Medical Devices: Arduino Version
Part of the hands-on work for one of my classes was to create versions of common medical devices using microcontrollers and additional electronics. Here are some of the main ones.
Arduino Baby Incubator
Objective: Design a arduino-based incubator that will maintain a temperature chosen by the user and alert the user if temperature is abnormally high or low for a baby.
Motive: Learn more about hardware-based closed-loop feedback systems inspired by a traditional medical device.
Key Specifications:
Be stand alone from the computer.
Display the current temperature and setpoint in units of Celsius.
At any point be able to change the setpoint and hysteresis.
Always bound check the setpoint and hysteresis, ignoring unreasonable values.
Alarm when the temperature is abnormally high or low
Responsibilities: All of the projects in this page were done by my friend Rishi Rajendran and I.
My responsibilities were:
Design the circuit.
Build all hardware components.
Troubleshoot connections using an oscilloscope and a multimeter.
Review the final version of the code.
Rishi's responsibilities were:
Code the user-interface to interact with the heater circuit.
Analyze the circuit's performance and choose temperature boundaries.
Design an alarm system for the user.
Review the final version of the circuit.
Since we worked as a team, we often helped each other and overlapped duties.
Results: Since this was part of the lab component in my Medical Instrumentation class, we had a handout that guided us through some of the parts of executing this project so it could be completed under 3 hours. That included general instructions on the purpose of specific components and recommendations of what parts to use. In the end we still had to do design and build ourselves.
Our final product used two main circuits: a Heater Circuit and an Interface Circuit. The former was as follows:
It consisted of a thermistor, a relay, a transistor, resistors, and diodes. The Interface circuit consisted of an Arduino Every, buttons and an OLED.
This is our final mid-fidelity prototype for this project:
Takeaway: This was a great experience for me to design something even more similar to medical devices. I learned a lot about each component as I researched specifics about each one and troubleshooted the circuit. While this project's goal was to end at this stage, I would have loved to design a PCB to increase fidelity and design a enclosure that would improve this product's interface.
ECG Biopotential Amplifier
Objective: Design a biopotential amplifier to be initially tested with a Instruments TechPatient Cardio ECG simulator and then used to obtain our ECG signals.
Motive: Learn about different ways to filter and manage biosignals to obtain clear results.
Key Specifications:
Battery powered.
Safe for human testing.
Be able to distinguish parts of the ECG waveform.
Not include a microcontroller.
Responsabilities: In this project, Rishi and I did a little bit of everything together, so we worked with the following responsibilities:
Design a highpass filter to block DC bias.
Design an inverting lowpass filter.
Build and design the final circuit.
Troubleshoot the hardware to refine the resultant waveforms.
Test and document ECG signals.
Results: The circuit used to amplify and filter the coming signal was as follows:
Here is a video showing live ECG signals:
As seen above, it is possible to detect with a reasonable resolution some of the intervals and segments of the ECG waveform. However, it is still very noisy and not ideal. Unfortunately we had a small amount of time to work on this, but if time allowed, we would have loved to add more filtering to clear these signals and increase the fidelity of our design.
Takeaway: One of the main difficulties in working with biosignals is the amount of noise and interference present. Therefore, this was a great opportunity to learn more about ways to overcome these difficulties and value the importance of each component in a circuit. It was also a great oppportunity to not use software to filter the signals and make the most out of the hardware available.
Ultrasound Imaging
Objective: Design and build a low-fidelity Ultrasound imaging system to clearly distinguish different distances.
Motive: Similarly to the biopotential amplifier, this project helped me learn how to manage an incoming signal to clearly output a desired outcome. It also explored important imaging components, such as image resolution, point spread function, etc.
Key Specifications:
Accurately measure a specific distance within the system's resolution.
Not include a microcontroller.
Distinguish an object in front of the sensor.
Responsabilities: Rishi and I also divided our responsabilities in this one.
Built and design each component of the necessary signal modifications needed.
Troubleshoot circuit components to refile image outputs.
Calculate circuit characteristics, such as ranging error, differences in frequency, cycle excitation, etc.
Results: This is the final circuit that we designed:
And this is an example of the system working to detect an object and output its measurements in an oscilloscope:
Takeaway: Even though this is a low-fidelity sytem, it introduced us to imaging concepts and design of ultrasound imaging. We had a ultrasound machine right next to our table, and it was interesting to see the correlation between the ultrasound machine and our outputs in the oscilloscope. With more time and resources, we could definetely progress to results closer to the ultrasound machine.
Photoplethysmography
Objective: Build a prototype of a medical device that can be tested using an oscilloscope and user data.
Motive: In this project, we had 3 weeks (versus 1 for the other project) to work on it. However, we did not have any blueprint to follow. It all had to come from our previous experiences and things learned in class. Therefore, it was a great opportunity to learn how to make things from complete scratch and apply what we learned.
Specifications: For the original project outline:
A prototype that functionally works as one of the available medical devices in the lab.
Uses only components available in the lab and used during the semester.
Produces outputs within 10% of error from the output from the medical device.
Has an innovative unique feature
For our project, we chose to do a photoplethysmography (PPG). Because of it, our specifications are also:
Measure heart rate using this system.
Be battery powered.
Include an interface that allows easy live output observation.
Has an additional enclosure that allows data capture.
Responsabilities: While we were both involved in all parts, my main responsabilities included:
Design a circuit to detect light differences in the finger.
Design a filter to reduce noise from the biosignals.
Troubleshoot and optimize circuit components to refine the output.
Rishi's duties were:
Code interface to correctly capture heart rate for every minute.
Design an enclosure to fixate the sensor on the finger.
Support circuit troubleshooting.
Results: Our main sensing and filtering circuit was as follows:
I am sorry for the low quality. Someone passed this image through a circuit in the lab. Just kidding.
The results were as follows when compared to the Devon Medical PPG available in the lab.
The results were decentely well. Even though there is still a considerable difference in certain scenarios, the expected trends from exercise were noticeable and the results are close.
Takeaway: I really enjoyed this opportunity to design something from scratch and create a reasonable product. The options for this project were the PPG, an ECG, a Spirometer and a Blood Pressure Meter. I really wanted to do the PPG because it seemed like the most electronically challenging, and I certainly learned a lot from it. If we had more time, we would have loved to design a better enclosure for the device, and create a PCB to clear the use of so many wires.