Medical Imaging - A Look

inside
María Helguera, Ph.D.

ING-340

Medical Imaging
•  Allows physicians to see what had previously been unseeable: bones, organs,
tissues, a beating heart…

•  Allows physicians to check if a bone is broken, detect brain tumors early and
monitor patient’s treatment, watch a fetus grow, etc.

•  Has evolved from simple pictures (X-ray) to an interactive operation where

doctor’s can take measurements, observe behaviors within the body, and
simulate and model organs and functions.

1
Medical Imaging
•  Human body consists of tissues and organs made up of primarily water,
bone, gases.

•  Search the electromagnetic

spectrum for a region suitable for
imaging inside the human body.

•  Two parameters are decisive:

resolution and attenuation

•  Wavelength must be under 1.0 cm

for resolution considerations

•  Too much attenuation è

transmission measurements difficult.

•  Too little attenuation è not enough

accuracy

Medical Imaging

Macovski, 1983.

2
X rays

[Link]/medphys/learning

Bushberg, 1994

Characteristic X-ray Bremsstrahlung X-rays

[Link]

X rays

Penetrate

Photoelectric
interaction

Release kinetic energy

Scatter
Compton
interaction

2D projection of a 3D structure.

3
Computed tomography
• Tomogram is an image of a slice taken through a 3D volume

•  A CT scan is a procedure that uses x rays, radiation detectors, and

computers to produce images of planes through the body.

•  Mathematical basis for CT was developed by Radon in 1917. An image of

an unknown object can be produced if one has an infinite number of
projections through the object.

•  Practical applications started in 1960s. (Cormack). Was made possible by

the invention of the computer

•  First practical scanner was built in 1972 (Hounsfield). It was the first
modern imaging modality that made it possible to probe the inner depths of
the body, slice by slice.

•  Cormack and Hounsfield won the Nobel price in 1979.

Bettyann Holtzmann, Naked to the Bone. Rutgers Universiyt Press, 1996

Computed tomography

Macovski, 1983

Toshiba, 2012

4
Computed tomography

[Link]

PET

A positron is a subatomic particle that is

produced when radioactive material decays.
Proton -> Neutron + Neutrino + Positron

• The moment a positron encounters an

electron, they annihilate each other. This
collision creates a pair of high-energy gamma
rays that speed away in opposite directions.

•  In a PET scanner, the escaping gamma rays

strike a ring of detectors. When detectors on
opposite sides of the scanner fire within a
nanosecond, the event is recorded.

•  Image can be recreated by back projecting

the detected photons. [Link]

5
PET
The human body is not naturally radioactive. To create these particles,
the patient is injected with a mildly radioactive fluid about a half-hour
before the PET scan. Cancerous cells distinguish themselves by eating
more sugar-based nutrients than normal cells.

Popular Mechanics, 2002 [Link]

MRI
•  The magnetic resonant signal is the result of excitation of the individual
magnetized protons (hydrogen) by irradiation with radiofrequency (RF) energy
of a specific frequency.

•  A particle in a magnetic field can absorb a photon of a frequency dependent

on its gyromagnetic ratio. ν = γβ For hydrogen γ = 42.58MHz/T.

•  Energy absorption causes the displacement of magnetic moment from

equilibrium è excited system.

•  As system returns to equilibrium, MR signals are emitted in proportion to the

number of excited protons in the sample.

6
TI (ms) TR=1000 ms TR =2000 ms

MRI 50

•  MR images depend on many

variables including proton
100
density, T1 and T2 relaxation
effects, flow effects, diffusion
effects, and susceptibility effects.

•  Choosing the proper protocol 250

maximizes contrast between

tissues of interest.

500

Inversion
Recovery
750

TI (ms) TR=1000 ms TR=2000 ms

[Link]

Ultrasound

•  Sound is a wave, i.e. traveling variations of some quantity (pressure).

•  Sound involves mechanical motion in the medium through which it

travels.

•  Pressure variations cause particles of the medium to vibrate due to

increase and decrease of density.

7
Doppler Ultrasound
•  Color-Doppler imaging
presents 2D, cross-
sectional, real-time blood
flow or tissue motion
information along with
2D, cross-sectional,
gray-scale anatomic
imaging.

• 2D, real-time
presentations of flow
information allow to
locate regions of
abnormal flow.

Philips Medical Systems. Renal transplant.

•  Doctors prescribe the imaging procedure

•  Technicians perform the study

•  May not be all that knowledgeable about physics and

engineering of different imaging systems

•  Must know anatomy

[Link]

8
0 PET 255

0
Multimodal imaging

MRI
255
K. G. Baum, M. Helguera, and A. Krol, A New Application for Displaying and Fusing Multimodal Data Sets, 2007
Proceedings SPIE Symposium on Biomedical Optics, January 2007.

Study
The goals of the study are to investigate:
1) Understanding of each technique.
2) Training/learning time of each
technique.
3) Diagnostic efficiency of each
technique.
4) Technique’s effect on feature location,
size and shape.
5) Technique’s effect on perceived
values.
6) Difficulty using each technique.
7) Technique preference.

