Cee is not installed! Fortune Cookies not available!
01 Apr 2005
Computers Rant
Subscribe to Digest
Unsubscribe from Digest
Tue, 12/04/2005
Zigbee Design Contest

In January of this year I ran into a wireless design contest being put on by Freescale Semiconductor (part of Motorola) which involved coming up with novel applications for their Zigbee wireless protocol. It was one of these things that I couldn't resist and shortly after sending off my proposed project, I had a design kit consisting of three wireless enabled boards, each with a HCS08 microprocessor with 56 Kb of flash RAM and 4 Kb of RAM.

The development system that came with the package was Metroworks C compiler for HCS08 and it is a very easy to use system and in a few hours I was modifying the demo C code and flashing the parts with my modifications.

Being a doctor, I don't have the luxury of spending days on this project like I would have done in my student days, and so I stuck strictly to the proposal I'd submitted even though it was a fairly basic use of the devices.

I have a strong interest in sleep apnea and I have had over 100 patients of mine undergo home pulse oximetry to determine whether or not they are likely to have sleep apnea. A pulse oximeter is nice, but I wanted more data and my project proposed to augment the pulse oximetry with a movement sensor (3 axis accelerometer) and a pressure sensor to do breath by breath timing. Both of these functions were implemented in wireless boards which are battery powered and easily fit into a modified tape cassette case. The board with the 3 axis accelerometer is attached to the patients leg and the pressure sensor is attached to a modified face mask which makes for easy sensing of breathing related pressure changes. Originally I had planned on attaching the unit with pressure sensor to the head with some webbing, but decided that I would instead have it strapped to the patients chest so that the 3 axis accelerometer could detect vigorous chest movements that occur when people are attempting to breath when their airway is blocked (thus allowing one to directly diagnose obstructive sleep apnea). Both wireless units transmit simultaneously; finding that just letting them transmit data packets at a rate of ~30/second resulted in very few collisions greatly simplified the project and allowed me to avoid getting sidetracked in programming complicated wireless protocols. The third board from the kit picks up the data packets from the two remote units, checks them for validity and transmits valid packets to a bedside laptop via a serial link. I also have serial output from the pulse oximeter going into the same computer and all serial data is timestamped and streamed to disk.

This weekend I decided it was time to get this project done and, somewhat to my surprise, it worked. The biggest headache was interfacing the MPXM201GST1 pressure sensor to an A/D line. The full scale output of this unit is only 25 mV and for some reason my 741 based amplifier circuits didn't work. In desperation, I used a dual FET input op amp (TL082) to buffer the pressure sensor output and applied the output ot a 1000X non-inverting amplifier using a 741 and it finally worked. When the design contest is over I'll post complete schematics and code of the project on my web site, but until 27/4/2005 they will remain private.

I've tested the system on myself and it works, but I now have to write code to take the data files I've collected and display them in a useful format.

Having wireless sensors is very liberating as I can walk around my apartment wearing the accelerometer and all movements are picked up by the reciever unit. My goal with this system is to have a completely ambulatory physiologic monitoring system which will store data on FlashRAM memory cards, and for this purpose I've purchased an ALFAT board which accepts serial data and streams it to memory cards. As memory cards now have capacities up to 4 Gb, one has essentially unlimited storage since even 24 hours of multiple signals will only occupy a few hundred megabytes. Further thoughts on this system will be left for the future as I still have my initial system to finish debugging.


by Boris Gimbarzevsky  Comments

by Boris Gimbarzevsky  Comments

Banning Bextra - more stupidity from Ottawa

A few days ago the oxymoronically named "Health Protection Branch" decided that Valdecoxib (Bextra) was too unsafe to be used by Canadians and ordered its removal from the market. No matter that the older COX-1 agents are far more dangerous to use in the elderly and I guess in Ottawa it is far more acceptable to die of GI bleeding from COX-1 NSAID's than face the ever so slight possibility of getting a rash from Valdecoxib.

I probably shouldn't complain; just like natural disasters are very good for the economy, every such asinine move on the part of statist bureaucrats results in an increase in office visits by patients either seeking a new analgesic or worried that some other drug they are on might be unsafe. It won't be long before all of the people who were previously on Viox that I switched to Valdecoxib will be in my office and I'll see them a number of times during a short time period while we try to find something that works for their pain.

