In this Presentation we talk about :-
What is PACS (Picture Archiving and Communication System).
Functions carried out by PACS.
Storage Devices in PACS
RAID Techniques
Cloud Based PACS
2. What is PACS?
• PACS is an acronym for Picture Archiving and Communication System:
i. Picture-referring to radiographic images and radiology reports
ii. Archiving-referring to the film file or film jacket component of storing
images
iii. Communications-referencing multiple viewers of images and reports at
virtually unlimited viewing sites called workstations.
iv. System-fostering the concept that a complex coordinated network makes it
all possible
3. PACS
• PACS is more than just picture display and archiving, the way in which the image is captured for
PACS is digitally and not analog film/screen processing.
• Because the image is an electronic data set, it can be stored into what is called memory in
computer language. Memory is the key concept in archiving, which takes up less space than a
counterpart x-ray film jacket.
• PACS software can interface with most computers commonly used in medicine to include the
hospital information system (HIS) and radiology information systems (RIS). Patient information
and radiology reports can be displayed with radiology images eliminating the need to store paper
information as well.
4. Basic functions carried out by a PACS system
• Image Capture
• Image Transfer
• Short Term Storage
• Long Term Storage
• Retrieval
• Image viewing
• Networking
5. Storage Devices in PACS
• Short-term storage of PACS documents
• The purpose and function of short-term storage
• The Direct Attached Software (DAS)
• The Network Attached Server (NAS)
• Storage Area Network (SAN)
• Redundant Array of Independent Disks (RAID)
• Back-up storage systems for network servers
• Long-Term Storage and data back-up
• Servers and image/data storage
• Optical disk and optical disk jukebox
• Digital linear tape storage.
6. Short-term Storage of PACS documents
• Images stored on the PACS network by storage servers are said to be
in short-term storage.
• Short term storage refers to those image documents available on the
server that can be immediately viewed from any network
workstation.
7. Short-term Storage of PACS documents
• Direct attached storage (DAS) is using the hard drives that come with
the Server.
• A 5U server might have the capacity for 10 to 12 drives to be added,
typically yielding a capacity of around 2 to 3 TB.
• This is the least expensive approach for small storage requirements
but also provides the least amount of redundancy.
• If any component fails on the server and stops the application, access
to the data will also be lost. Another challenge of DAS is the likelihood
of running out of space.
8. Short-term Storage of PACS documents
• Storage area network refers to a network dedicated to storage access at the block
level. A SAN uses the Fibre Channel communications protocol to directly access
storage devices. The file system remains on the computer, but low-level block-
level access is routed through an external network of storage devices.
• A SAN is perfectly suited for a busy information services department running
dozens or even hundreds of servers. A SAN can be used to handle the storage
needs of servers running different operating systems, although each system sees
only its own blocks of data.
• A SAN is the most frequently used storage method for PACS database
redundancy.
• SAN is the highest performance type of scalable storage architecture because it
provides such low-level access—but it is also the most expensive.
• One reason is that optical connections are more costly than copper ones.
9. Short-term Storage of PACS documents
• This is usually accomplished by a specialized server such as
a Network Attached Server (NAS) which makes access to PACS images
faster than it they are stored on a separate storage network.
• The recommended memory capacity of a NAS if used as a short-term
storage device is usually several terabytes which is expandable to
petabytes.
• Storing digital data is only one of the many functions of the NAS; it
must also retrieve errorless data files to network workstations, WEB
servers, and to printers as well.
10. Short-term Storage of PACS documents
• Patient information such as the medical record number, name, date-
of-birth, type of study, date of study and the like is entered into the
PACS record through a data bridge. Thus one of the most important
functions of the data bridge is to transfer selected information from
HIS and RIS into PACS so that the patient's electronic file is
established.
• Since each patient has an exclusive medical records file number, it can
be used to retrieve electronic documents
11. Short-term Storage of PACS documents
• The process of the workflow manager pulling all studies of a patient
file to the main server when the patient medical record number is
added to the worklist is called prefetching. Prefetching is analogous to
the file clerk retrieving a patient's jacket for the radiologist in case
prior studies are needed for comparison. When a workstation viewer
opens the patient file, smooth seamless acquisition of all image
studies are presented on the menu.
• The point here is that the workflow software in PACS will bring
forward all images in a patient's file whether new data is added or not
whenever a medical record number is accessed.
12. Short-term Storage of PACS documents
• The PACS workflow manager anticipates that studies in the
"electronic file" may be needed for comparison or that reviewing of
previous reports may be desired, and therefore makes them readily
available as a pre fetched function.
• By anticipating files of image documents are retrieved as a group to
be easily accessed from the NAS.
• If a new study is added, the entire sets of files are saved as a group.
This keep images updated so they are not discarded for a minimum of
5-7 years following inactivity.
14. Short-term Storage of PACS documents
• The NAS provides data protection through a Redundant Array of Independent
Disks (RAID) formatting.
• Within the NAS server are several hard drives with multiple disks (at least 2 or
more) that are designed for reliability of data storage and optimal retrieval
performance.
• One type of array in which identical copies of data are written on two different
disks drives within a server is called disk mirroring.
• Mirroring provides data back-up so should a system failure occur on any one disk
its identical disk will respond so that the workstation user experiences no loss.
• Disk mirroring is also called a RAID Level 1 format. Drives in a mirror array must
be added in pairs because they also function as a back-up data retrieval system.
There are several types of storage arrays that back-up data from complete loss.
15. • RAID 0, striping, splits data to 2 hard drives simultaneously. The benefit to striping
is that the storage system can be twice as fast at reading and writing data
because it utilizes 2 drives. The problem with striping, however, is that there is no
data redundancy. If either drive fails, all data on both drives are lost.
• RAID 1, mirroring, copies the data to a shadow hard drive. RAID 1 provides
complete redundancy in the event that 1 of the hard drives fails but does not
enjoy any performance benefits because the user is still reading from a single
drive. The other challenge to RAID 1 is effectively the loss of the capacity of the
second drive.
• RAID 10 is called a nested strategy and combines RAID 1 and RAID 0 levels. Two
hard drives are striped, and both of those hard drives have a mirror copy in case
of failure. This simple technique provides the performance benefits of 2 drives
while allowing for failure. The drawback of RAID 10 is that half the storage is used
making a backup of the data.
16. RAID 0
• Following are the key points to
remember for RAID level 0.
• Minimum 2 disks.
• Excellent performance ( as blocks
are striped ).
• No redundancy ( no mirror, no
parity ).
• Don’t use this for any critical
system.
17. RAID 1
• Following are the key points to
remember for RAID level 1.
• Minimum 2 disks.
• Good performance ( no striping. no
parity ).
• Excellent redundancy ( as blocks are
mirrored ).
18. RAID 5
• Following are the key points to
remember for RAID level 5.
• Minimum 3 disks.
• Good performance ( as blocks are
striped ).
• Good redundancy ( distributed parity ).
• Best cost effective option providing
both performance and redundancy. Use
this for DB that is heavily read oriented.
Write operations will be slow.
19. RAID 10
• Following are the key points to
remember for RAID level 10.
• Minimum 4 disks.
• This is also called as “stripe of mirrors”
• Excellent redundancy ( as blocks are
mirrored )
• Excellent performance ( as blocks are
striped )
• If you can afford the dollar, this is the
BEST option for any mission critical
applications (especially databases).
20. Long-term storage and data back-up
• The main server or the network attached server can handle most of
the data storage capacity for a large institution; however, there are
medical legal requirements for long-term storage of medical
information that must also be considered.
• The PACS system must carry out the functions of data storage that
include long-term storage of records. These records can easily be
retrieved from PACS just as a film jacket can be retrieved from a
remote location if needed or purged if outdated.
21. Long-term storage and data back-up
• A storage function not easily accomplished with analog films is the
recording of back-up files in case of a disaster that destroys the main
systems. A back-up disaster image file for analog film is primarily
accomplished using a microfilm scanner or a minifier system.
• These may require some darkroom processing and a lot of time to
produce and store duplicate images. PACS is able to provide disaster
file using optical disk technology and/or digital linear tape technology.
22. Long-term storage and data back-up
• Most institutions save data on network servers, optical disk within an
optical disk jukebox (ODJ), and on digital linear tape (DLT), or in a DLT
jukebox, or combinations of all of the methods mentioned.
• An optical disk jukebox is a long-term storage hardware device which
encompasses optical disk drives, optical disk storage slots, and
associated robotic arms and software for fetching data disk(s).
23. Long-term storage and data back-up
Pictures above shows an Optical disc jukebox (white arrow) used to store images for long-term storage.
The pictures with the blue arrows are of an optical disc with a 2.3 gigabyte capacity and OD burner
commonly used to back-up CT images
24. Long-term storage and data back-up
• Because the data stored in an ODJ must be fetched from its location
by a robotic arm to the driver to be read, the time required to process
images from an optical disk jukebox to the main server can be
upwards to 30x retrieval time of a NAS.
• This is why the workflow manager uses software protocols to pre-
fetch image files when a medical record number with prior studies is
sent to the workflow list.
