MPHY-501
Bushberg Chapter 5
Medical Imaging Informatics
During the next two years you will learn the basics of Diagnostic Imaging. Ultimately after this period you will enroll in a Medical Physics Residency Program and finally take the Board examination that, once you pass, will certify you as a medical physicist.
1
Digits, Base 10, Binary
Base 10
The number: 3406 = (3 X 103) + (4 X 102) + (0 X 101) + (6 X 100)
Base 2 (Binary)
The number: 1011 = (1 X 23) + (0 X 22) + (1 X 21) + (1 X 20)
Conversion Binary to Base 10
1011 (Binary) = (1 X 23) + (0 X 22) + (1 X 21) + (1 X 20) = 8+0+2+1 = 11 (Base 10)
We write the binary number 1011 as 10112 and the Base 10 number 11 as 1110
Analog Versus Digital form in Imaging
Image Capture is in analog form
Today analogue information is converted to digital immediately for storage, manipulation and display.
You may find some rare fully analogue devices including Fluoroscopy, Nuclear Medicine and Ultrasound.
Transistor Transistor Logic (TTL) Pulse:
Digital electronics: A TTL pulse is 5 volts considered High
Otherwise 0 volts Considered Low
The 0 volts or 5 volts constitute the Binary logic of digital electronics
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
16 bit chip
Pins and pin numbers
Pins and pin numbers
16-Bit Addressing:
From: 0000,0000 0000,0000
To: 1111,1111 1111,1111
Two 8-bit words
Understanding word order:
From: 0000,0000 0000,0000
To: 1111,1111 1111,1111
Two byte words
From: 0000,0000 0000,0000
To: 1111,1111 1111,1111
Two byte words (16 bits)
From: 1111,1111 1111,1111
To: 0000,0000 0000,0000
Two byte words
Higher order
digits
Lower order
digits
8-bits = 1 byte
Higher order
digits
Lower order
digits
Little Endian
Big Endian
Middle Endian
This needs to be verified, could be wrong
Nonetheless order matters
Other uses of Words (Use your words wisely)
0000,0000
The first 4 bits are numbers and the second four are decimal numbers
The first 4 bits are numbers and the second four powers of 2 (20000)
The first bit is a sign bit where 0 = + and 1 = -
Hence 0000,0001 = 2
And 1000,0001 = -2
Examples of Digital Voltage Representations:
0000
0
10
20
30
10
5
0
0101
1010
Sampled Voltages
Time (in milliseconds)
10 volts
5 volts
0 volts
Unipolar representations
Pulses of 0 or +5 volts only
Disambiguation: Digital Memory Versus Digital Storage
Units for describing Computer Memory Capaciy and Information Storage Capicity
Computer Memory Capicity
1 Kilobyte kB 2^10 bytes 1,024 bytes onethousand bytes
1 megabyte MB 2^20 bytes 1,024 kilobytes 1,048,576 bytes onemillion bytes
1 gigabyte GB 2^30 bytes 1,024 Megabytes 1,073,741,824 bytes onebillion bytes
Digital Storrage Device of Media Capicity
1 Kilobyte kB 10^3 bytes 1,000 bytes onethousand bytes
1 megabyte MB 10^6 bytes 1,000 kilobytes 1,000,000 bytes onemillion bytes
1 gigabyte GB 10^9 bytes 1,000 megabytes 1,000,000,000 bytes onebillion bytes
1 terabyte TB 10^12 bytes 1,000 gigabytes 1,000,000,000,000 bytes onetrillion bytes
1 petabyte PB 10^15 bytes 1,000 terabytes 1,000,000,000,000,000 bytes onequadrillion bytes
How to interpret .11111111 in Binary
Right of Point 1 1 1 1 1 1 1 1
Power of 2 -1 -2 -3 -4 -5 -6 -7 -8
Then the fractional part (right of the point):
= 1 x 2^-1 + 1 x 2^-2 + 1 x 2^-3 + 1 x 2^-4 + 1 x 2^-5 + 1 x 2^-6 + 1 x 2^-7 + 1 x 2^-8
= 1 x 1/2 + 1 x 1/4 + 1 x 1/8 + 1 x 1/16 + 1 x 1/32 + 1 x 1/64 + 1 x 1/128 + 1 x 1/256
= 0.99609375,
Quiz
What is:
.11111111 (binary) X 2 01001111 (binary)
Mapping ASCII and Unicode
American Standard Code for Information Interchange (ASCII)
Maps the American alphabet, upper and lower case, numbers 0 9, punctuation marks and several carriage control characters to 8-bit.
