Digital Therapeutics for Disease Control
We are building the worlds first digital therapeutics using AI/ML to analyse biological and electromagnetic signals emanating from human bodies and using the analyses to alter the underlying mechanisms to enable personalised prevention.
Let’s unpack each or these concepts. First, at every instant we are alive, we interact with our environment. We receive signals from our environment, we give out signals to our environment. For example, right now I am receiving signals from my environment in terms of the time I have for this presentation, the cues about my audience and so on. At other times I may have received cues and signals in the form of food and beverage and at all times I revive air to breathe and the remaining five senses to interact with the world around me. At the same time, I send out signals. My brain sends out electromagnetic waves, my heart send out electrical signals, the flow of blood in my veins and capillaries send out signals that can be captured and processed, and my blood components and metabolites send out signals to the environment every time I pass urine and faeces, reached out chemicals, and effervesce body odour. My heart rhythm, my respiration patterns are unique to me and together these signals define my physical, mental, and social health snapshots. These cues and signals are altered when I am healthy as opposed to when I am ill. Doctors, in particular internists pick up many of these signals to decide my treatment for illnesses and decide what needs to be done. Until now, there have been two limitations of this approach: (1) almost all of the signals and data that we humans as doctors collect form fellow humans to manage conditions are retrospective, meaning that we get to measure these metrics after an event. For example, for a patient with diabetes, the doctors use a metric glycosylated haemoglobin or HbA1c to plan and monitor their treatment. Glycosylated haemoglobin is a measure of the concentration of glucose in the blood that is attached to a persons red blood cells. Red blood cells have a lifespan of three months and therefore the higher the concentration of HbA1c, the higher the three monthly average of glucose level in the blood. This fact is utilised to monitor patients with diabetes to monitor their level of control and plan treatment. If the level of HbA1c is kept at a level of 48 mol/mol on three subsequent measurements, doctors would claim that such a person with diabetes has reversed the state of diabetes. If the level of HbA1c is maintained at a level of 55 mol/mol, the experts claim that it is possible to avoid serious cardiovascular complications for such a person with diabetes.
This last point is important because here you can see that the chemical signal is treated as predictor or what can happen afterwards. At the same time, the measure itself is a retrospective measure because it can provide any indication only after the fact. This is a limitation we can address using signal analysis and real time monitoring of individuals and then using AI/ML tools to develop predictive algorithms. This is where we are making a difference in our approach.
Our bias is as follows. If we are able to identify, capture, and measure electronic signals that result from underlying biological or physiological or pathological processes, then combine these signals and use AIML to develop individualised models and determine prognosis, and then iteratively attempt and measure such processes. T In turn these help to control the condition.