Design Article
Developing the world’s first real-time 3D OCT medical imaging system
Dr. Kohji Ohbayashi, Graduate School of Medical Science, Kitasato University, Japan and team members
4/25/2011 6:55 AM EDT
This team created the world’s first real-time 3D OCT imaging system
by using optical coherence tomography (OCT) and a 320-channel data
acquisition system combining NI FlexRIO field-programmable gate array
(FPGA) hardware and GPU processing, .
OCT is a noninvasive imaging technique that provides subsurface, cross-sectional images of materials. Interest in OCT technology continues to grow because it provides much greater resolution than other imaging techniques such as magnetic resonance imaging (MRI) or positron emission tomography (PET).
Additionally, the method does not require much preparation and is extremely safe for the patient because it uses low-level laser outputs and does not require ionizing radiation. OCT uses a low-power light source and the corresponding light reflections to create images – a method similar to ultrasound, but it measures light instead of sound. When the light beam is projected into a sample, much of the light is scattered, but a small amount reflects as a collimated beam, which can be detected and used to create an image.
OCT is a promising diagnostic tool in many medical fields. In OCT applications, imaging speed is crucial for fast inspection and achieving good image quality without motion artifacts. To inspect the human eye, which can be held relatively still using a chin rest, we must use a fast A-scan rate to eliminate all motion artifacts. However, in endoscopic OCT, such as the digestive and respiratory systems, the tissue being imaged cannot be fixed in place, so we must use ultrahigh-speed OCT methods to eliminate motion artifacts. Moreover, in noninvasive real-time optical biopsy, the imaging speed must be fast enough to display the 3D image in real time for immediate diagnosis, just like a conventional endoscope.
A few previous methods have been proposed for ultrahigh-speed OCT, but none have succeeded in real-time display of 3D OCT movies. This article describes how a team led by Dr. Kohji Ohbayashi, Kitasato University, Graduate School of Medical Science, Japan, accomplished this impressive goal.
To read it, click here.
|
|
|
|
|
pip_010
4/27/2011 6:11 AM EDT
well mri is 3D volume, not surface ! also 7T-mri can provide amazing detail with resolution : fraction of the millimeter in voxel size. anatomical scans do not require toxic fluids for contrast etc. and it is noninvasive and safe. bad sides : digestive and respiratory systems are notoriously hard in MRI, though SWIFT seems to mitigate the issue. real-time is challenging due to immense post-processing, though using GPU and distributed-cluster should be a viable option for MRI too.
Sign in to Reply
SpeedEvil
5/2/2011 1:29 PM EDT
MRI is great.
MRI is also hideously expensive.
OCT is potentially a lot cheaper than MRI, with faster scanning, but it does have severe limitations as to the depth of objects it can sense.
It's another tool in the arsenal, not a magic bullet.
Sign in to Reply
iniewski
5/2/2011 6:35 PM EDT
Agreed with SpeedEvil: it is an exciting but largely complementary technology to MRI or PET...I am editing a book on medical imaging for Wiley, looking for someone to write a chapter on OCT, if interested pls email: kris.iniewski@gmail.com, Kris
Sign in to Reply
agk
6/21/2011 3:50 AM EDT
Hi iniewski i am interested to write a chapter on this. I will refer to various articles on this topic and prepare for you. Will it be allright
Sign in to Reply
iniewski
6/21/2011 10:43 AM EDT
Hi agk, great, pls email me at kris.iniewski@gmail.com
Sign in to Reply
solarpower44
5/5/2011 3:54 PM EDT
This is not true.
This is not the first real-time OCT.
Too bad they don't cite other works.
for example:
Graphics Processing Unit Accelerated Non-uniform fast Fourier transform for ultrahigh-speed, real-time Fourier-domain OCT,” Opt. Exp., vol. 18, Iss. 22, pp. 23472-23487, Oct., 2010
Real-time 4D signal processing and visualization using graphics processing unit on a regular nonlinear-k Fourier-domain OCT system,” Optics Express, Vol. 18, Issue 11, pp. 11772-11784, 2010
Sign in to Reply
particle60
5/19/2011 3:18 AM EDT
I checked the works cited by solarpower44 because I got interested in. Although they actually showed some real-time OCT, I think it is
insufficient for MEDICAL IMAGING both in the image size and the image penetration depth. It seems this work's designs fit a bit better to target medical use as the title shows.
Sign in to Reply
Benchuchu
6/20/2011 11:08 PM EDT
This work is surely no better than the examples given by solarpower44, because of the physical/optical feature of OCT imaging itself.
For OCT imaging, which is limited by the scattering of tissue itself, the imaging penetration depth is up to 1~2mm for normal biological tissue and may be up to 10mm for ophthalmological structures like cornea.