00000000 through 01111111
Unicode incorporates the worlds languages and ACSCII
May require 4 bytes (32 bits) to represent one character.
Data transfer
Data is moved from one point to another in the form of digital pulses.
If only 0 volts and +5 volts are used this is called Unipolar digital encoding
The pulses can represent numbers, or other maps
A group of wires used to transfer data between several devices is called a Data Bus
Only one device at a time can transmit on a bus
Typically one one device at a time received the data
The sending device sends both the data and the address for the device that is intended to receive the data
Data: Digital Versus Analogue
Pros Digital Format
Data is not distorted by amplification
Immune to noise
Inexpensive
Error detection with redundant data
Pros Analogue Format
Can be faster than Digital
Can be continuous data
Cons Analogue Format
Loss of signal
Signal distortions
External Noise can influence data
More expensive
No error detection
Cons Digital Format
Slower that analogue
Not continuous
If signal is interrupted data is lost
Rounding errors (data loss)
Analogue to Digital Conversion (ADCs, or A/Ds)
Most Transducers, Sensors and Detectors are analogue
Conversion from analogue to digital is called Digitization
Two Steps to Digitization
Sampling (data in-between sample points is lost)
Quantization (rounding errors and data lost in terms of dynamic range step)
Analogue samples are continuous in time
It is impossible to sample the signal in every point in time
Points are sampled for at discrete points in time and for a specific duration dwell time this is called SAMPLING
Each sample is converted into a digital sample Quantization”
Continuous Vs Sampled & Quantized Data
Volts
Volts
Nyquist Violation
Frequency lost
Continuous Vs Sampled & Quantized Data
Volts
Volts
Amplitude Lost
Dwell Time, Sampling Rate, Bandwidth
Dwell Time:
Sampling time/ number of samples
Because the Nyquist sampling theorem states that we must sample at twice the frequency
Bandwidth = =
Pixel Depth and Contrast Resolution (dynamic range)
The number of bits used to store the grey scale of an image is determined by the contrast resolution required.
Ultrasound ~ 8-bits
X-ray CT ~ 12bits
Pixel Size, Septh and Storage Requirements
Filed of View divided by number of pixels = pixel size
Example:
Fluoroscopy field of view 23cm and matrix size of 512X512
Pixel size = 230mm/512 = .45mm/pixel
Image Size = 5122 x 16-bits = 5122 x 2-bytes = 524 kilobytes or 0.5MB
2-images/Megabyte
2,000 images/GB
For a disc that can store 60GB of data one can store just less tha 120 images (should be about 114,000 images)
Some Computer Terms
Hardware
Software
Console Computer
Image Reconstruction Plane
Image Reconstruction Engine
Software
Application Software
PACS
Radiological Information System
Monitor
Terminal
Key Board
Input Device
Pointer Device
Printer
Display
Work Station
Platform
Storage Devices
Tape
Disc
Magnetic Floppy
Optical
WORM
ROM
DVD
Solid State Drive
Hard Drive
RAID
Jukebox ? Analogue Jukebox!
Flash Memory
Cloud Solution
Mirror
Network Terminology
TCPIP
UIS
IT
VPN
Telnet
FTP
Router
HUB
DICOM
HL7
SMPT
Domain
Subnet
NIS
NOC
https
AE title
Image Display
Film
Monitors TG18, TG 270
CRT
Flat Screen
LCD
LED 2k, 3k, 4k, 8k
QLED
Monitors
Three types of monitors are used for medical imaging.