Rafferty, K., Baum, K.G., Schmidt, E., Krol, A., Helguera, M., “Multimodal Digital Techniques with Application to Breast
Imaging”, RIT Digital Media Library, 2007

9
Baum, K.G., Helguera, M., Schmidt, E., Rafferty, K., Krol, A., “Evaluation of Genetic Algortihm – Generated Multivarate
Color Tables for the Visualization of Multimodal Medical Fused Data Sets”, 2008 IEEE NSS-MIC

All together I1
Fig. 2. Original MRI image. Fig. 3. Original PET image. Fig. 4. Image created
fusion technique. Corre
__ T7

ghtes Mpace.
[Link] SreatiounC rl3L Fig. 7. Image created using HSL color L;.O5. imlage createci oy
Hue=PET, Saturation=CT, Lightness=MRI. space. Hue=PET, Saturation=constant, MRI volumes. PET i
HSL color space. Hue = PET, Saturation = CT, Lightness = MRI Lightness=MRI. Hue values are allowed using a fire color tabl
vary from cyan to green to yellow. grayscale col
3.2 Color Mixing 3.3 Use of Other Color Spaces

Color
Baum, K.G., Helguera, M., Hornak, J., mixing
Kerekes, is a technique
J.P., Unlu, that
M.Z., Feiglin, can
D.H., be A..
Krol, used to take for
Techniques number
anyFusion of A powerful, yet more complex
of Multimodal
Images: Application to Breast Imaging,
one channel
IEEE Proceedings (N)2521-2524,
of ICIP,
images and create 2006a fused RGB image. We technique for the creation of
perform channel mixing using Eq. 1. Here R, G, B represent the fused images involves the use of

red, green, and blue channels in the displayed image respectively, different color spaces [11, 12,
Si represents the intensity in the ith source image, Ri, Gi, Bi are the 13]. A few examples include F

weighting factors for the red channel, green channel, and blue CIE XYZ, CIE L*a*b*, HSV, ia
channel. They determine the contribution of source i to each of the and HSL. Each source grayscale 25i
output channels. image can be used as a channel in Fig. 5.
RI GI B1 the color space. The resulting create Fig
R2 G2 B2 color image can then be
converted to the RGB color space for displa
(SL S2 Sx (R G B) registered PET, CT, and MRI images 10 are ava
used as the lightness, CT as the saturation, a

(Fig. 6).
aiX
X) ~~~~~~~(1) This technique is perhaps better suited fo

sources, but none the less can be used for the f

Synthetic Imaging

Baum, K., MacNamara, K., Helguera, M., "Design of a multiple component geometric breast phantom", Medical Imaging,
Proceedings of SPIE, 6913, 69134H-1-12, 2008.

Digital Simulations

11
US Applications in Biology

f = 15 MHz, Δx = 50 mm

Image Registration

http://
[Link]
m/toc/tea/16/12

12
Artificial Tissues: USWF

Artificial Tissues: USWF

Garvin, K.A., Dalecki, D., YoussefHussien, M., Helguera, M., Hocking, D. Spatial Patterning of Endothelial Cells and Vascular
Network Formation Using Ultrasound Standing Wave Fields, J. Acoust. Soc. Am. 134 (2), 1483-1490, 2013

13
Artificial Tissues: USWF

Image Analysis

YousefHussien, M., Garvin, K., Dalecki, D., Saber, E., Helguera, M., “Three-dimensional Volume Analysis of
Vasculature in Engineered Tissues”, IS&T SPIE Electronic Imaging, 2013.

14
Tortuosity

GLCM

15
Quantitative Ultrasound: Cancer

Lavarello, R.J., Ridway, W.R., Sarwate,

S.S., Oelze, M.L., Characterization of
Thyroid Cancer in Mouse Models Using
High-frequency Quantitative Ultrasound
Techniques, Ultrasound in Med. & Biol.,
39(12), 2333-2341, 2013

Quantitative Ultrasound: Artificial

Tissues

Mercado, K.P., Helguera, M., Hocking, D.C.,

Dalecki, D. Estimating Cell Concentration in
Three-dimensional Engineered Tissues Using
High Frequency Quantitative Ultrasound, Annals
of Biomedical Engineering. DOI: 10.1007/
s10439-014-0994-8, 2014

16
Quantitative Ultrasound: Biofilms

Vaidya, K., Osgood, R., Ren, D., Pichichero, M.E., Helguera, M. Ultrasound Imaging and Characterization of Biofilms
Based on Wavelet De-noised Radiofrequency Data, Ultrasound in Med. & Biol., 40(3), 583-595, 2014

SMURF

Elegbe, E.C., McAleavey, S. A., Single Tracking

Location MethodsSuppress Speckle Noise in Shear
Wave Velocity Estimation, Ultrasonic Imaging, 35(2),
109-125, 2013.

17
SUMMARY

•  Create images of organs and functions

•  Serve doctors to diagnose/monitor disease

•  Determine underlying mechanisms
•  Therapy

•  Good quality images

•  Contrast ✓
•  Resolution ✓
•  Noise ✗
•  Artifacts ✗
•  Distortion ✗
•  Accuracy ✓

18