I have a friend which is working with COVID-19 cases in Marseille Hospital. For her patients that are intubate with machine they are treated with medicines like morphine that they cannot dose precisely, they do medical examination each 4 hours to grade the pain of the patient with they level of crispation as they cannot speak. But in heavy cases they use Curare that inhibit all the muscles for a better performance of the air pump. In this case they have no mean to know the level of pain of the patient and if the patient is awake or not. So they give standard dose for everybody of pain killer and hypnotic hopping for the best.
EEG are now easy to buy and some cheap model are available in the market. With low prices and much more easy to use. Before an EEG was minimum $10 000 with the need of a medical assistant to install it and use it. Lots of studies have been published on the capacities of EEG to detect different state of the brain like sleep or heavy pain. Medical care needed for pain in hospital are based on morphine derivative need specific dosage as the effect can vary between people. Too low dosage will not be effective and too high may put the patient in danger as his body will not react properly anymore.
What it does
Measure the brain waives with a light EEG, depending on what is is observed is able to indicate the sate of steepness of the person and the level of pain and is able to do action on medication.
How I built it
I have a Muse EEG that have an old open library for Android. I reuse EEG-101 witch is an open source project issue from Mozilla Open Leaders X for teaching basic of an EEG with Muse. I create a real time socket connection to upload the data to a server, process them and make action depending of the result on an electronic device. The ESP8266 that I have used put it led on to simulate action on medical device like perfusion of medicine.
Challenges I ran into
Working with different programming languages is always tricky as to do the same thing you have different command and you can't mix them. Here is Javascrip/react-native, java/android, python, c/embedded system.
Accomplishments that I'm proud of
Stable real time network communication. Link between EEG device/Smartphone/server/electronic
What I learned
I have an engineer background on electronic, signal processing, system and network and 15 years of experiences on various development languages. So not much technically that I already know but it is still fun.
What's next for EEG State Monitoring
Use EEG existing database for machine learning to improve the detection of state, the biggest public one is 4 terabit of data. Work on the safety of the patient data and security. Work with regulation have it certified as a medical device. Work with manufacturer to designed a more adapted one for the hospital.
java, muse, python, react-native