Home Articles Downloads Forum Products Services EBME Expo Contact
Previous Thread
Next Thread
Print Thread
Rate Thread
Joined: Apr 2005
Posts: 27
Edwin Offline OP
Dreamer
OP Offline
Dreamer
Joined: Apr 2005
Posts: 27
Our department looks after a few Dixwell UV cabinets and over the past few years we have calibrated these units in different ways.

At the moment we attach a contraption to the cage which put the UV sensor of our GigaHertz X96 UV-meter at 20cm from the tubes at waist height. We measure at all 4 sides and calculate a mean figure. We were advised to take 15% - 20% of this figure to compensate for the fact that there is no patient in the cabinet. But we found that the cabinet readings also drop when someone goes inside. In fact the cabinet's UVB output display shows a greater drop in measured output in mW/cm2 than our X96 UV meter does (UVB 311 nm).

As I am trying to establish an accurate calibration factor, I wondered if anybody out there have already done the research to work out this factor. Or maybe it has to be done for each individual cabinet, a different conversion factor for UVA and UVB?

I would also like to know what other methods are being used. Someone recently suggested using a Dolly smilewink

A side issue is that NPL has implemented a new scale onto the spectroradiometer calibration which is used to calibrate our X96 UV meter. This has resulted in a significant rise in displayed output, affecting the recalibrated cabinet's dosimetry.

In a nutshell my question is: How do you calibrate your UV cabinets?

Many thanks, Edwin

Joined: Jul 2004
Posts: 20
Novice
Offline
Novice
Joined: Jul 2004
Posts: 20
If you start looking into the calibration of whole-body UV cabins, you rapidly open a huge can of worms.

The simple answer to your question "maybe it has to be done for each individual cabinet, a different conversion factor for UVA and UVB", is that YES, you need to measure each wavelength seperately for each cabin, and if you are issuing factors (rather than tweaking the machine), you need to ensure that the operator is applying the correct factors to the correct treatment unit.

The biggest problem is your phrase "Define an accurate calibration factor". We accept the cabin factors as "Arbitry units", and keep our calibration measurements to ourselves. If a transfer record is required, we can generate transfer records with absolute values rapidly.

If your center does not use prescriptive dosimetry (i.e. the Dr says - "Give this person 40J over 16 fractions"), but instead the operator increases the dose in line with the patients reactions to the treatment, then I do not believe absolute calibration is essential.

Waldmann and Cosmedico (the 2 big players at the moment) lock the calibration factors of the sensors into the comuter memory at the factory and you cannot change that. I have just completed the factory training course for service engineers at Cosmedico, and they won't even let their engineers change it!

As to dosimetry & calibration - well, where would you like to start. Some folk belive it is best to measure at 20cm out, some at 2cm out, some use a full size body "simulator" made from foam ... there really are as many different techniques as there are centers providing the service. I don't want to teach you to suck eggs, but technique and repeatability of technique is really important. If you cannot accurately repeat your measurements, then there is no point in doing them in the first place. Also, always document what you do - we are in a very litigatious society and given you are directly affecting a patients treatment, documentation will save you neck in court. I was always told - "It's the work that gets you paid, it's the paperwork that keeps you out of jail".

Anyhow, to try to answer your question. I would always consider it prudent to consider calibrations on a like for like basis. i.e. if the cabin's dosimetry is calibrated with a body in-situ, then that is how you should measure the UV intensity in the cabin. It is true that a body present in the cabin will attenuate the readings on the sensors, but then again, most sensors only look at one (or at most 2) banks of lamps.

Just for illustration, the Cosmedico caibration procedure, calls for the engineer to enter the cabin, wearing a white Tyvek jumpsuit (with hood - great fun in July, when it was 39ÂșC in the factory and then being sat in the cabin with all the lamps on for 3 minutes!!), and to measure the output along the mid-line of each bank of lamps, 2cm out from the clear, protective guard. The maximum value obtained (irrespective of lamp bank) is used. Only the factory specified dosemeter is approved for this. Personally, I measure the output using an IL1400A (with appropriate sensor) held on a tripod, at midline, mid height, 20cm out from each bank with a lab-coat hanging on the tripod to provide attenuation. I measure both intensity and calculated dose for a given exposure. These values are compared with the set/displayed values, and provided the % error does not change (significantly), the unit is deemed fit for use. At present. this has the advantage of permitting the patient to swap from one machine to another without correction factors - the factory calibration is still valid (and is still within acceptable error)

I would be careful in issuing new calibrations/correction factors on a regular basis - is the cabin really changing that much?

I believe it is really important that should you decide to change your calibration technique, you cross check with your old technique to ensure silly mistakes haven't been made (I nearly let one by some time ago when doing just that, but a double check caught it!).

There is a (what I hope to be good) meeting in December at Gloucester, targeted at those providing a QA/Service to W/B UV users. I shall be there and I hope to gain a lot from it.

Feel free to contact me at david.richardson@essexrivers.nhs.uk if you want any more info, or failing that, catch up with me at Gloucester in December.

Joined: Jul 2003
Posts: 108
Savant
Offline
Savant
Joined: Jul 2003
Posts: 108
One suggestion for a further contact on this subject, that may be of some help.
Brian Diffey, of Regional Med Physics, i beleive based in the Newcastle(upon tyne)hospitals (sorry i don't have a direct contact these days).

We used to use Brian for advice on our cabinets and number of years ago and i think there is still a regional service that carries out QA checks. You should be able to track the service down fairly easily via internet.

Could i ask, is the Gloucester meeting open to other trusts ? or does anyone have more details?
If it is open, then i would enquire as to sending someone down for updates etc.
Thanks
Steve

Joined: Sep 2006
Posts: 2
Newbie
Offline
Newbie
Joined: Sep 2006
Posts: 2
The Gloucester meeting is open to all interested parties I believe. Further details are available from Dr. Chris Edwards on 01633 238560 or email him on chris.edwards2@gwent.wales.nhs.uk
There is a PDF flier available which either myself or Chris could email

Joined: Apr 2005
Posts: 27
Edwin Offline OP
Dreamer
OP Offline
Dreamer
Joined: Apr 2005
Posts: 27
Thank you all very much for your input. I have booked myself a place at the Gloucester Dosimetry Course on 7 December. Looking forward to meeting you there!

In the meantime I have been in touch with Cosmedico in France who look after the Dixwell cabinets. They explained that taking the readings at 20cm is ok and that a person would go inside to take these measurements. Once outside the average plus 20% (to allow for the fact that the cabinet is now empty and more light will fall on the sensors) is entered into the calculator panel.

Cosmedico France has also confirmed that as long as the cabinets display is about the same as what a person inside is measuring at 20cm from the tubes, all is ok.

Has anybody got any thoughts on the change in scales used for calibration by NPL?

Many thanks,

Edwin.


Moderated by  DaveC in Oz, RoJo 

Link Copied to Clipboard
Who's Online Now
1 members (daisizhou), 435 guests, and 401 robots.
Key: Admin, Global Mod, Mod
Newest Members
MDevins, EmsR16, Fergusben1, Guilherme Gomes, Salvatore
10,173 Registered Users
Forum Statistics
Forums25
Topics11,055
Posts73,707
Members10,173
Most Online5,980
Jan 29th, 2020
Powered by UBB.threads™ PHP Forum Software 7.7.5