Transcript Presentation(in MS PowerPointFormat)

Medical Imaging:
Techniques and Applications
Amy Schnelle
Computer Science
University of Wisconsin-Platteville
Why Medical Imaging is an
Important Field of Study
•
•
•
•
Earlier Diagnosis
Easier Diagnosis
More Accurate Diagnosis
Less Invasive Diagnosis and
Treatments
• Greater Sharing of Knowledge
A Very Brief History of Medical
Imaging
• 1895 - Professor Roentgen accidentally discovers x-rays
while experimenting with Crookes tubes
• 1946 - Felix Bloch and Edward Purcell discover the
presence of magnetic resonance in solids and liquids
• 1960’s - The ultrasound is developed thanks to sonar
development during World War II
• 1972 – The computed tomography scan becomes a
reality due to breakthroughs in digital computers
Image Capturing Techniques
• Radiography
• Magnetic Resonance Imaging
• Computed Tomography
• Ultrasound
• Nuclear Imaging
Radiography
• Radiography is the process of creating an
image by passing x-rays through a patient
to a receptor.
• Traditional
-converting to digital
+scanning
+sampling
+conversion
• Digital
Radiography
© EarthOps.org
© Radiology Centennial, Inc
“So excited was the public that each newly radiographed
organ or system brought headlines. With everything about
the rays so novel, it is easy to understand the frequent
appearance of falsified images, such as this much-admired
"first radiograph of the human brain," in reality a pan of cat
intestines photographed by H.A. Falk in 1896.”
-Penn State University College of Medicine
Magnetic Resonance Imaging
• Magnetic Resonance Imaging (MRI) uses
magnetic energy and radio waves that
have varying sensitivity to the presence of
water within tissues.
• The magnet
• The magnetic Field
• Image Capture
– Acquisition
– Reconstruction
Magnetic Resonance Imaging
In this photograph, you can see a fully
loaded pallet jack that has been sucked
into the bore of an MRI system.
© How Stuff Works, Inc.
Computed Tomography
• Computed Tomography (CT) imaging
involves taking a series of x-ray images
and combining them to form a 2D or 3D
cross-sectional image.
• Components of the CT
• Image capture
• CT variables
Computed Tomography
© Imaginis
© Imaginis
Ultrasound
• Ultrasound uses high frequency sound
waves and their corresponding echoes to
create images of the internal structures of
patients.
• How the image formed
• Why ultrasound is popular
Ultrasound
© Radiology Info
© Photo Dynamic Imaging Limited
Nuclear Imaging
• Nuclear Imaging involves detecting
radiopharmaceuticals given to a patient for
diagnostic purposes and then creating
images of the collected data.
• Types
– PET: positron emission tomography
– SPECT: single photon emission computed
tomography
• Combined Systems
Nuclear Imaging
PET scans. Uptake of tracer in the lymph nodes involved with lymphoma
in the groin, both axilla, and neck (red areas). Image coutesy of
Dr. Jorge Carrasquillo, Nuclear Medicine Department, Clinical Center,
National Institutes of Health.
Image Capturing Review
• Radiography
– X-rays pass through patient to receptor
• Magnetic Resonance Imaging
– uses magnetic energy and radio waves that have varying
sensitivity to the presence of water within tissues
• Computed Tomography
– taking a series of x-ray images and combining them to form a 2D
or 3D cross-sectional image
• Ultrasound
– uses high frequency sound waves and their corresponding
echoes
• Nuclear Imaging
– detecting radiopharmaceuticals given to a patient for diagnostic
purposes and then creating images of the collected data
Techniques for Visual Rendering of
Captured Images
• Real-time Volume Rendering
• Finite-element modeling
Real-time Volume Rendering
• Volumetric rendering allows for the visualization
of the internal structure of objects, making it a
highly important contribution to medicine.
• Very computationally expensive
• “rendering a dataset of 2563 16-bit voxels at 30 Hz requires 32
MBytes of storage, a memory transfer rate of 1 GByte per second,
and approximately 5 billion instructions per second. This problem is
aggravated by the continuing trend towards larger datasets. Highresolution sampling devices, faster supercomputers, and more
accurate modeling techniques will make 10243 and larger datasets
the norm”
• Process of creating 2D images from a collection
of voxels.
Real-time Volume Rendering
• Rendering methods:
– Indirect
– Direct
• Image Order
• Object Order
• Hybrid
Real-time Volume Rendering
VolumePro
Hardware
acceleration
card
© TeraRecon Inc.
“VolumePro uses a distributed interleaved memory, several parallel processing pipelines, and an
Innovative parallel dataflow scheme that requires no global communication, except at the pixel level.”
Segmented volume
rendering of Cardiac
CTA
Plate transfixing tibial
metaphysis. Screw transfixing
tibial plateau. Proximal tibial
diaphysis.
