Hydration & Microdosing Caffeine
Executive Summary
Paracelsus — “dose differentiates a poison from a remedy”.
We use machine learning (ML — TensorFlow/Scikit-learn) and real time health data to tell the user when to microdose medications.
We access watch (wearOS 3 / iWatch) sensor data, health (Google Health Connect / Apple HealthKit) and even personal (calendar) information to determine the best time to dose a medication using our trained machine learning (ML) models.
Our initial product is targeting caffeine. We notify the user to take a micro shot of caffeine dependent on the watch biosensors (heart rate / blood pressure) and health data (eat, drink, sleep, workout etc …).
It’s the first time in history an energy drink is controlled by a machine learning algorithm on a mobile device.
We are starting with caffeine but if we get this working we will approach pharmaceuticals about using machine learning with health data to time the microdosing of cancer drugs.
Our system is very secure as we only read data. We do not write, share or otherwise alter / transmit any data. Again, we only utilize local data to determine when to microdose.
We use rapid development:
And a very advanced mobile architecture:
Website:
Original concept video:
Our background is in machine learning (ML), mobile development and biotech.
We can provide more info about this project to utilize physiological data to know when to microdose medications which are precisely timed to maximize benefits and minimize side effects.
Hummingbird Fuel
Energy all day & Sleep all night
We introduce a machine learning (ML) mobile application (Android & iOS) to help with real time tracking & notification of hydration / caffeine for peak performance.
We start by exploring how hydration and microdosing caffeine can help us maintain our energy levels.
Dehydration causes fatigue
Dehydration has many negative physiological effects.
“If you’re thirsty, you’re already mildly dehydrated, and that can cause symptoms like headache, fatigue, dizziness and more. Dehydration can contribute to life-threatening illnesses like heatstroke.”
~~~
Dehydration leads to many negative consequences.
“In HIGH the restricted water intake resulted in a significant increase in thirst and a decrease in contentedness, calmness, positive emotions and vigor/activity. In LOW, increased water consumption resulted in a significant decrease in fatigue/inertia, confusion/bewilderment and thirst and a trend to lower sleepiness compared to baseline.”
~~~
Hydration is important for a healthy body.
“ The effects of water on daily performance and short and long-term health are quite clear. There are few negative effects of water intake and the evidence of positive effects is quite clear from the literature.”
~~~
Hydration will fight fatigue.
“Consuming a sufficient amount of fluids in beverages and water-filled food (such as fruits, vegetables, and soup) will help replenish the water your body loses throughout the day and can help you maintain your energy”
~~~
The first important point:
- To maintain high levels of energy throughout the day we must be adequately hydrated.
Microdosing Caffeine
Microdosing caffeine has been shown to have many health benefits.
~~~
Liquid ☕️ vs Pill 💊 vs Spray 💨
We look at the absorption, distribution, metabolism, and excretion (ADME) response of caffeine in multiple forms.
There has been recent interest in the ergogenic effects of caffeine delivered in low doses (~ 200 mg or ~ 3 mg/kg body mass) and administered in forms other than capsules, coffee and sports drinks, including chewing gum, bars, gels, mouth rinses, energy drinks and aerosols.
We see that Caffeinated aerosols still need more studies …
Caffeinated aerosol mouth and nasal sprays may stimulate nerves with direct brain connections and enter the blood via mucosal and pulmonary absorption, although little support exists for caffeine delivered in this manner.
~~~
The Second important point:
- Microdosing caffeine has many positive effects while mitigating the negative ones.
Physiological Mobile App (Android & iOS)
Hydration & Microdosing Caffeine
App uses personal physiological information (Google Health Connect / Apple HealthKit) and real time health data (WearOS / iWatch watch biosensors ) for calculating hydration levels and microdosing caffeine notification times.
Google Health Connect
On Android we use health signals from Google Health Connect .
Health Connect supports more than 50 different data types that give users the ability to measure everything from their cycling speeds to their body temperature, and more.
- Activity: This captures any activity that a user does. It can include health and fitness activities like running or swimming, meditation, and sleep.
- Body Measurement: This captures common data related to the body. It includes capturing a user’s weight, or a user’s basal metabolic rate, among other data types.
- Cycle Tracking: This captures menstrual cycles and related data points, such as the binary result of an ovulation test.
- Nutrition: This captures hydration and nutrition data types. The former represents how much water a user drank in a single drink. The latter includes many optional fields, from calories to sugar and magnesium that record which nutrients the user consumed.
- Sleep: This captures interval data related to a user’s length and type of sleep.
- Vitals: This captures essential information about the user’s general health. It includes everything from blood glucose to body temperature and blood oxygen saturation.
~~~
Apple HealthKit
On iOS we use health signals from Apple HealthKit.
