At its simplest level, the delivery of healthcare can be broken down into two modalities — “diagnostics” and “therapeutics.” Diagnostics is about analyzing information from life — this can be digitized. Therapeutics is about affecting change upon life — this will fundamentally require a physical or molecular component to it. However, there is plenty of room on the diagnostics side of the health equation for digitized and virtualized delivery of care.
Many of these diseases develop while expressing themselves through non-molecular pathways, depending on the disease, typically as subtle, occasional, or intermittent physiologic or behavioral "signs." As humans, we subsequently experience these diseases through symptoms, often later in a disease's stage, and sometimes too late.
No. The signs of many diseases do not need to be limited to molecular measures, but can and should be measured non-molecularly (phenotypically) as well. LIFEdata can be generated without needing to physically or invasively "access" an intermediary such as blood or tissue – yielding unprecedented scalability – a key hallmark of the digital revolution.
LIFEdata will provide more objective, convenient, affordable, and frequent insights into our dynamic health.
We define LIFEdata as sensor-measured information digitally generated from life and around life. We are converting LIFEdata (learnable insights from expressions) into digital biomarkers for disease detection and drug response through machine learning and artificial intelligence.
Some examples of LIFEdata include heart rate (PPG), bodily motions (accelerometer), behavioral inputs (touchscreen), vocal sounds (microphone), visible features in the face and eyes (camera), and many more.
The time dimension has been largely absent in our approaches to healthcare today. Intermittent (i.e. "snapshot") measures of our health in Medical setting visits are ineffective.
Disease “velocity” – a function of time – is just as important as disease “position” or diagnosis. Many developing diseases have a dynamic pattern of progression that can be better detected through more frequent physiologic measures – enabled via digital biomarkers.
Earlier detection catalyzes timelier diagnoses and interventions, which can yield improved clinical outcomes and treatment cost savings – or ROeI (Return On earlier Intervention) – for the net system as a whole.
Consumer smart devices with sensors that can capture LIFEdata are starting to rapidly populate our Life settings. These digital devices provide individuals with the opportunity to contribute to and access novel healthcare innovations outside the Medical setting. Even IoT home devices not originally purposed for healthcare applications can be a very valuable source of LIFEdata which will be processed by our technology platform. We believe subtle and quantifiable, non-molecular and "digitizable" signals are under-discovered, but very insightful objective "markers" for many developing diseases and disorders.
Take a smartwatch that can now continuously measure heart rate – a type of LIFEdata. AI can process and analyze millions of retrospectively measured heart rate data points across a large population to determine specific changes in heart rate patterns that mark most accurately to a cardiovascular abnormality – making it a digital biomarker. Users of heart rate measuring smartwatches (programmed with this digital biomarker algorithm) can then be notified in real time if they are developing a heart condition such as arrhythmia, and seek clinical intervention in a medical setting much earlier than they may have done otherwise.
BioTrillion's first set of digital biomarkers are focused on Neurology and providing insights into the human brain.
Smartphones, smartwatches and smart devices provide the Consumer a number of sensory modalities to more conveniently, frequently, and affordably generate LIFEdata. Existing chips, hardware, and sensors already in your smart devices have already put powerful tools in your pocket - novel and digitally downloadable software can enable their full potential as health tools. And these smart devices become more powerful each year, effectively broadening their health application potential.
Human vision evolved to detect variations at the Macro scale – yielding innovations of analogous automation. Computer vision now allows us to quantify visual features hidden at the Milli scale – yielding innovations of novel enablement.
BioTrillion is developing radically novel digital biomarkers by algorithmically training BioEngine4D via machine learning and artificial intelligence — premised around clinically-known and causal pathophysiology.
Our digital biomarkers have the potential to be more effective and scalable than molecular, physical biomarkers of disease and present compelling value propositions across the healthcare stakeholder paradigm, for both served and under-served diseases and population.
Some of these value propositions include improving time and cost inefficiencies in the drug development process (pharma), catalyzing earlier clinical diagnoses and interventions, which can lead to savings in treatment costs (payors), and more time-effectively augmenting clinical decision or triage-making abilities through individualized data-driven analytics (provider), all while improving quality of life and longevity outcomes for patients.
We have only just begun to scratch the surface with the number of possibilities: n! / r! * (n − r)! ...
Solutions that will make us realize our health had been in the dark 99% of the time in retrospect.