The imaging size mentioned by particle60, which called lateral scanning range in OCT, is just determined by the number of A-scans in one volume, which is not related to the imaging processing part.
Overall, OCT is kind of novel and upcoming imaging modality, which high resolution but shallower imaging depth compared to US or CT or MRI. I think the most powerful field for OCT is the ophthalmology applications, since eye itself is the most transparent part within the body.
Sign in to Reply
iniewski
6/21/2011 1:04 AM EDT
To @Benchuchu, interesting comments. I am editing a book on medical imaging for Wiley and planning to add OCT to standard modalities (MRI, US, CT etc). Would you be interested in contributing a chapter? kris.iniewski@gmail.com
Sign in to Reply
iniewski
5/19/2011 9:41 AM EDT
To @particle60: what is the image penetration depth? Kris
Sign in to Reply
particle60
5/23/2011 7:13 PM EDT
Hi, The image penetration depth is the length of actually acquired image
in the direction of depth from the surface.
Sign in to Reply
Benchuchu
6/20/2011 11:00 PM EDT
I strongly agree with solarpower44. The FIRST realtime 3D (or 4D)OCT system was developed by the group at Johns Hopkins using GPU for both OCT image reconstruction (A-scan processing)and volume rendering. While this work uses FPGA for reconstruction and GPU volume rendering. The FPGA method has also been tried before [1,2]. Personally I prefer the GPU method with CUDA since it's the most cost effective approach. The FPGA module should also realize the function but it would cost much more money than a regular GPU.
[1] T. E. Ustun, N. V. Iftimia, R. D. Ferguson, and D. X. Hammer, “Real-time processing for Fourier domain optical coherence tomography using a field programmable gate array,” Review of Scientific Instruments 79, 114301 (2008).
[2] A. E. Desjardins, B. J. Vakoc, M. J. Suter, S. H. Yun, G. J. Tearney, B. E. Bouma, “Real-time FPGA processing for high-speed optical frequency domain imaging,” IEEE Transactions on Medical Imaging 28, 1468-1472 (2009).
Sign in to Reply
agk
6/21/2011 4:00 AM EDT
3D OCT system by Dr Kohji works at 1310 nm center wave length and 250 narrow spectral band.They achieved OCT imaging of 60 million axial scans per second and demonstrated a 16 kHz frame rate with 1,400 A-lines per frame, a 3 mm depth range, and 23 µm resolution. With 14 bit resolution 50Mega samples per second a nearly 1GB data rate the processing is done by FPGA's with DSP. Great work and results. With this no more painfull biopsys
Sign in to Reply
AsicProf
10/1/2011 10:22 PM EDT
Hi AGK,
You seems to be knowledgeable in area of medical imaging. I want to know your opinion on medical imaging company , who claims to generate real time 3d images.
Recently I came across a small company imaging3.com, who claims to create real time 3d image. They claim to construct 3d volumetric image of hand / skull in in real time.
Their technology is not using FBP reconstruction algorithm. They basically use fluoroscopy to measure the distance reading from the surface of the targeted volume , by rotating X source 360 degree around the object.. These distance measurement eventually used to create 3d base map (volume envelope). Now once 3d base map is created, S/W just paste 2d picture on the 3d base map. This process gives a very fast construction of 3d volume image.
The gantry and cart computer are synchronized on 3d base map ....ie. when user rotates the image on the computer screen ....a message is sent to gantry computer to re-scan new coordinate ....On the opposite side ...if target moves inside the gantry ....gantry computer re-scans the new view and send a message and new image to cart computer for update ...
Below is the link of their technology patent.
http://www.google.com/patents?id=piISAAAAEBAJ&printsec=abstract&zoom=4&source=gbs_overview_r&cad=0#v=onepage&q&f=false
I am wondering whether this technology is really feasible, given that big medial imaging companies like GE, Toshiba, Philips, and Siemens could not think of this technique.
I will really appreciate if you can give me your comment on this technology.
AsicProf
Sign in to Reply
iniewski
10/2/2011 1:49 PM EDT
Very interesting technology, has anyone used Imaging3 products? Kris
Sign in to Reply
AsicProf
10/3/2011 12:36 AM EDT
Hi Kris,
They are waiting for 510 pre-market FDA approval before they can market the product.
What you think about market potential of this technology , if it works as advertised.
Sign in to Reply
iniewski
10/3/2011 12:11 PM EDT
Hi AsicProf: Yes, it looks very promising...I will be happy to invite them to present at CMOSET conference in Vancouver in 2012 to present (I am the chair, www.cmoset.com), Kris
Sign in to Reply
AsicProf
10/4/2011 12:46 AM EDT
Hi Kris,
They will be attending RSNA in Chicago in November.
Hopefully by 2012 , they will get FDA approval.
You can contact company CEO at dean@imaging3.com
AsicProf
Sign in to Reply