f65b36de5e091e431093c4e3885fba06.ppt
- Количество слайдов: 50
VISIBLE HUMAN DATASET - MILANO MIRROR SITE® L. Guglielmi - F. Pinciroli vhd-mms/tutorial/base/introcopertina
Visible Human Project the idea. . . z 1987 - The National Library of Medicine long time working draft contains the following item: (NLM should). . ”thoroughly and systematically investigate the technical requirements and feasibility of instituting a biomedical images library” vhd-mms/tutorial/base/intro-1. 11/2
Visible Human Project the idea. . . z 1988 - During a meeting about 3 D anatomical images, held at the NLM Lister Hill National Center for Biomedical Communications, the present people recommend NLM to phisically contribute to the progress of the applicative fileld, directly supporting the development of a image data base of an entire male and female body. The Visible Human Project Idea vhd-mms/tutorial/base/intro-1. 12/2
Visible Human Project the evolution. . . z 1989 - NLM Board of Regents starts a study to evaluate the role of NLM inside the field, instantly variable, of Electronic Imaging. z 1990, April - NLM Board of Regents approve a report that textually said: (The)”NLM should undertake a first project, building a digital image library of volumetric data representing a complete normal adult human male and female” vhd-mms/tutorial/base/intro-1. 2
Visible Human Project the announcement. . . z 1991 - NLM agreed with some medicals of the Colorado University with the aim to search for a good cadaver, and for acquiring the images needed by the project. z 1994, November 28 - NLM give the announcement of complete availability of a data set containing digital images of male anatomy. vhd-mms/tutorial/base/intro-1. 3
Visible Human Data. Set CT and MR images acquisition z Because pilot studies prove that CT and MR images could have a degradation because the body freezing operations, the images, where applicable, are been taken before this frozen procedure y MR images acquisition at Colorado University Hospital (male) x x format: General Electric Genesis 256 x 16 -bit matrix operations start: 18 hours after dead operations end : 21. 75 hours after dead storing: via ethernet to the Medicin School of the Colorado University, and then directly on 4 -mm DAT cartridges vhd-mms/tutorial/base/intro-2. 11/3
Visible Human Data. Set CT and MR images acquisition z CT images acquisition need to immobilize the cadaver with a shape-retaining plastic material y CT images acquisition at Colorado University Hospital (male) x step dimension: 1/3/5 -mm x operations start : 22. 5 hours after dead x operations end : 25. 5 hours after dead vhd-mms/tutorial/base/intro-2. 12/3
Visible Human Data. Set CT and MR images acquisition Male z Freezing of the cadaver yoperations start: 26 hours after death yput into a special freezing chamber until October, 15 1993 ymoving to another maintaining freezing cell with constant temperature of -7°C z. Frozen CT images acquisition y spacing of 1 -mm y first step: (head to midcalf) August, 29 1993 y second step: (midcalf to toes) October, 14 1993 vhd-mms/tutorial/base/intro-2. 13/3
Visible Human Data. Set Tecnica di fotografia z All the slices photographed in red, green and blue (RGB) z All the photos in the 3 tonality (together forming the fullcolor version) has been acquired in TIFF format 2048 x 14 -bit (“raw images”) vhd-mms/tutorial/base/intro-2. 2
Visible Human Project Male choice z Search for the cadavers ysetup of a consortium of SABs (State Anatomical Board) of Colorado, Texas and Maryland. Human bodies that become available for medical research and teaching are btained in mosta states through gifts by their citizens to the SAB. yinventory of a lot of possible cadaver, divided by sex, and setup of a screening process reviewed all the medical records of each candidate ycadaver final choice vhd-mms/tutorial/base/intro-2. 31/5
Visible Human Project Male choice z Cadaver setup ydead at 12: 31 AM on August, 5 1993 (by court-ordered lethal injection at the Texas Department of Corrections in Huntsville) ybody reaches the SAB of Texas about 1. 5 hours later yfinal tests and cadaver setup (previous exams had showed the “normality” of the candidate; only an appendectomy at age 21 and a left orchiectomy at age 15; number 14 teooth had been extracted at age 38) ybody reaches the SAB of Colorado about 8 hours after death yat 12. 5 hours after death were obtained and later digitized 15 radiographs vhd-mms/tutorial/base/intro-2. 32/5
