Will detect higher Hz vibrations. up to 100Hz. Could be a second hill in Beau_Gunderson 's FFT graph. Possibly hill is at 5 Hz because your current app only samples 5 times per second. More likely 10 times.
the data is interpolated (necessary for feeding uniformly spaced data to the FFT) so my guess is it’s pretty accurate right now—I plan on running all of the newer files through when I get a chance though, and of course it would be nice to compare a 10hz sample just to see how it differs
100hz. you are using 10 Hz I think
I ordered a cheap Android phone to try out SensorRecord - should arrive Friday…
I’ve been wondering: is my tremor worse when heart rate and blood pressure higher? Tremor is sometimes treated with beta-blockers, which are also used to treat high blood pressure.
Beta-blockers, also known as Beta-adrenergic blockers, beta-blockers are a class of drugs that block beta receptors in the brain. Beta-blockers decrease the effect of the sympathetic part of the autonomic nervousbsystem and are commonly used to reduce high blood pressure and treat migraine. They can also suppress tremors in many patients with ET. There are several types of beta receptors. It is thought that the development of drugs that could act specifically on individual beta receptors would lead to more effective tremor control. Propranolol (Inderal®) is a beta-blocking drug.
So, is my tremor worse when my heart rate is high? Tonight I did not feel very tremory, so I wanted to check with the g-Force sensor, and so I did the test right after measuring my blood pressure and heart rate. The three BP/HR measures are below. My median BP/HR in the last 30 days 128/80/64. Here BP is normal (for me), while pulse is higher.
HR is probably high because I recently got back from coaching my daughter’s soccer team, which is always a lot of exercise.
Tremor, on the other hand, appears to be somewhat low, which is what I’d guessed.
sensor.csv (341.4 KB)
I did some more reading on beta-blockers and tremor, and at least one recent source explains that the mechanism of action of Propranolol in reducing tremor isn’t through lowering heart rate and blood pressure, but through interfering with “effects of peripheral noncardiac beta-2 receptors located in the muscle spindles.” (See Pharmacotherapy of Essential Tremor by Peter Hedera, František Cibulčík, and Thomas L. Davis
So the BP/HR and essential tremor idea may be unproductive.
Nice papers! Indeed, it would seem like HR/BP and tremor are both affected by adrenaline (/epinephrine), and that is why they’re both affected by beta-blockers - which work by inhibiting the effects of adrenaline.
From the same paper (Pharmacotherapy of Essential Tremor by Peter Hedera, František Cibulčík, and Thomas L. Davis’): ‘Epinephrine upregulates the sensitivity of muscle spindles, leading to increased rhythmic afferent activity and, thus, higher synchronization of afferent signals and enhanced reflex activity.’
Thank you Valeria. I think what you’re suggesting is that upstream causes can influence both HR/BP and tremor, and therefore we might expect them to vary together. That makes sense to me.
Also, an “epistemological” conclusion from looking at the scientific literature on tremor is that I’m unlikely to understand it deeply enough to learn how to influence my own tremor. Just as it’s perhaps possible to master the coding skills to develop my own app for tremor measurement, it’s theoretically possible for me to make a discovery in the physiology of tremor - but it’s not practically possible in either case. Instead, I’m better off making (as simple as possible) measurements of what’s actually occurring, and tinker with the problem directly.
I’m still curious about BP/HR so I decided to keep exploring.
Pre-measurement predictions: 130/80,70; Tremory:4
Measurement result: AVG: 139/82,68 Tremor:4/5
(Actual measurements: 129/89,68 | 144/77,68 | 145/80,69)
sensor-1.csv (262.3 KB)
I define tremory as: a feeling of vulnerability to tremor that is noticeable in daily activities without intentionally inducing the measurement tremor.
- very low tremory
- somewhat low tremory
- typical tremory
- somewhat high tremory
- very high tremory
However, I do not have this level of precision in my measurement, since I’m just eyeballing the graph. In this measurement, there is clearly somewhat high tremor, but I don’t have an objective definition of the difference between somewhat and very high. I hope to get some help from a friend in scoring my data so I can have a 5 pt tremor score.
Prediction: 130/80,70, Tremory: 2
Measurement: AVG: 128/73,68 Tremor: 3
(Actual measurements:141/86, 67 | 117/68, 66 | 125/69, 66)
sensor-2.csv (301.7 KB)
This is a case where my estimate of tremory is off. I was more tremory than I thought.