“HealthKit provides a central repository for health and fitness data on iPhone and Apple Watch. With the user’s permission, apps communicate with the HealthKit store to access and share this data.”
Using all this data we notify the user when to drink and/or take a shot of caffeine.
CardialKit
A reference for connecting to Google & Apple Health API.
“CardinalKit empowers the digital health community to rapidly prototype and build modern, interoperable, scalable digital health solutions on a variety of platforms.”
“We provide you a suite of tools to build your modern, interoperable digital health experience from the ground up, from the app itself to storing and analyzing collected data in the cloud.
CardinalKit is designed to accelerate rapid prototyping of digital health applications by reducing costs by as much as 75% (~$150,000) and timelines by 12 months. Available on iOS and Android!”
Source
Android Source
“An Android application for the CardinalKit framework, developed with Jetpack Compose, Material Design 3, Firebase, Health Connect and Android FHIR SDK. CardinalKit aims to accelerate the development of digital health applications by providing a pre-integrated template application in this repository that you can use as a starting point for your project.”
Follows best practice — Clean Architecture
- Create and deploy healthcare surveys using HL7® FHIR® and a drag-and-drop survey builder
- Schedule daily tasks for your users to complete including surveys, medication doses, activity goals, etc.
- Collect and upload health data using Health Connect
- Simple Jetpack Compose templates that you can customize
- Supports Dynamic Colors based on the user’s wallpaper or color choices
- Pre-integrated with Google Cloud Firebase backend
iOS Source
“This repository contains a fully functional example in the
CardinalKit-Exampledirectory that you can use as a starting point for building your own app. To get started, clone this repository and follow our simple setup instructions.”
- Informed consent process using ResearchKit.
- Track day-to-day adherence with CareKit.
- Monitor health data with HealthKit.
- Collect and upload EHR data.
- CoreMotion data demo.
- Awesome SwiftUI templates.
- Zero-code customizable configuration file.
- GCP Firebase Integration.
Hydration Tracking
Our mobile app tracks hydration data. More details coming soon …
- Our mobile App tracks real time hydration levels.
→ Full article explaining how the hydration app works coming soon … ←
If the user is dehydrated they might become over caffeinated. So it is very important to be fully hydrated when microdosing caffeine.
Microdosing Caffeine
For our initial product we will start with an aerosol delivery system because it provides fine grain control over dosing. The user can always increase the micro dose by just increasing the number of pumps. Pills, gums and bars do not allow for the same flexibility and convenience of dosing. Also, some research shows that caffeine has heightened bioavailability in a atomized aerosol mouth spray.
Below is an image of our atomized aerosol pump dispenser:
Our initial formulation (for example) is a standard 2oz energy drink.
Formulation has three levels:
- Air Nectar
- Water Nectar
- Fire Nectar
Cost $7.99
- Spray bottle (reusable)
- 2 fl oz of Energy Drink
- ML App [Android & WearOS, iPhone & iWatch]
Caffeine Tracking
Mobile app shows the amount of caffeine in the body at any given time. The darker the purple color the larger the amount of caffeine in the body.
Example:
- Shot of caffeine was taken at 3pm (yellow bar)
- Half efficacy at 4pm (the intensity of the purple shows the amount of caffeine active in the body)
- Very little effect at 5:45pm (almost no purple color after 5:45pm)
- User is ready to sleep after 7pm with no caffeine in their system.
→ Full article coming soon … ←
Sleep Tracking
Mobile app shows the stages of sleep throughout the night. The darker the purple color the deeper the sleep.
Example:
- Bed @ 11pm (Blue Dot)
- The darker the purple color the deeper the sleep.
Purple wedges: 11pm — midnight = Light N1 sleep, midnight — 1am = N2 sleep, 1am — 3am = REM sleep, 3am — 4am = light N1 sleep, 4am — 5am = N2 sleep, etc …
- Woke up @ 7am (Yellow Dot)
Full article:
Fitness Tracking
List of activities can be displayed by clicking the center of the circle.
Directions:
- Click on center icon to see the next percentage of activity.
→ Full article coming soon: ←
Meditation
Reduce stress with biofeedback from heart rate monitor.
Directions:
- Breathe in as blue circle gets bigger (expand lungs)
- Breathe out as blue circle gets smaller (exhale)
- Red heart beat is actual heat rate from WearOS / iWatch sensor
- Concentrate on reducing heart rate and relaxing mind.
Full article:
This is still a work in progress and any feedback is welcome.
Machine Learning
We use the above (Google health connect, WearOS, Apple HealthKit, iWatch) as input to our ML model to make predictions for the next timed shot of microdosed caffeine.
In the next part of the series we will explain (at a high level) how the ML predictions work.
Thank you!