Colon Polyp
Deformable Models with Finiteelement Modeling
• Needed to realistically model moving tissues
• Allows for a quick and accurate representation of
the volume as well as the surface of soft tissues
• Model is made up of layers of nodes
• Each Node has properties that determine its
deformability
• The nodes are overlaid by a mesh that models
the surface of a specific layer
Specialized Applications of Medical
Imaging Technology
• Surgery Simulation
• Computer-aided Diagnosis
• Drug Development
Surgery Simulation
• Used to train surgeons
• Need to have a good UI, preferably force
feedback
• Very computationally expensive
-Precompute
-Limit deformable area
Computer-aided Diagnosis
• Has been shown to greatly improve
diagnostic accuracy of doctors
• Methods
– Neural Networks
– Expert Systems
Neural Networks
• Modeled after human brain
• Contains a set of Nodes (artificial neurons)
• Each node has a set of inputs and an
output
• Nodes fire output depending on the inputs
• The connections between the nodes have
weights, and these weights can be
adjusted to receive a known output from a
known input.
Expert Systems
• Emulates the decision making of a human
expert
• Composed of:
If It’s raining and It’s cold, then Amy is not happy
antecedent
consequent
– Set of Rules
– Inference Engine
• Forward and backward chaining
Drug Development
• Spatial properties of organs and organ
systems
• Testing with computerized models
• Clearer understanding of drug effects
• Easier to share findings with other
researchers
References
[1] Balázs, C. (2001, May). Interactive Volume-Rendering Techniques for Medical Data Visualization. Institut f¨ur Computergraphik und Algorithmen.
[2] Bro-Nielsen, M. (1997, October 28). Finite Element Modeling in Surgery Simulation. HT Medical, Inc.,
[3] CAD software helps distinguish benign, malignant nodules seen on
CT scans. (2004, Dec 21). Cancer Weekly, 115.
[4] Dellingette, H., Ayache, N. (2005). Hepatic Surgery Simulation
Communications of the ACM, 48 (2) 31-36.
[5] Hadjiyski, L. Ph.D., Sahiner, B. Ph.D., Chan, H. Ph.D., Bogot, N. M.D., Cascade, P. M.D., Kazerooni, E. M.D. (2004, November 29).
New computer-aided diagnosis (CAD) methods for cancer. News-Medical.Net. Retrieved Feb. 22, 2005, from
http://www.news-medical.net/?id=6495
[6] Hibbs, C., Vance, G. (2005). The Effects of Combined Imaging Technology on Healthcare Planning and Design. Continuum Solutions Consulting.
Retrieved Feb. 22, 2005, from http://www.aia.org/aah_a_jrnl_0401_article5&grandCh=yes
[7] Huang C.R., Sheu B.S., Chung P.C., Yang H.B. (2004). Computerized diagnosis of Helicobacter pylori infection and associated gastric inflammation from
endoscopic images by refined feature selection using a neural network. Retrieved Feb. 22, 2005, from
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&listuids=15243882&dopt=Abstract.
[8] Imaginis Corporation. (2005). Computed Tomography Imaging. Retrieved Feb. 22, 2005, from http://imaginis.com/ct-scan/
[9] Imaginis Corporation. (2005). History of Medical Diagnosis and Diagnostic Imaging. Retrieved Feb. 22, 2005, from http://imaginis.com/faq/history.asp
[10] Kaufman, A., Brady, M., Lorensen, B., Kitson, F., Pfister, H. (1998). Why is Real-Time Volume Rendering No Longer a Year Away?.
State University of New York at Stony Brook. Retrieved Feb. 22, 2005, from
http://64.233.167.104/search?q=cache:1z8yxW-Om8cJ:csdl.computer.org/comp/proceedings/vis/1998/9176/00/91760497.pdf
[11] Kaufman, A., Lakare, S., Kreeger, K., Bitter, I. (2005). Virtual Colonoscopy.
Communications of the ACM, 48 (2) 37-41
[12] Lin, Q., (2003, April). Enhancement, Extraction, and Visualization of 3D Volume Data. Department of Electrical Engineering,
Link¨oping University, SE-581 83 Link¨oping, Sweden
[13] McRobbie, D., Moore, E., Graves, M., Prince, M. (2003). MRI: from Picture to Proton. Cambridge University Press 2003
[14] National Cancer Institute. (2005). Nuclear Imaging (PET and SPECT). Retrieved Feb. 22, 2005, from
http://imaging.cancer.gov/imaginginformation/cancerimaging/page5
[15] Nondestructive Testing Resource Center. (2005). Computed Tomography. Retrieved Feb. 22, 2005, from
http://www.ndted.org/EducationResources/CommunityCollege/Radiography/Adv ancedTechniques/computedtomography.htm
[16] Park, K., Montillo, A., Metaxas, D., Axel, L. (2005). Volumetric Heart Modeling and Analysis. Communications of the ACM, 48 (2) 43-48.
[17] Radiological Society of North America, Inc. (RSNA). (2005). Ultrasound – General. Retrieved Feb. 22, 2005, from
http://www.radiologyinfo.org/content/ultrasound-general.htm
[18] Dr. Saini, S., Dr. Digumarthy, S. R., Dr. Sobrino, H., Dame, M. (2002). Drug Development and Medical Imaging Technology. Pharamatech.
[19] Sprawls, P. Jr. (1995). Physical Principles of Medical Imaging. Perry Sprawls and Associates
[20] TeraRecon, Inc. (2004). VolumePro™ - Volume Rendering Hardware Solutions. Retrieved Feb. 22, 2005, from
http://www.terarecon.com/products/volumepro_prod.html
[21] Vilanova Bartroli, A. (2001, September). Visualization Techniques for Virtual Endoscopy.
.
[22] Webb, S. (1998). The Physics of Medical Imaging. IOP Publishing
Any Questions?