@Agaricus I think I’ve reproduced this (much thanks to the screenshot and slack responses from @Beau_Gunderson). I tried adding an additional measure I think you wanted – reporting the peak frequency and the magnitude at that peak.
20190829 MPB tremor analysis.ipynb (14.1 KB)
I can help walk you through running this if you’d like. It’s a manually uploaded file, so Open Humans doesn’t provide any OH-specific value for this (yet), but does give you a standardized personal Jupyter instance.
Thank you @madprime! OK, now I have to install Python. How hard can it be?
I didn’t make predictions this morning or have a chance to to make my regular measurements before coffee and food,=, but during our “quantified heart” PLR decided to measure: so this one is an “at work” measurement.
Measurement: AVG: 142/88, 58 Tremor: 3
(Actual measurements: 136/85, 57 | 149/87, 58 | 142/92, 58)
201908301016.csv (504.5 KB)
@madprime messaged me to point out that I didn’t need to install anything, I could just use Open Humans. Right!
I signed into Open Humans, uploaded the file, and could see the output. Peak magnitude is 58.5. I’ll have to work with this a bit and see how it goes.
I’m been away for Labor Day weekend and not attempting to keep my measurement schedule. I did have a chance to measure just the tremor this morning. I felt very tremory and predicted a 5. But the results were actually unusually low:
090220190734.csv (269.7 KB)
@Agaricus I’ve updated this notebook a bit and shared it in @gedankenstuecke’s “exploratory” – hopefully this makes it a bit less confusing how to open and use it. (Among other things, there’s an “open in notebooks” button on this page.) https://exploratory.openhumans.org/notebook/71/
It’s likely you see this and think “great but now I want to do X”, let me know, I can think of a variety of next steps (and others might have more ideas too). In theory this is becoming increasingly re-usable for other people, should you find anyone willing to try to reproduce your project on themselves.
@madprime Thank you.
I got a bit tired of this project after last week, thinking: “Nobody knows how to fix it, why am I taking so much time measuring it?” However, I continued to notice (without measuring) that the tremor seemed to vary a lot. It is only very annoying a few times each day, so if I can reduce these incidents the benefit will have been worth it. I also wondered: How much time is this really taking? Is it possible I’m thinking “takes too long” because I’m just a bit discouraged for other reasons? Let’s see how long it really takes.
So, my goal is to do three days of frequent measurement, starting this morning and ending Friday. I hope to take at least 4 per day. The spreadsheet with results is here: Tremor and Blood Pressure Spreadsheet
This morning’s results:
Prediction: 130/80,70, Tremory: 1
Measurement: AVG: 132/85,60 Tremor: 4.68
Most interesting result this morning is that I predicted very low and my results were also very low. I barely felt any feeling of “tremory.” This showed in the data. I’m interested in this because I hadn’t yet tried to do any fine motor tasks. Is there an overall sensation of tremory that may be based on subconscious noticing of difficulties performing motor actions or on other factors? I’d like to think my intuitions are reliable but it hasn’t always been the case that my predictions are accurate. Image of the measurement:
Finally: It has taken over forty minutes to complete this process. That’s a lot of time! Some of it is working out a relatively new workflow and also writing the notes. I’d really like to do this in less than 15 minutes each time, otherwise I can’t do it more than once a day.
I’m about halfway through my 3 day intensive tracking protocol. I’ve reduced the measurement time to about 7 minutes per session, which is easily manageable. But I’ve done this by focusing on data collection only - no analysis. I collect the BP measurement, which is stored in the monitor and transmitted to my phone via Bluetooth, using the BloodpressureDB app. (The app has worked well so I’m giving it a link.) I message the data from Physics Sensor Toolkit to myself using the native iOS sharing tool. I don’t update the spreadsheet or post here. I’ll check back in here at the end of the three days, assuming this continues to go well, with what I’ve learned.
Hi Gary, when I open the spreadsheet I see it empty!
My project stalled due to high complexity and I’m going restart with a simpler approach. I’m going to measure daily between 10 am and 11 am (which is a time when I often notice my tremor) and see what the variation is on a daily basis. I’ll try to complete 10 measurements. On Tuesday I tested the latest notebook iteration:
Today, after troubleshooting by @madprime, I got another graph:
I felt that today’s tremor was worse. I have a very basic question: do the numbers match my estimate of tremor severity, the way they seemed to when I looked at the pictures at the beginning of this project?