HomeMade CT [update_first _3D_pictures] / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

HomeMade CT [update_first _3D_pictures]

previous 1 2 3 4 next

Move Thread

LAN_403

Linas

Sat Jun 25 2011, 09:37PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

Grenadier wrote ...

A $7,000 sensor!

ok, i am out :D
Still, must be easy way to do that. just need fat CCD ( but i believe CCD is capable directly convert X ray to picture ? or it's just my imagination? )

Ash Small

Sat Jun 25 2011, 09:44PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

An array of scintillators and photomultipliers were used in the early scanners, around ten, I think.

There was a bloke on another site who was dismantling one recently and was offering the parts for sale.

I'll try and find a link if you are interested.

(I would have been interested in the magnets if I'd had somewhere to put them)

hboy007

Sat Jun 25 2011, 10:18PM

Registered Member #1667 Joined: Sat Aug 30 2008, 09:57PM
Location:
Posts: 374

if you have to ask for the price it means you can't afford it... or just ask nicely and get a few meters for free...

Wolfram

Sun Jun 26 2011, 04:26AM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

Well, you have a few alternatives, but not all of them are good.

You could make an axial CT (CAT), those only require a line sensor or a single scintillator, but they are very complex mechanically, and the single-sensor kind is extremely slow and even more mechanically complicated.

There is also digital tomosynthesis, which is quite easy to do, but it's less interesting as it doesn't give you a full 3D view of the scanned object.

Cone beam CT (like in the video I sent you earlier) is the best kind for hobby use, but requires a 2D sensor array, as you know. The resolution of the x-rays does not have to be that high, I've seen good results from input images only a couple of hundred pixels on each side, but you need a lot of images for good results. So a fluorography setup (camera and intensifying screen) should work fine. If you want short exposures, you need a moderately fast intensifying screen, but not so fast that it reduces the resolution too much. I guess "Regular" or "Fine" would work well, depending on the size of the object you want to scan. You also need a good camera with low noise and fast optics to get away with a short exposure time. A modern DSLR with a 50mm f/1.4 would probably be ideal, but cheaper stuff would work too if you increase the exposure time.

Reconstruction (making a DICOM stack from the projections) is the difficult bit, luckily there's open source software for cone beam reconstruction

. It's a bit tedious to use, and there are a lot of cryptic settings, but it does the job just fine, and alternatives seem non-existent unless you have a lot of money to spend.

You also need a viewer to interpret the DICOM stack and for doing volume rendering, but this is the easy bit, as there are tens of free programs that do this.

If you're serious about trying this, I can sugges the book "The Essential Physics of Medical Imaging" which contains a lot of good info on the subject.

Linas

Tue Jun 28 2011, 10:52AM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

ok, i was doing some research, and find out that with camera i will be able get some normal results (compared 70k$ detector pictures with mine)
next question, how many pictures i need to take ? ( because i can find just 1.8^/step motor that is around 200 photos)

Wolfram

Tue Jun 28 2011, 09:07PM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

I think 200 would work fine. If I remember correctly, the sample set for OSCaR is around 320 pictures, so 200 would probably work OK too. Do you have a proper camera or are you planning on buying one for this project?

Linas

Tue Jun 28 2011, 09:16PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

my father is in photography, so i have a right stuff for screen capturing.
after i saw photos of real cone beam detector, i feel quite confident that i will get something

here is remote camera for earlier photos, but for my experiment i will use DMC-FZ20,
does X ray can hurt cameras?

ok, this is quality i am getting with low distance running 300W and 2s exp time

any ideas is enough ?

Linas

Fri Jul 22 2011, 04:05PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

CT scanner driver , now just need upload program and close the door for 30min

testtest

Fri Jul 22 2011, 10:09PM

Registered Member #3271 Joined: Mon Oct 04 2010, 02:29AM
Location: Canada
Posts: 159

Linas wrote ...

CT scanner driver , now just need upload program and close the door for 30min

Care to share more details about the device in your photo. Is it a uP based stepper motor driver or does it do more?

Linas

Sat Jul 23 2011, 04:45PM

Registered Member #1143 Joined: Sun Nov 25 2007, 04:55PM
Location: Vilnius, Lithuania
Posts: 721

it's MSP430F2132 based motor driver with Bipolar-unipolar motor support, shutter driver for Lumix camera, X ray generator bridge power interrupter,Geiger counter, and lcd with some user interface for delay, motor type, step count , delay for shutter setup.
Still writing program, but since is first time in microcontrollers, it will take some time to program it.
for now i don't know how to print integer
like:
int j;
j=3;
PrintToLCD(j);
For now just can print string

previous 1 2 3 4 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob