Physics questions 3 read the PowerPoint slides carefully and answer the questions. MOHY 501 Homework Lecture 3 Medical Imaging Informatics Use the c

Physics questions 3
read the PowerPoint slides carefully and answer the questions.

MOHY 501
Homework Lecture 3
Medical Imaging Informatics
Use the class book or notes to answer.

1. What is the binary number 1101 in base 10
2. How many volts is a TTL pulse?
3. Describe Big Endian and Little Endian give examples
4. What is a sign bit?
5. Can digital voltages represent a continuous wave form? Discuss why or why not.
6. What is: .11111111 (binary) X 2 01001111 (binary)
7. How is Unicode different than ASCII
8. What does ASCII stand for
9. What are the pros of digital formats for transmission
10. What are the cons of analogue formats for transmission
11. What is dwell time? Give a formula for Dwell time.
12. Is amplitude preserved in digital quantized sampling?
13. Write the formula that relates dwell time to bandwidth.
14. How much storage is required to store 28 images that have a 16bit depth a field of view of 25cmX25cm and a matrix size of 256pixles X 256pixels?
15. What does TCPIP stand for
16. What AAPM task group deals with modality display monitors
17. Define luminance and give SI units for Luminance
18. What is the dynamic range of a monitor, give two mathematical definitions
19. What is the minimum max luminance required for a Mammography Primary display monitor?
20. Define Black Level
21. Define contrast ratio in terms of black level (use an equation and label all terms)
22. Define JND
23. Define Threshold Contrast use an equation and label all terms.
24. How would you calculate your maximum viewing distance, give a mathematical example.
25. What is GSDF?
26. What is a presentation value with GSDF and what is the goal of doing this?
27. Describe how to use this image:

28. What is this equation: J = jmin + p (jmax jmin)/pmax define each term and where the data comes from for each term in this equation.and how you would get the data for each term in this equation.
29. Why are fiberoptic cables so useful?
30. What is a Packet
31. Does every network interface have a network address?
32. What is an internet protocol used for?
33. What is a link?
34. In TCP/IP what is layer 3?
35. Name two Layer 4 functions of the sending computer
36. Name four layer 4 functions for the receiving computer
37. What is ivP6, do you think that it will be used or even needed in the future, why or why not?
38. What does SMPT stand for?
39. What is a picture archiving systems responsibilities?
40. Define HL7
41. Define IHE
42. How much storage does a cardio-catheterization fluoroscopy unit need
43. Is it legal to use lossy compression in mammography?
44. What is the maximum room illuminance for a radiological reading room. What are the correct units to measure this in?
45. What is a hanging protocol?
46. Name three PACS quality control parameters
47. Define HIPAA
48. Where can one find the rules and regulations for HIPAA
49. Name the 4 goals of HIPAA
50. Define a PACS security plan that a medical physicist should require. Diagnostic Imaging Physics
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 a group of human observers.

Threshold Contrast = JND/Lave

Visual Degree

1o

60cm

Best perceived image contrast at ~ 5 cycles per visual degree

http://dawnsbrain.com/how-to-draw-a-human-eye-worksheet/

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