Wearable Therapy Tracker

     When I started this project, I wanted to create a piano on a stuffed animal so that my grandmother, who has Alzheimer’s, could play the piano again. I found that a lot of other students were using sound as the communicating element of their projects, and I decided I wanted to create something that didn’t use sound but that would still be usable for my grandmother. My next project idea, that I suddenly scribbled into a notebook while someone else was presenting their idea (sorry about that!), was to create a tracker inside a stuffed animal that when pressed would record a data point. The idea being that when an Alzheimer’s patient was anxious or felt uncomfortable, the amount of data points would go up, helping the care provider to know if something was bothering them, even if they couldn’t verbally tell them. I posed the problem to my mother (ICU nurse and my grandmother’s primary caretaker) to bounce some ideas back and forth. Her suggestions were to create a tracker with a time element and a gps that could be placed inside a person’s comfort item to track when and where the patient felt anxious. For example, a tracker could be sewn into the corner of a blanket for a patient with nonverbal autism, nonverbal down syndrome etc. to help understand what is making the patient anxious. Identifying what the problem is can be extremely difficult for patients that are nonverbal or can’t communicate with you because they can’t simply tell you what the problem is. The other suggestion, which is what I went with, was to create a wearable tracker for behavior modification therapy in patients with things like Anxiety, OCD, ADHD, etc. These can be extremely debilitating for the patient– interrupting daily life. Therapy can also be extremely difficult because of the time it may take for a therapy to work, which can be frustrating and discouraging, and the judgement from others against those with mental illness (“It’s all in your head” “Get over yourself” “I get nervous sometimes too”). Therapy itself can be be a trigger for some because they worry that the therapist and others may judge them if the therapy doesn’t work immediately. Many patients will give a false positive report that the therapy is working because they worry the therapist or others will judge them if it doesn’t. In the long run, this hurts the patient even more because they aren’t able to find a therapy method that truly does work for them. By creating a therapy tracker, the goal was to help alleviate any anxiety caused by therapy and to give an objective way for therapists to see if the therapy is helping the patient. The basic principle of the tracker is like a stress ball that can count for you. The patient will press on the center of the striped fabric when they feel anxious and the arduino records a data point. By hiding the number count within the arduino rather than in something like a tally counter, it can help prevent any anxiety that could be caused by seeing the number go up. This data could then be seen by the therapist as a way to understand if the patient is anxious. By comparing the number of presses to a baseline it would be possible to see if the therapy is working. If the number is lower than baseline, then the therapy is working. If the number is higher or the same as the baseline, then the therapy is not working. There is a two-fold purpose in hiding the tracker in a bracelet. First, by hiding it in a bracelet, it looks like normal clothing preventing any anxiety or judgment from knowing that someone is wearing a tracker. Second, people tend to play with clothing, jewelry, pens, etc. when they are nervous. This allows the person to continue those calming, physical ways of alleviating any nervousness or anxiety in a way that can be quantified in a helpful way.

     Once I had the idea of what I wanted to do, my first major hurdle was coding the arduino. I do very poorly when it comes to coding, so I owe a big thanks to Professor Antonius for helping me write the code. In order to record data into the Arduino, we wrote in an EEPROM record function. When the button, ultimately the pressure sensor, was pressed it would record in the EEPROM an increase in the variable. However, with this alone, it may record three, six, or a whole bunch of points depending on the length of the press in the EEPROM instead of just once when the button was pressed. Professor Antonius suggested to use a debouncer into the code to make it record one data point in the EEPROM when the button was pressed once. In order to do this we wrote in a Boolean variable. When the button was pressed it would register that as a “false” and update the EEPROM once. It would then change the “false” to “true”. If the button was still registering as “true” (ie holding the button) it would not update the number. When the button was released, it would go back to “false.” So in simple terms, even if you press and hold the button, it will still only record one data point, making it just like a stress ball. To record six data points, press six times, etc.

      Once the Arduino was coded, I could start on transferring the hardware from a bulky breadboard and button into the  slim(mer), wireless bracelet that I wanted to create. I asking the equipment room for a prototyping shield with a breadboard so I could slim down the design, all they had with a breadboard was a cellphone shield, but it still ultimately worked for my prototyping. I put the button, a resistor, and two 3.7 V battery packs onto this using wires. I used all black wires so it would blend in with the black fabric. In order to make sure I didn’t short circuit it, I used tape to create a system I could understand for plugging all the wires in. Masking tape went to the same battery and green electrical tape were the positives. The sensor was created by using two pieces of conductive tape and two places of stranded core wire. The wires are plugged into the Arduino and then each attached to one of the pieces of tape respectively. The wires are stripped and spread on the end connected to the tape. I spread the strands apart and used electrical tape to press them onto the conductive tape. Touching the two pieces of conductive tape together completes the circuit, acting like a button-press.

