which is the best DICOM Toolkit ?
Blech
I'm writing a diploma about DICOM. One Part is to find a provider for a
DICOM Toolkit.
After my internetrecherche the favorits are :
- MERGE
- DEJARNETTE or
- MITRA
The Price is not the problem.
Has anyone experience with one of the DICOM toolkits from these
providers ?
Is there anyone who can tell me wich ist the best (best Support, best
reliability, performace, manuals, actual updates ...) ?
Many thanks in advance for your hint.
---------------------------------------------------
Michael Blech
FH Brandenburg - University of Applied Sciences
Informatik - Computer Science
Magdeburger Straße 50
4770 Brandenburg an der Havel GERMANY
http://www.fh-brandenburg.de
mailto:bl...@fh-brandenburg.de
---------------------------------------------------
King of the Wizzers
My company paid 2,500 for version 10.5, had problems and then upgraded
to 11.5 for 250 and it still did not work !!!!
I sent a problem report and an example to their tech support and never
got back a response Dec 1999
I then asked in Jan 2000 and was told that yes the problem was real
and that it would be made an A priority. But have not heard anything
back since.
Since I refuse to pay another 250 for another "upgrade" I have dumped
Lead Tools and Will respond in this fashion any chance I get.
I then spent 1 month using learning UC Davis Toolkit
and Dave Clunie's tool kit in the WIN32 environment and have managed
to get everything working.
Better yet these toolkits are FREE
DO NOT USE LEAD TOOLS !!!!!!!!
Miro Landa
I'm using MERGE toolkit and have my opinion on it, which I want to share
with you.
This toolkit are written on plane C language and is DOS-oriented.
I'm using it in VC++ 6.0 under WIN NT 4.0 environment thought and have no
serious problems with it.
It seems to be fully DICOM standard complaint either.
BUT:
- it supports just basic DICOM Service Class behavior (NO Asynchron
Association Protocol, NO Relational Queries), that is a very serious
restriction in multitasking environment
- huge inconvenient setup...
- which still doesn't allow me to change Association Negotiation the way I
like (so I'm forced to allow everything in setup, gather info after
association established, maintain my own client list and program all the
security and compliance checking manually)
- it seems to be slow (more than 90% of working time my programs seems to
run inside their subroutines)
- they don't supply the source codes, so debugging is a mess (I must deal
with assembler, registers, stack, memory etc.)
- the documentation is not so good (confusing sometimes).
Their support team reaction time is very good, but it is not a big assist
since poor system architecture (I mean, if your car has no reverse gear
nobody can help you to park it in the crowded street).
Considering above, I think that it is suitable for plane simple applications
only, but in this case it works fine, it is bug-free and convenient.
Regards,
Miro.
dee csipo
Glad to hear that money is not a problem for you because all three of the
toolkits you picked carry a heavy entry fee (about $7500) and a rather
selective and arbitrary process of deciding to whom the companies will
license the toolkits. But here is a brief summary.
Merge:
The merge toolkit has been well maintained over the years. It has solid
support and commitment behind it. Don Van Syckle is the WKG6 chair, it
would not look good if they did not keep up with the standard. The toolkit
follows a socket model (you make an association send and image wait for
response release) It is fairly intuitive. I have not personally kept up
with the higher level support changes, but last i checked you still have to
know an awful lot about DICOm to be able to use it.
DeJarnette:
This is real close to the concrete, but gives you all the flexibility you
need. You have control over everything. There is almost no DIMSE layer
implementation, you have to do that yourselves. ( I doubt that it had
changed since I left that company.) The big advantage that it will run on
almost anything that has an ANSI C compiler. The interface to the library
is real simple. You treat the disk files the same as any other protocol.
Last i checked the toolkit did not have DICOMDIR support, you may need to
negotiate that with the company, i know they have it written for some of
their products.
Mitra
Similar to the Merge toolkit in the level of support, but it is closer to
the CTN in design philosophy than the other two. It uses callbacks. It is
slightly more OO than any of the previous ones.
If i were you i would get one of the free ware libraries. The Oldenburg
Offis Toolkit is probably your best choice since it is better documented
then the CTN, and works right out of the box on NT. The UCDavis code is an
interesting package if you want to see how an OO approach can be used to
implement DICOM.
Remember none of these toolkits will get you out of learning about DICOM
message structures and DICOM communications basics.
dee
;-D
ge...@my-deja.com
pretty satisfied with it.
They keep it up to date with standards changes and I have had very few
problems with bugs in their new releases. I just rebuild everything
with the new release and it works. I initially used this on Unix but
now am using it on NT and had to do close to no code changes related to
their toolkit to port my code.
It has documented both C and C++ interfaces. I have been
exclusively using the C++. Their documentation could be better as it is
just a reference manual and there is no effort to tie it all together
in more of a developer's guide that explains why to do something one
way over another. They do provide sample code for some of the basic
DICOM functionality in both C and C++, and I see something about Basic
in their NT release, but I haven't looked into it. Since I use the
toolkit in an object-oriented fashion, it was pretty straight-forward
using the sample code as a guide-line. The sample code shows
an individual function and I was left to my own to put it all together,
which was a little confusing at first. But Mitra has also been good at
getting back to me about any questions and put up with a lot of
learning-curve DICOM questions also.
Other than the price, I have been pleased and found it relatively easy
to understand. Learning about DICOM was a much bigger effort in the
beginning than learning to use the toolkit.
Hope this helps,
Ginny
Analogic Corp.
Sent via Deja.com http://www.deja.com/
Before you buy.
walter...@my-deja.com
we are currently evaluating the ETIAM
(www.etiam.com) toolkit.
Seams nice : powerful and easy to use, if you
don't want to go deeply into DICOM literature.
Have a look at DICOM Eye tool also, for test and
debug of your app.
Regards,
Klaus
In article <3A1159C7...@fh-brandenburg.de>,
Sent via Deja.com http://www.deja.com/
Before you buy.
keesh
I've had some experiences with Accusoft's MISDK...
Although they eventually solved my problem
(a limitation in their API), the process was anything but smooth.
They seem more interested in celebrating their successes
(cruises, trips, etc.) than making a world class product.
Maybe I should help them with their efforts,
at least I'll get a free trip out of it.
We're still hoping for the best
while we evaluate them as a whole.
Sincerely,
keesh
-----
<walter...@my-deja.com> wrote in message
news:8van2h$rka$1...@nnrp1.deja.com...
Dave Harvey
> I'm writing a diploma about DICOM. One Part is to find a provider for
a DICOM Toolkit.
At the risk of advertising, you could try DicomObjects if you want
something commercial, with good support, but a low entry cost.
ActiveX - works in most PC environments, shields you from much of the
nitty-gritty.
Dave Harvey
Medical Connections