to achieve meaningful clinical relevance in CNS; holding back the promise of precision medicine.
New research, supported by recent advances in multimodal machine learning, is revealing that fusing multiple biomarkers - what we call Biomarker Constellations™ - can add unprecedented value for new drug and medical device development, and ultimately in clinical care.
The Need for
MultiModal Biomarkers
This Sankey diagram visualizes our database of biomarker-CNS indication correlations that Deliberate has curated and updates regularly.
By hovering over or clicking on the various biomarker modalities and/or CNS indications, you will see the scientific evidence we have reviewed. While this is exploratory and illustrative, it demonstrates the vast opportunity.
​
Please sign up for access and regular updates to this database.
Introducing
A powerful platform for CNS Biomarker Discovery and Development.
Deliberate offers the life sciences industry the first industrial-grade platform that extracts hundreds of behavioral and physiological features from ordinary video recordings. Multitude™ also facilitates feature selection, helping researchers to identify the features that are most closely correlated with disease processes of interest. Further, Multitude™ facilitates the fusion of different biomarker modalities, including those from outside our platform, and includes a variety of tools that will help data science teams to build high-performance models to discover, enrich and validate different kinds of Biomarker Constellations™. With Multitude™, the journey between raw data and useful models has gotten a lot shorter and smoother.
NATIVE INPUT MODALITY
AUDIO/VIDEO
RETINA SCAN
BLOODBASED PANOMICS
GENOMICS INFO
CONTINUOUS PASSIVE SENSING
NEURO IMAGING
FUSING OPTIONAL
EXTERNAL MODALITIES
FACIAL EXPRESSION
BODY MOVEMENT
FEATURE EXTRACTION MODULES
HEAD MOVEMENT
OCULOMETRY & PUPILLOMETRY
HEART RATE
VOCAL ACOUSTICS
SPEECH BEHAVIORS
BREATHING
COMPUTATIONAL LINGUISTICS
DISCOVERY MODULES
BIOMARKER AND PHENOTYPE EVALUATION
FEATURE SELECTION
CLASSICAL ML AND DEEP LEARNING
MODEL TRAINING AND EVALUATION
BIOMARKER FUSION