Primary Display Monitor
Secondary Display Monitor
Modality Monitor
Each type has its own use and required performance
Performance Criteria for each type are discussed in AAPM reports
AAPM report TG18 Primary and Secondary Display Monitors
TG18 is a bit aged and focuses on CRT technology
AAPM report TG270 Modality Monitor
TG270 as more information on digital display monitors
Monitor Performance
In depth analysis of Photometrics of Displays and Human perception
Grey scale levels
Contrast resolution
Dynamic Range
Spatial Resolution
Linearity
Background Illumination
Display luminance (Amount of light energy emitted from or reflected from a surface)
Intensity
Maximum
Minimum (Black Level)
Uniformity of Luminance over the viewing area
Viewing distance
Glare
Spatial Distortion
Lag & Refresh rate
Luminance
Rate of light energy emitted or reflected from a surface
Per unit area
Per solid angle
Corrected for the photoptic (normal daylight color vision) spectral sensitivity of the human eye
Unit of luminance is candela per square meter (cd/m2)
Doubling luminance a human eye only perceives a small brightness increase
Dynamic Range of a Monitor
Defined in two different ways
Difference between Luminance Max and Luminance Min ( Lmax Lmin )
Ratio of Luminance Max and Luminance Min ( Lmax / Lmin )
American College of Radiologys (ACR) Technical Standard for Electronic Practice of Medical Imaging
Monochrome Interpretation Workstation Monitors
Lmax > 171 cd/m2
For Mammography Lmax > 250 cd/m2
For optimal contrast Lmax > 450 cd/m2
Radiological monitors have
built in light sensing hardware
feed back loop keeps luminance constant
Black Level, Contrast Ratio & Veiling Glare
Black Level
Measurement of Lmin with lights off & room in total darkness
Difference between Black Level and Lmin is Veiling Glare:
Veiling Glare
Reduces Contrast
When scattered light reflects off the face of a monitor
Diffuse reflection of ambient light
Veiling Glare = Lmin Black Level
Contrast Ratio
like dynamic range except using black level instead of Lmin)
Contrast Ratio = Lmax / Black Level
Spatial Linearity
The preservation of geometric shape
Digital monitors maintain straight lines because they are based on a matrix pattern
Mostly a problem with CRT monitors (almost extinct issue)
Perceived Contrast, Threshold Contrast & Just Noticeable Difference (JND)
Perceived Contrast = (Lmax Lmin)/Lave
Just Noticeable Difference: smallest luminance difference = (L2 – L1) where difference is noticeable by only ½ a group of human observers.
Threshold Contrast = JND/Lave
Visual Degree
1o
60cm
Best perceived image contrast at ~ 5 cycles per visual degree
https://monitors.com/collections/monitors/products/dome-s2c-2mp-color-medical-led-monitor
5-cycles
20-cycles
Your monitor may alias this frequency and produce 1D Moire pattern
Contrast Sensitivity Function
Webvision: The Organization of the Retina and Visual System, Kolb H, Fernandez E, Nelson R, editors., Salt Lake City (UT): University of Utah Health Sciences Center; 1995.
https://www.ncbi.nlm.nih.gov/books/NBK11509/figure/ch25kallspatial.F23/?report=objectonly
5
Max contrast sensitivity is about 5 cycles per visual degree
50
1/100 contrast sensitivity at 50 cycles/degree
Look Up Table (LUT)
One technique to map bit ranges is window and level
P1
P2
Window = W
W/2
Level
Window and Level control (mapping the LUT)
Level sets the middle range of values
Window sets the extent of the range
Lower level = Level –
Upper level = Level +
All values below lower level are set to 0
All values above Upper level are set to maximum pixel value possible
In between pixel values are set:
P1
P2
W = Window = Upper Level – Lower Level
W/2
Level
Pij = (Pij Lower Level) * (Max allowed Pixel value/Window)
Pij = Pixel Value before Windowing
Pij = Pixel Value after Windowing
Luminance Calibration of Display Systems
Display Function: Luminance produced as a function of driving signal
If it is a digital signal this is called Digital Driving Level
Effects perceived contrast
Inherent/generic display function of a monitor is not-linear.
Input to display does not increase linearly with linear increase of input driving level
Can be different for every monitor
Changes over time (Re-calibration needed during the life time of the monitor)
Radiological display systems modify display function to enhance contrast
Various options may be available for specific tasks like mammography
Automated Contrast Equalization
Conspicuity can be increased
Grayscale Standard Display Function (GSDF)
Because different monitors have different Display Functions contrast of a displayed object would differ from monitor to monitor. DICOM has addressed this problem by creating GSDF to standardize contrast across monitors.