Visible Human Data. Set Final Male choice z. September, 2 1993 NLM’s Visible Human Project Selection Panel, made up of anatomists and radiologists discuss about CT and MR images of two anonymous cadaver. The panel decided unanimously to proceed with SAB number 6022 as the Visible Human Male. So it gave start to all the technical procedures need for sectioning the cadaver and acquiring relative further images. vhd-mms/tutorial/base/intro-2. 33/5
Visible Human Data. Set Female choice Characteristics z. Death: September, 1993 z. Motivazione: Arresto cardio-polmonare z. Housewife z. Etnia caucasica z 58 years old vhd-mms/tutorial/base/intro-2. 34/5
Visible Human Data. Set Female choice General conditions z No fracture or subluxation z Standard bony mineralization z Bilateral degenerative variations at L 4 -5 and L 5 -S 1 level z Minimum degenerative variations present at cervical level z Feet show a contraction, but they don’t seem abnorms z Ventricles have normal dimensions vhd-mms/tutorial/base/intro-2. 35/5
Visible Human Data. Set Sectioning of the cadaver Male z The cutting system was developed at the University of Colorado Medical School in the former Department of Anatomy y. The machine consisted of a milling device whose controls had been automated and drives upgrated to retain accuracy when subjected to the intense cold y Cutting blade: xdisk of 14 inches in dieameter x 20 specially hardened teeth mounted around its edge xspeed rotation: 300 rpm xit succesfully shaved blocks containing tissues to 0. 1 -mm intervals vhd-mms/tutorial/base/intro-2. 41/5
Visible Human Data. Set Sectioning of the cadaver Male z On January, 20 1994, the cadaver was transfered from the walk -in freezer (-7°C) to the freezing chamber (-70°C). After 2 days, the cadaver was removed from the freezing chamber. z On January, 24, 1994, the frozen cadaver, immobilized in its Alpha Cradle mold, was positioned and clamped to the bed of the backsaw. y The initial 4 blocks obtained consisted of: xlegs, ankles and feet xthighs and knees xabdomen and pelvis xhead, neck and thorax y. The kerf loss between them was approximately 1. 5 mm vhd-mms/tutorial/base/intro-2. 42/5
Visible Human Data. Set Sectioning of the cadaver Male z Each block was embedded in a suitable medium, which when frozen would provide stability for the cutting process z Once the gelatine was solid, the mold was transferred to the Ultra Freezer (-85°C) for at least 12 hours. The mold was then removed and partially disassembled. Mold was finally reassembled, with another layer of gelatine siphoned into it. When the final layer was solid, the mold was again transferred to the Ultra Freezer for complete freezing and storage. z Cutting operations y start: February, 17 1994 y end: May, 19 1994 vhd-mms/tutorial/base/intro-2. 43/5
Visible Human Data. Set Sectioning of the cadaver Female z. The Female dataset, instead of the Male one, has a different sectioning step along the Z direction. y. Female sectioning step = 0, 33 -mm y. Male sectioning step = 1 -mm vvhd-mms/tutorial/base/intro-2. 44/5
Visible Human Data. Set Sectioning of the cadaver Female z. The initial 4 blocks obtained consisted of: y 1 st y 2 nd y 3 rd y 4 th head, neck and thorax abdomen and pelvis thighs and knees legs, ankles and feet vhd-mms/tutorial/base/intro-2. 45/5
Visible Human Data. Set Alignment of CT/MR/Fullcolor’s images z Individual fullcolor, CT and MR images have been analyzed to establish the correspondence between them, to provide anatomic alignment. (The number of the images are not the same for all the three types, and something has been lost in kerf) z Anatomic and CT slices wer renumbered to reflect the correspondence, starting at 1, 001 and proceeding to 2, 878. The MR images were renumbered to reflect some anatomic correspondence to the CT and anatomic slices. vhd-mms/tutorial/base/intro-2. 5
