Physiofunctional circadian metabolism...

PFun Glucose Demo

Interactive simulation of circadian-ultradian glucose dynamics (over a 24-hour period)

A quick prelude...

(EDITED: 2025-08-19)

Before you launch into these results... I'll give you some extra special ~~~insider information~~~👀👀...🤫👀

I have another demo that's almost ready to show off...

Likely I'll have some time to write a sufficient blog post in the near future.

For now, I'll leave you with a tasty hint...

aRtIfIcIaL 🤖🚗 nEuRaL 🤖🚗 nEtWoRkS 🤖🚗...🚗🚗🚗... . 🤖🚪...🤸🤸🤸

~~~🤔🤔🤔...~~🫢~~~...🤫.~~🤭~~~~

---

 

An interactive biophysical simulation of glucose dynamics, estimated over a single 24-hour period.  The curves shown represent: Meal-likelihood [0.0, 1.0], Glucose [0.0, 2.0], Insulin. This is a simplified version of the full model to avoid too much computational overhead (we're really maxing out Desmos here).

Messing around with music & stuff

(Updated: 2023-05-16)

I've been messing around with the read/write methods in scipy.io.wavfile to make some "music" using the output of some of my biophysical models.

Example 1

So for example, here's some simulation output with an audio track (caution, may be loud):

So another interesting point... these sounds can be parametrized...

Example 2

Here's another example with different parameters (again, caution of loudness):

So as you can see, there's a wide range of different sound patterns that can be produced, even with just this one model.

Next Steps...

I'm thinking I'll make a public API at some point in the near future so you can make your own physiological music... 😉

Update!

Some new examples... with rhythm!

Thanks to my friend's sage advice, here are a few examples with a subset of parameters mapped to band pass filters.

Example 2-0

We can get some rhythm by utilizing a binary column as a low-pass filter...

Example 2-1

Example 2-2

Here's another with a few different band-pass filters that produces an interesting pattern...

Dynamical systems model of cardio-respiratory interactions

Currently I'm working on a model of cardio-respiratory interactions. Given human experimental data, the model reproduces the average heart rate, the average respiratory phase duration (inspiration/expiration), and the Respiratory Sinus Arrhythmia (RSA). The most sophisticated part of the model is the respiratory central pattern generator (previously published). My current goal is to optimize the model so that it reproduces RSA for several individuals.

After I get it working with the current dataset-- which was measured while participants watched Disney's Fantasia-- the next step is to fit the model to real clinical data from septic ER patients. The end goal is for our collaborators to use our model as a basis for a machine learning algorithm that predicts the risk and most likely cause of death for septic patients.

Below are some pretty pictures generated by the model.