Digital Imaging & Communications in Medicine (DICOM)
Presentation Values (digital pixel values) P-Values with GSDF
Provides same contrast between monitors
Provides same contrast between a monitor and printed material
Provides same contrast from monitor/printed material to the human eye
Electro optical measurements can be different than human eye perception
!!! GDSF pertains to grey scale monitors only !!!
Presentation Values (P-values) Digital Driving Levels (DDLs)
Input display systems LUT levels are: Presentation Values
Output systems LUT values are: Digital Driving Levels
Use different LUT for each monitor to achieve GSDF
+ LUT = GSDF
SMPTE Pattern 1) Contrasts Quick View: 0% – 5%, 95% – 100%
2) Line Pairs (resolution
3) Iso-Intensity squares
4) Lines (Geometry measurements)
5) Just noticeable difference calibration Squares
Just Noticeable Difference (JND)
JND: Input value to the GSDF, such that one step in JND results in a luminance difference that is a just noticeable difference
0
100
200
300
400
500
600
700
800
900
1000
1100
Luminance (cd/m2)
10,000
1,000
100
10
1
0.1
0.01
DICOM Grayscale Standard Display Function
250 cd/m2
2 cd/m2
Min = 2 cd/m2 & Max = 250 cd/m2 for this monitor hence only this portion of the GSDF is used
JND Index
J = jmin + p (jmax jmin)/pmax
Image Printers
Typically Digital
Silver based (red or infrared-sensitive) film
Dry thermal development
Exposed with a red scanning laser
Still minimum requirement for Mammography
Computer Terminology
Wide Area Network (WAN)
Local Area Network (LAN) i.e. within a department
Server: Provides service to other computers
File Server: Stores files from other computers on the network
E-mail Server: .
Thick Client: Client computer provides information processing
Thin Client: Client computer serves to display information processed elsewhere
Network Connectivity
Wi-Fi
Coaxial Cable
Other hardwires
Fiberoptic cables (not effected by electrical noise)
Data Passage
Data is passed between computers using Packets
Packets headers contain Network Addresses of transmit and destination nodes
Packets can be lost or dropped
Network-Interfaces between node and network have unique Network Address
Devices connect to network via network-adapter/network-interface
Internet Protocols: are used to check packets integrity and passage
Network Switches: pass packets between network segments
Each device on a network is a Node (computer, switch, printer, etc.)
Links: Communications between nodes
Network Bandwidth: Maximum data transfer rate
transmission control protocol/Internet protocol (TCP/IP)
Uses 5-layer stack:
Applications Applications
Transport Transport
Network LAN Network LAN
Data Link Data Link
Physical Physical
Router
Router
Layer 5
Layer 4
Layer 3
Layer 2
Layer 1
Layer 5
Layer 4
Layer 3
Layer 2
Layer 1
Network
WAN
Transmission Line
Network
WAN
Applications
Transport
Network LAN
Data Link
Physical
Passes information to transport with Destination Computer Address and Requested Application on destination computer
FOLLOWING TCP: Divides info into packets, adds header to each packet with info on packet sequence and error detection
FOLLOWING IP: Further subdivide packets, adds source address and destination address
FOLLOWING specific LAN or point-to-point protocol, encapsulates packets into packets for transmission over LAN, gives each packet another header such as LAN address of destination computer. Example: if protocol is Ethernet, data link encapsulates packets into ethernet packets and sends data to physical layer where they are converted to electrical, infrared or Wi-Fi and transmitted
Layer 4
Under TCP:
Layer 4 of sending computer
initiates dialogue with Layer 4 of destination computer
negotiates packet size and other aspects
Layer 4 of destination computer
Requests re-transmission of missing or corrupted packets
Places packets in correct order
Recovers information from packets (such as requested application)
Sends information to correct application
IPv4 & IPv6
IPv4 uses for 1-byte binary this permits > 4billion addresses
Looks like 1111.1111.1111.1111 (binary) or 255.255.255.255 (base 10)
IPv6 uses six 1-byte binary > 340 undecillion addresses
Should look like 1111.1111.1111.1111.1111.1111 or 255.255.255.255.255.255
Other Internet Protocols
Simple Mail Transfer Protocol (SMPT)
Hypertext Transfer Protocol (HTTP)
Universal Resource Locater (URL)
Teletype Network (TELNET) provides remote command line control
File Transfer Protocol (FTP)