Visible Human Data. Set Digital image processing Male z All digital images were reduced from 42 -bit (14 bits of red, green and blue) to 24 -bits images by independently and logarithmically compressing the 14 bits of each channel to 8 bits. z The images were cropped in the anteroposterior (AP) dimension of the anatomy from 2, 048 to 1, 216 pixels. vhd-mms/tutorial/base/intro-2. 6. 1
Visible Human Data. Set Data format Male z Fullcolor: 24 -bit “raw” file format z CT: General Electric Genesis format y. Fresh y. Frozen z. MR: General Electric Genesis format vhd-mms/tutorial/base/intro-2. 6. 2
Visible Human Data. Set Dataset content Male z MRI images: y y axial of head and neck spacing of 4 -mm coronal of the body spacing of 4 -mm 256 x 256 pixels (12 bits of gray per pixel) General Electric format z CT images: y axial spacing of 1 -mm y 512 x 512 pixel (12 bits of gray per pixel) y General Electric format vhd-mms/tutorial/base/intro-2. 6. 31/5
Visible Human Data. Set Dataset content Male z FULLCOLOR images: y axials (aligned with the CT ones) xspacing of 1 -mm for man xspacing of 0. 33 -mm for woman (more than 5, 000 images only for the anatomic dataset) y 2048 x 1216 pixel (24 bits of color per pixel) y raw format, UNIX compressed (e. g. xxx. raw. Z) vhd-mms/tutorial/base/intro-2. 6. 32/5
Visible Human Data. Set Dataset content Female Fullcolor images z. Resolution : voxel 0. 33 mm (same spacing on all the 3 dimensions) z. Photographic format: TIFF (with 14 significative bits per RGB channel) z. Images format: raw format (8 bits per RGB channel) vhd-mms/tutorial/base/intro-2. 6. 33/5 24 -bit
Visible Human Data. Set Dataset content Female Characteristics z. Fullcolor images: yblock 1 (head, neck, thorax) 2048 x 14 bit x RGB TIFF Format yblocks 2, 3, 4 (the other ones) 2048 x 1280 x 14 bit RGB TIFF Format yexcept: first 34 images of the block 2 (abdomen and pelvis) 1760 x 1280 x 14 bit RGB TIFF Format vhd-mms/tutorial/base/intro-2. 6. 34/5
Visible Human Data. Set Dataset content Female X-Ray, CT and MRI images z. MRI Format: General Electric Genesis format 256 x 16 bits z. CT Format: General Electric Genesis format spacing of 1 mm z. X-Ray Format: TIFF vhd-mms/tutorial/base/intro-2. 6. 35/5
Visible Human Data. Set Dataset dimensions z Man: about 15 Giga. Bytes z Woman: about 40 Giga. Bytes vhd-mms/tutorial/base/intro-2. 6. 4
Visible Human Data. Set Fullcolor example Male vhd-mms/tutorial/base/intro-2. 7. 1
Visible Human Data. Set CT fresh cadaver example Male vhd-mms/tutorial/base/intro-2. 7. 2
Visible Human Data. Set CT frozen cadaver example Male vhd-mms/tutorial/base/intro-2. 7. 3
Visible Human Data. Set MRI example Male vhd-mms/tutorial/base/intro-2. 7. 4
Visible Human Data. Set example of a visualization application vhd-mms/tutorial/base/intro-2. 7. 5
Visible Human Data. Set Data set access Directly form the National Library of Medicine (http: //www. nlm. nih. gov/visible_human. html) z Via Internet with FTP z Acquired on cassettes (both 4 mm or 8 mm) vhd-mms/tutorial/base/intro-3. 1
Visible Human Data. Set Dataset amministrative access Directly from the National Library of Medicine (http: //www. nlm. nih. gov/visible_human. html) z Signature of a License Agreement directly with the Nationa Library of Medicine yassignment of a couple of userid/password to grant the access to the dataset via FTP on the NLM server (free access) (Free contract download) ydelivery of the dataset on cassettes (by explicit request) (only charge mail expense and hardware support) vhd-mms/tutorial/base/intro-3. 2
Visible Human Dataset - Milano Mirror Site® Motivations z To solve all the problems European continental users had to download the data directly from NLM, due to network constrictions z To offer a repository service for all the applications built using Visible Human Project data set z To offer to the dataset users a simple, powerful and quick interface to find the data they neeed z To establish an European reference point for the NLM Visible Human Project z To found a reference and development point for applications based upon VHD vhd-mms/tutorial/base/intro-4. 1