• The workflow manager server will bring forward all data records from
short and long term stores such as the NAS and optical disk jukebox,
but never from digital linear tape stores.
25. Long-term storage and data back-up
• Data back-up in case of a disaster that could destroy the NAS or Storage
Area Network (SAN) servers is handled by Digital Linear Tapes (DLT).
• The DLT recorder stores all images in PACS on 1/2 inch magnetic tape.
These tapes can be combined into a jukebox, but are most often stored at a
remote location since the tapes are not accessed by PACS for routine use.
• The information they contain must be loaded into PACS should a disaster
occur; therefore, testing should be done whenever hospital wide disaster
drills are held.
• DLT offers the cheapest method of backing up medical imaging files, and
requires the least space for storing disaster recovery files.
26. Long-term storage and data back-up
The two pictures are of a digital linear tape system used to make disaster tapes of all images recorded onto
PACS. This is provided in addition to other storage servers for image viewing. The tape shown looks like an
old cassette type and is less than 5 inches long. These tapes are stored at a remote location away from the
institution.
27. Transfer/Storage of Data
• Both short-term and long-term data can be stored on a separate sub-
network of dedicated storage devices called a Storage Area Network (SAN).
• The speed of data transferred over a network is dependent on type of line
over which data is transmitted such as a fiber optic line, and the network
architecture.
• Most digital images are several megabytes in size each; in order for a SAN
to handle data from so many inputs it is important that it uses fiber optic
lines and Fiber Channel architecture to move duplex data at transfer rates
of near 100 megabytes per second.
• At this speed a separate network for image storage is reasonable;
otherwise a network attached server is much faster at retrieval.
28. Cloud based PACS
• Cloud-based PACS are a type of PACS architecture that stores and backs up a
healthcare organization's medical imaging data in an offsite server, not within the
organization's physical location.
• Users with permission to access medical images can do so at any time through
the cloud PACS. Cloud PACS can also enable medical personnel to view medical
imaging data from any of their approved devices.
• Thus, there is an immediate reduction in the cost of the storage and archiving for
the hospital, clinic or imaging center, since there is no longer a need for investing
in these areas. Absence of an onsite data center reduces the ancillary
investments that would be associated in running it, such as the emergency power
backup or uninterruptible power supply (UPS), as well as the heating, ventilation
and air conditioning (HVAC) systems. An offsite data center managed by the
vendor puts the responsibility of data backup, disaster protection and recovery
responsibility on it.
29. Cloud based PACS
• There is a further reduction in overall costs, since technology management
for IT systems is outsourced to IT experts. With the absence of many
systems, valuable square footage is saved that can be utilized by the
hospital or clinic to set up more beds and accommodate more patients.
• Apart from economic benefits and provision of better quality care, PACS
also can help to reduce the carbon footprint. With the reduction in the
overall number of hardware equipment pieces, there should be a lower
amount of energy demanded in the production, running, cooling and,
ultimately, disposing of these systems.
• With lower energy demands, there also would be a reduction in the
relative carbon dioxide emission, helping the organization as a whole adopt
a “greener” stance in its operation.
30. Cloud based PACS
• Despite the benefits, the adoption of cloud-based systems is not easy. The
presence of a reliable, high-speed, bandwidth Internet connection is the
most important factor when considering such systems, as it is the
operational backbone. Speed and bandwidth should be considered when
setting up such a system, as the speed of accessing and downloading
medical images from the online archive should be within acceptable limits,
even when several different users are accessing the “data cloud”
simultaneously.
• Cloud-based PACS with offsite data centers hold data security of utmost
importance. Strict third-party checks and audit trials are required in order
to ensure that these data centers are secure in storing medical data and
capable of maintaining confidentiality.
31. Cloud based PACS
• Currently, hospitals and healthcare providers may be interested in the cloud-based
systems; however only medical images, surgery data and research data are uploaded
onto an online data cloud.
• The patient’s highly sensitive general details are mostly kept in onsite data storage
devices as there is still a level of fear and apprehension when putting medical data in the
cloud.
• For hospitals interested in such systems, it might not be possible currently to transfer
large amounts of data or even to change their cloud service provider. Such systems are
still in their infancy, with business models yet to be developed on a larger scale. Thus,
there is apprehension about adopting a new system, such as cloud-based PACS.
• There might also be problems in terms of bureaucratic barriers in some countries, as it is
illegal to actually send medical data outside the premises of the hospital or transmit the
medical data to individuals not employed by the hospital. Thus, there is a huge barrier
that does not allow the data to be stored at the offsite data center that may not always
be within the country, with the concept of outsourcing some of the hospital IT duties.