World Wide Web (WWW)
Picture Archiving and Communications Systems (PACS)
Digital archive to store medical images
Display workstations to display images
Computer network to transfer images between
Archive
Imaging modality
Workstations
Database to track location of images and related information
Exchange information with other systems
Hospital Information System (HIS)
Radiology Information System (RIS)
Electronic Medical Record (EMR)
Other connected Systems
CD/DVD
Film Digitizer
Web Server
Health Level-7 (HL7) Integrating the Healthcare Enterprise (IHE)
Health Level-7 (HL7)
a set of international standards used to transfer and share data between various healthcare providers
Integrating the Healthcare Enterprise (IHE)
Initiative by healthcare professionals and industry
Seeks to improve the way computer systems share information
Seeks to improve the interoperability of healthcare information systems
Typical Image Sizes
Study Typical Storage (in Megabytes)
Chest Radiographs (PS and Lateral 2K by 2.5K) 20
Standard Head CT series (50 5122 images) 26
TI-201 Myocardial perfusion SPECT study 1
Ultrasound (5122, 60 images to PACS archive) 16
Cardio-Catheterization Study (Coronal and LV images (maybe s-ray Bi-plane study) 450 3,000
Digital screening mammograms (2cc and 2 MLO) 60 – 132
Image Compression may be needed to store images
!!!Mammograms may not be compressed with lossy!!!
Display Monitors/Viewing Workstations
Work Station is a monitor plus computer with dedicated software
Software includes
Specialty oriented measurement tools
Dentistry
Orthopedics
Neurology
Oncology
Monitor size and placement are critical for reading
High Maximal luminance (400-500 cd/m2)
When the diagonal size of the monitor > 50 inches viewing angle becomes a problem
Curved monitor
Shifting positions to read entire screen
Monitor Performance
Room lighting < 50 lux Reflections from other monitors/light sources (multiple workstations) Viewing distance SMPTE Pattern Software for Workstations Hanging Protocols How the radiologists wishes to see images displayed Magnify (Zoom) Roam (Pan) Window and Level Specialized software Orthopedics (specific angular measurements) Angiographic (Subtractions) Ultrasound (circumference measurements) Image reconstructions and filtering PACS Quality Control Patient Name Right exam right patient Right electronic medical record Proper connection to RIS & HIS Security and Availability Privacy: only those authorized to see data can see it Integrity: error detection to ensure that data is not corrupt Redundancy, back-ups, off-site of data Authentication: Sender-receiver verification Non-repudiation: finalization of reports, transmission of payments are not rejected Availability: Information and services when needed Encryption: Protection against intentional modification Health Insurance Portability Accountability Act (HIPAA, 45 CFR 164) Confidentiality, integrity, availability of data Protect against anticipated threats or hazards to security and integrity of data Protect against reasonably anticipated impermissible uses or disclosures Ensure compliance by the workforce PACS and Teleradiology Security Plan Perform risk analysis Establish Written policies and procedures Train Staff in policies and procedures On and off-site back-ups (fault tolerance) Anti-virus software with up-to-date signature libraries Forbid external data sources (flash drive, e-mails, etc) User authentication Terminate access to former employees Log off workstations when not in use Limit access and control on a user / group basis Secure transfer (encrypted) Physically secure room where data is stored/encrypt the data Destroy information on items before they are disposed of Install security patches on syste Install firewalls Audit trails: recording of access to protected information Establish an emergency disaster recovery plan (fault tolerance) Plan for risks: against human error against natural disaster For emergency services Redundant system Applied Sciences Architecture and Design Biology Business & Finance Chemistry Computer Science Geography Geology Education Engineering English Environmental science Spanish Government History Human Resource Management Information Systems Law Literature Mathematics Nursing Physics Political Science Psychology Reading Science Social Science Home Homework Answers Blog Archive Tags Reviews Contact google+twitterfacebook Copyright © 2021 HomeworkMarket.com
Recent Comments