Today the latest CSM demo video went live on the CardioSolv site. It showcases the use of our mapping interface, which makes it easy to create useful maps of activity in simulation models.
It’s currently non-trivial to show movies in papers, so instead we do time-lapse type things called activation maps. These show the activation times as a series of lines (‘isochrones’ or ‘isochronal lines’, meaning that all of the points on the line are activated at the same time) or bands of color representing the same thing. We can extend this to also show repolarization times, or non-sequential data such as action potential duration maps and dominant frequency maps.
Here’s a sample activation map of a wave moving across a sheet from right to left:
Activation Map Right to Left
And here’s one of a spiral (this with 20ms isochrones):
Activation Map of a Spiral Wave
To give you an idea of the correspondence between an activation map and a movie of the simulation, here’s a movie of that spiral:
There’s a lot more to this — for instance, deciding when a cell has activated or repolarized, and back-end processing. We use a program I wrote that does the analysis in parallel, making it rather quick to analyze even huge datasets, provided you have the computing power.
If you have any questions about the process I’d be happy to answer them here or on the CardioSolv post.
A little while ago I embarked on a search for a blue canary night light for my daughter. After not finding one, I decided to make one. Several prototypes later, I’m almost ready to sell them (at my site, bluecanarynightlight.com).
Today I got some new LED night light bases to try. They aren’t going to work for the final sale models, but I was finally (with the aid of a plug rotater) able to get a blue canary night light set up by the light switch in my daughter’s room:
The final model won’t have legs, and will be bigger (as is the current prototype), but I was glad to be able to make use of this older prototype. I think I’ll be able to get better light diffusion with the final models as well — they’ll be dyed instead of painted.
If you want to hear the story of how I moronically made sparks shoot out of my wall installing the combo switch/receptacle today, let me know in the comments.
Today I’m really excited to finally show you something that’s been in the works, both in implementation and in the planning stages, for a long time. The CardioSolv Simulation Manager.
Running cardiac electrophysiology (and mechanics) simulations has traditionally been really complicated. It involved learning a bunch of UNIX command-line tricks, dealing with queuing systems and their associated script files, and so on. Furthermore, there are many, many options in a sophisticated cardiac simulator, and the novice user (and even the expert) can easily get lost in all of the choices.
We’ve taken years of experience setting up, running, and analyzing simulations to build a really cool (excuse my excitement) web interface that handles all of the dirty work, and guides the user through the important choices when running simulations.
The video below is my first demo. In it, I demonstrate how to create a plane wave moving across a sheet of tissue, then create a spiral wave, all from the web interface.
Today I added two more accountability partner types to didyoudo.it; habit and fitness accountability partners. Is there another type of partner that I should list? Let me know by commenting or one of the other methods listed on my contact page.
The whole article is here.
The HPC service lets the small, five-employee company do the heavy lifting that would otherwise cost a fortune. “With what we could purchase out of pocket, we’d have to bootstrap very slowly, or look for VC [venture capital] funding,” said Dr. Brock Tice, the vice president of operations at Cardiosolv, a privately funded medical research firm. Instead, Tice uses a new HPC on-demand service from Penguin Computing called Penguin on Demand.
While Cardiosolv has its own small cluster on the premises for calculations, Tice estimates the resources he rents from Penguin would probably cost $500,000 to build, and other cloud options weren’t suitable.
“We can’t use [Amazon’s Elastic Compute Cloud] EC2, since there’s a lot of latency between the nodes,” he said.