      I used black neoprene for the bracelet body because it was stretchy for going onto different sized wrists and it the fabric was thick and dark to hide any color on the Arduino and blunt any sharp edges. To make it easier to sew together, I used a water bottle to act as a wrist form kind of like a dress form you would use for creating a T-shirt. This kept me from sewing it so small that I would be the only person it would fit on. I cut out a piece of fabric long and wide enough to create a bracelet when the ends were sewn together. I also added a section at the top that could be folded over the Arduino to create a rudimentary pocket-hem-like-thing. I sewed the ends of the bracelet into the circle shape while it was inside-out so the seam would be on the inside. I then put this on the water bottle so I could add the electronics. I had the Arduino, shield, sensor, and batteries all wired together at this point so I detached the batteries so I could sew it on. I placed the Arduino and shield on the top and folded the fabric over it, sewing the sides down to hold it in place. I left the end open so it was still possible to access the spot for the computer plug-in and see the LED to make sure the Arduino was actually receiving power. I then cut another long piece of the black neoprene so I could make another band over the rest of the band to create a pocket for the batteries to sit in and to act like a rubber band to hold on the Arduino as securely as possible. I taped the batteries together in a V-like shape so they would be more ergonomic on the wrist. I also taped their connection wires together and to the side. I sewed the band around on the rest of the band starting with a seam just around the top first. I then put the batteries in the half pocket that was created. I pulled the connection wires of the batteries close to the Arduino and the Arduino power wires to the same side. When I did the seam around the bottom I did a very long stitch over these wires to hold them in place for easy access but to also make sure they didn’t move once they were wired together. To create an opening for the Arduino in the bottom like I did with the pocket-sem-like opening, I did what I’ve done before to create gathered lace. I did a simple running stitch through the fabric without connecting it to the bottom piece of fabric. Once I got to the other side, I pushed the needle through both piece of fabric and pulled the thread tight to bunch the fabric together around the opening. It’s kinda like the opening in a sock without the elastic. You want it to stretch and be big enough to get your foot in, but not so big it falls off. Gathering the fabric here makes it so you can access the Arduino when you need to, but it holds it in place otherwise. Having this opening also allowed for the wires going from the sensor to the Arduino. I then went to sew the sensor on top of the bracelet, but I discovered one of the wires came untaped so I had to bust it apart to retape everything. Once I repaired it, the wires were still very touchy and required constant fixing. Even when I couched the wires in place they still tended to pull out of the tape, so I should have looked for another method to attach the wires. Here I used 2 pieces of striped woven fabric to hide the conductive tape sensor because stripes are very popular right now (makes it look even more like regular clothing) and they were already the size I needed for my project. In retrospect, I should have used a serger hem on the edge of these pieces because they were beginning to pull apart. After the sensor was sewn on, it covered a seam on the top band and the project was completely put together barring any problem.

     But, there are always problems.

     Once I had it all together, I had a problem with getting it to record data. After showing it to Professor Antonius, we found that I had wired the breadboard incorrectly. Fixing the breadboard wiring did ultimately fix that problem. I then put electrical tape over the top of this so that the wires wouldn’t disconnect when I put the Arduino back in the bracelet. Luckily the opening I left in the bottom was just large enough to remove and put the Arduino back in without having to rip the seam out.

     After this problem was fixed, the project worked just like I had intended, even if the sensor was a little finicky.

     For the future of this project, I would like to add a GPS tracker and a clock to add in the potential to see if particular times of day or a particular place causes the patient anxiety. I’d also like to make the microcontroller and battery pack much smaller and more flexible. A button switch for the battery pack would also be an ideal addition so you are able to turn it on and off without unplugging it.

Further projects with this could include other versions of it for other applications. For example, create a sensor that could be placed on object for OCD patients, such as light switches or faucet handles. It would also be ideal if there was an App on a smart device like a phone or tablet to have a patient recorded therapy evaluation (similar to current models) that could be coupled with the data received by the device to provide additional information for the therapist.

The presentation I created for the IMA show has some additional info in an easily viewable way:

https://docs.google.com/presentation/d/155P_FPRdW8fv-aWIwAznEU0mTvYO4r_xhWvdi5pL_Fg/edit?usp=sharing

     

-Sarah Brooker

#include <EEPROM.h>
int addr = 0;
int val;
boolean pressed = false;

void setup() {
  // put your setup code here, to run once:
  pinMode(13, INPUT);
  Serial.begin(9600);
  val = EEPROM.read(addr);
}

void loop() {
  // put your main code here, to run repeatedly:
  if (digitalRead(13) == 1) {
    if (pressed == false) {
      val++;
      EEPROM.update(addr, val);
      pressed = true;
    }
  } else {
    if (pressed == true) {
      pressed = false;
    }
  }
  delay(100);
  Serial.println(val);
}

Leave a Reply