Sparksam's Blog

AI-Powered MCPs

Samuel Klutse · · 3 min read
Updated
AI cybersecurity science

MCPS? AI? Let’s quickly define a few terms before you skip this article for the next one. An MCPS simply means Medical Cyber-Physical System and Internet of Things (IoT) devices such as Pacemakers, and Generic Patient Controlled Analgesia (GPCA) fall under that category. AI stands for Artificial Intelligence and I bet this one you knew it. AI has a diverse domain of applications and touches every sector of business such as transport, healthcare, banking, retail, entertainment, and e-commerce to only name a few.

A recent project I worked on for one of my classes consisted of engineering a MCPS constituted of a physical device with sensors with processing ability, mobile software, and a cloud platform that will store and analyze the aggregated from different sources. The focus however for that project was on the security aspect, the different Trusted Platform Modules (TPM), the protocols for encrypting the data while it is transferred from one device to another, and also the pros and cons of using a certain technology instead of another one (Bluetooth or Wifi for example). I will give a more detailed article about the Heart Rate Monitoring System that I engineered in a later article.

Digital thinking — a 3D rendered brain on a gradient background Photo by Milad Fakurian on Unsplash

Where Is All This Talk Leading Us?

Well, an idea struck me while working on that project.

Let’s dream a little bit.

Imagine a world where we can read dreams.

At least, record them. (Yes, I know recording nightmares too but let’s just focus on the dreams).

Let me explain…

According to Blocka, K. (2018), an EEG or Electroencephalogram is a test used to evaluate the electrical activity in the brain. Brain cells communicate with each other through electrical impulses. An EEG can be used to detect problems associated with this activity. In healthcare, EEG tests have been used to confirm or rule out various conditions such as seizure disorders, head injuries, encephalitis, brain tumor, sleep disorders, encephalopathy, stroke, and dementia. They are usually painless and very safe for patients. One of the cons is that several types of movements such as pulse and heartbeats, and breathing can cause EEG recordings that mimic brain waves and it’s up to the EEG interpreter to take those into account.

Many EEG headsets made their way onto the market recently and are available for the general public. According to Farnsworth, B. (2019), many EEG headsets can range from $99 to $1000.

I believe the market availability and the low price of these devices create a new set of opportunities. An MCPS can be built so that the brain waves recorded from these headsets (during sleep probably) can be sent to a mobile application that will then sync in real-time to a cloud platform. The aggregated data can then be analyzed using AI, to either find patterns or predict some abnormalities for some patients.

A Few Use Cases

A pattern can be found in a certain user’s data showing a rise in brain activities during his sleep around a certain time of the month or the year, which means the user during that time is stressed for example.

Another use case can be a certain patient presenting a brain wave signature recorded from people with brain tumors, thus implying the user should get checked for tumors and early detection can mean saving a life.

The idea behind this MCPS is not new and is already being used by our Apple watches which record how well we slept or take our pulses to let us know whether we are stressed or not.

Yes, the MCPS I described in this article does not cover the dream translation part but in a near future, a technology to translate the brain waves will see light and therefore shed some light on our dreams and nightmares.

References

Related Posts