Visible Human Dataset - Milano Mirror Site® Goals z A official continental Europe copy of the entire datasets present on NLM servers y Male dataset : about 15 Giga. Bytes y Female dataset : about 40 Giga. Bytes z Storing the entire datasets on RAID disks (about 60 Giga. Bytes!) for on-line download z WEB interface to the data, downloadable via FTP z European reference and help desk point for who need informations about VHD download and use vhd-mms/tutorial/base/intro-4. 2
Visible Human Dataset - Milano Mirror Site® Promoters z Agreeement between Politecnico of Milano (Department of Bio. Engineering) and CILEA (Consortium of the Italian North. West Universities) for the setup and the maintenance of the Continental Europe VHD mirror site vhd-mms/tutorial/base/intro-4. 31/2
Visible Human Dataset - Milano Mirror Site® Sponsors z Sponsorships to maintain the VHD-MMS services at very high level of quality vhd-mms/tutorial/base/intro-4. 32/2
Visible Human Dataset - Milano Mirror Site® Technical criteria z Availability on-line 365 day/year of all the datasets, storing them on SCSI disk with RAID technology (ca 60 Gb) z WEB interface also for FTP download process z Excellent network access for European and Italian users y on a GARR node (Italian academic network) y near a gateway between GARR and commercial ISPs y near the main commercial ISPs italian gateway point vhd-mms/tutorial/base/intro-4. 4. 1
Visible Human Dataset - Milano Mirror Site® Hardware/Software • HP Netserver LX PRO • Bi-processor Pentium 200 • 9 disks RAID, each one of 9 Gb • SCO UNIX Open Server • Netscape Web Fasttrack Server vhd-mms/tutorial/base/intro-4. 4. 2
Visible Human Dataset - Milano Mirror Site® Internet connection >8 Mb/s Internet MIX NSP #n 2 Mb/s > 34 Mb/s NAP #n NAP CILEA 512 Kb/s GARR 2 Mb/s VHD-MMS vhd-mms/tutorial/base/intro-4. 4. 3
Visible Human Dataset - Milano Mirror Site® Web site vhd-mms/tutorial/base/intro-4. 4. 4
Services of the Visible Human Dataset - Milano Mirror Site® Qualified and continuos help desk z Coordinator z Technical Manager Prof. Francesco Pinciroli Tel. (39) (2) 2399. 3335 e-mail: pinciroli@biomed. polimi. it Dott. Luciano Guglielmi Tel. (39) (2) 26995. 267 e-mail: guglielmi@cilea. it http: //vhd-mms. cilea. it/ Informations: VHDmilano@cilea. it vhd-mms/tutorial/base/intro-4. 5. 1
Services of the Visible Human Dataset - Milano Mirror Site® Continuos access to the data via FTP, FTP also using a WEB interface http: //vhd-mms. cilea. it/ ftp: //userid@vhd-mms. cilea. it vhd-mms/tutorial/base/intro-4. 5. 2
Services of the Visible Human Dataset - Milano Mirror Site® To scrape together products based on the “Visible Human Dataset” based products Built of a collection of VHD based product at the Library of the Department of Bio. Engineering of the Politecnico di Milano, made available both from single editors or found on the market Web access to this repository for queries about availability and borrow purpose, will be ready as soon as possible. vhd-mms/tutorial/base/intro 4. 5. 3. 1
Services of the Visible Human Dataset - Milano Mirror Site® Web Site VHD-MMS in all the European language Availability of the Visible Human Dataset - Milano Mirror Site® Web site in all the European languages vhd-mms/tutorial/base/intro 4. 5. 3. 2
Services offered by Visible Human Dataset - Milano Mirror Site® Disk space availability for storing, and delivering via internet, any kind of applications built using the Visible Human Dataset (with special attention to those made up by individuals or groups coming from continental Europe) z availability confirmed z waiting for requests. . . vhd-mms/tutorial/base/intro 4. 5. 3. 3
Services of the Visible Human Dataset - Milano Mirror Site® Money and logistic contribution Contribution assignment, both money and logistic ones, with the goal to promote scientific Ph. D and university degree works, based on the use of the VHD and the optimization of the services offered by the Visible Human Dataset - Milano Mirror Site® vhd-mms/tutorial/base/intro 4. 5. 3. 4
Services of the Visible Human Dataset - Milano Mirror Site® ”Help us to serve you better…” Setup of a direct web line with the users, to collect: Advices Remarks Criticisms Suggestions Appreciations vhd-mms/tutorial/base/intro 4. 5. 3. 5
f65b36de5e091e431093c4e3885fba06.ppt