Valdecoxib is a drug that has not been associated with a single adverse event of any kind in my practice. Celecoxib, now the main COX-2 NSAID on the market has been associated with far more skin rashes in my experience, but all patients have to do is to stop taking it and try something else. The thing about NSAIDs is that what works for one person will not work for another. I've never heard an explanation for why this should be the case, but I have patients who have tried up to 10 different NSAIDs before they find one that works for them. Obviously if Valdecoxib or Viox were the only drugs that worked for someone, then they are out of luck in Canada.

The morons in Ottawa seem to believe that they should be the sole arbiters of what risks are acceptable to people. People's perception of risk is very skewed; yesterday I had a patient in my office who was ready to stop taking Bextra because of its removal in Canada. He was asking about whether he should undergo a colonoscopy to deal with a matter unrelated to his arthritis pain and I told him that the risk of a bowel perforation was about 1 in 2000 with the gastroenterologist I refer to. His response was: "Oh, that's a very small risk". His risk of developing rash from Bextra, or any other serious Bextra side effect for that matter, is << 1 in 2000.

by Boris Gimbarzevsky  Comments

Update of web site

I've always liked to have complete control of my web site and that means hosting it on one of my computers. If you're into computers and networks, it is a lot of fun to setup your personal networks that are connected to the internet but relatively isolated. Loki, the machine that hosts the web site is a 900 MHz Athlon machine that now is more or less limited to analyzing seti@home data and hosting the public face of drgimbarzevsky.com. If you poke around the web site you'll notice links to other machines through ports other than the standard HTTP 80, but eventually Loki will be home to the complete web site. I used to have a machine in my office for this purpose, but Telus decided they would no longer allow outgoing port 80 on their ADSL connections and hence drgimbarzevsky.com was moved to a Rogers cable connection. Google will index pages that are available through ports other than 80, but most people use the standard web browser interface which uses port 80 and hence the relocation of drgimbarzevsky.com.

This is the first update of this web site in 3 years as maintaining this site was more like work rather than something new and exciting like it was in 2000-2001. I guess the blog makes things interesting again, but will have to wait and see how long this lasts.


by Boris Gimbarzevsky  Comments

Dr. Gimbarzevsky blog

Don't have time to update this website as often as I like, but have decided that a blog format is the best way of putting information on this web site. Blogs are useful means of making ones thoughts public and I expect that this blog will deal with stuff that interests me on a day to day basis.

I need a place to keep track of all of my various ideas on things and a blog is ideal. The reason to make it public is that my ideas already are semi-public in that I communicate them with patients when I see them. As I happen to be doing almost entirely clinical work, I have no time to publish papers on my clinical experiments, but the internet is a great way of getting a wider distribution for various clinical observations that I make on a daily basis.

An obligatory disclaimer should be made here: anyone who uses this information does so entirely at their own risk. While I am a physician, the only setting in which my observations have any validity is if you are sitting in my office and I am rendering a diagnosis or recommendations for treatment after having first taken a detailed clinical history and having performed a focused clinical examination. Biologic systems are very diverse and assuming that something I say applies to you just because you happen to like what you've read on the internet is the height of folly. If you think this information is applicable to you, either take it to your physician to see what they think of it or do extensive cross checking from other sources.

The safest thing for me to do would be to say nothing, but I see so much crap on the internet that passes as medical information that I feel it necessary to inject my viewpoints into the mix. I'll attempt to back up any statements of mine with appropriate references, but unfortunately I won't be able to post full text of many of the references. Due to licensing conditions for many of the online medical papers that I have access to through the BC College of Physicians and Surgeons, I am allowed to download copies for personal use, but not post the full text of these papers on my web site. Fair use doctrine does, however, allow me to paraphrase the findings of the papers which is not the same thing as having the original paper, but close enough for most people.

As this blog is just starting, no organizational plan is in place and what will be posted is what I'm dealing with on a day to day basis although it is unlikely I'll have the time to update it on a daily or even weekly basis.


by Boris Gimbarzevsky  Comments

Posted at:Tue, 12/04/2005 21:45:19