Some Early Computers:
1938 – Zuse Z1 – The Z1 was a mechanical computer designed by Konrad Zuse from 1935 to 1936 and built by him from 1936 to 1938.
1942 – The Atanasoff-Berry Computer – The Atanasoff Berry Computer, later named the ABC, was built at Iowa State University from 1939-1942 by physics professor Dr. John Vincent Atanasoff and his graduate student, Clifford Berry.
1945 – The ENIAC ( 18,000 vacuum tubes. ) – The first programmable general-purpose electronic digital computer, built during World War II by the United States. In the United States, government funding during the war went to a project led by John Mauchly, J.
The use of vacuum tube memory
The use of magnetic core memory
The use of semiconductor memory
Hopkins CT Experience: 1980-2017
Siemens Medical Systems: DEC-PDP11
Pixar: Pixar I and II Image Computer
‘ image rendering’
‘volume rendering’
12-192 Megabyte picture data
NEXT
SUN Workstations 3/160 thru Sparc 2
Silicon Graphicsn Onyx: Infinite Grahics Workstations
Silicon Graphics O2 workstation
SGI O2 with VP500/VP1000 board
Dell 8100 with VolumePro 1000 board and/or
nVidia G-Force 4
InSpace on Leonardo Workstation (HP) with NVIDIA board
iPad 3 with NVIDIA boards
went from “binary classification” to “volume rendering”
“server side processing and post processing”
“multiplanar reconstruction”
- In 1967 Sir Godfrey Hounsfield invented the first CT scanner at EMI Central Research Laboratories using x-ray technology.
By the early 1980s, more than three million CT studies had been performed. Presently that number has grown to well over 100 million CT studies annually,8 with CT becoming the modern doctor’s “truth machine.
filmless radiology
Control computers
parallel processing imaging computers
convolution
backprojection
» The market for Computed Tomography imaging includes the 16-slice, 32/40-slice, 64-slice and ultra-premium (over 64-slice) markets. There is a clear tendency in this market toward higher slice systems because they produce more accurate images. This explains the higher growth of the 64-slice segment.
The development of PACS servers
archival storge costs have decreased
The trend toward cloud storage – acquisition, annotation, collaberation
Development of AI in diagnosis
it generates large digitized data sets that can be subjected to advanced analytics and deep learning.
dataexplosion – an enormous amount of data
“Deep Learning”: – neural network
‘decision trees’
‘random forests’
3D imaging
‘isotropc data sets’ – in multidetector CT
precision medicine
population studies
use of computers in training and educastion
‘enterprise networking’ in hospitals, and displaying images in surgery suite, due to networking – HL7
2013 – US probes that plug into smart phones or tablet
making 3D anatomical models:
three steps: image segmentation, mesh refinement and 3D printing
VR and medical imaging- diagnosis, treatment and medical education
Intuitive surgical
computational power – abilty to use graphical interfaces
“ultrasound on a chip” –
“capacitive micromachined ultrasound transducer”
Multi Core Platforms:
Currently the largest super computer for processing and analysis of mass volumes accumulating raw data provides 129,600 processor cores. A super computer however is not a suitable option for a hospital as they are very large and extremely costly installations. Only a few large research centers can afford such investments.
‘medical imaging servers’
Knowledge-Based Auto Contouring for Radiation Therapy
IMRT -image modulated radio therapy
an advanced form of 3D Conformal Radiotherapy that allows the physician to administer higher and varying doses of radiation to the tumor while sparing healthy surrounding tissue
IGRT – image-guided radio therapy
When undergoing IGRT, high-quality images are taken before each radiation therapy treatment session. IGRT may make it possible to use higher doses of radiation, which increases the probability of tumor control and typically results in shorter treatment schedules.
Moore’s Law and pixel scaling
‘spatial resolution’
the advent of (CCD) charge coupled devices
CMOS sensors – CCD and CMOS sensors power digital cameras. … Both CCD (charge-coupled device) and CMOS (complementary metal-oxide semiconductor) image sensors start at the same point — they have to convert light into electrons.
Parallel Processing Computers:
Cray I- 1976
AI and clinical decision making:
- disease detection
- lesion segmentation
- diagnosis
- treatment assesment
- response assessment
- clinical prediction (of response or future disease)
radiomics -mineable databases from radiological images
a method that extracts large amount of features from radiographic medical images using data-characterisation algorithms. These features, termed radiomic features, have the potential to uncover disease characteristics that fail to be appreciated by the naked eye.
ePAD platform – ePAD is a freely available quantitative imaging informatics platform, developed by the Rubin Lab at Stanford Medicine Radiology at Stanford University. Thanks to its plug-in architecture, ePAD can be used to support a wide range of imaging-based projects.
REFERENCES:
https://www.mayoclinic.org/tests-procedures/image-guided-radiation-therapy/about/pac-20385267
https://keck.usc.edu/radiation-oncology/patient-care/intensity-modulated-radiotherapy/
https://www.doc.ic.ac.uk/~jce317/history-medical-imaging.html
https://www.sciencedirect.com/science/article/abs/pii/S0895611107000262