Hi,performed.
I implemented C-Get SCP in a DICOM Archive application but I'm not sure how to handle the following situation:
The DICOM Archive contains MR images , some are in uncompressed format and some are stored in JPEG format. The client DICOM Application (viewer) is capable of handling both uncompressed and JPEG format for these instances.
When the Client DICOM Viewer issues a C-Get request for some series with MR instances, offers the following Presentation Contexts:
1 classUID: GETStudyRootQueryRetrieveInformationModel transferSyntaxUID: Implicit VR Endian
2 classUID: MRImageStorage transferSyntaxUID: Implicit VR Endian, JPEG Lossless
At the association negotiation phase the DICOM Archive need to pick (select) a single transfer syntax for MRImageStorage. The Presentation Contexts negotiation take place before the C-GET query is resolved and the resulting C-Store operations are
When the transfer syntax selection is performed we do not know the resulting instances. This will lead to possible unnecessary transformations (from uncompressed to JPEG or JPEG to uncompressed) if the selected transfer syntax do not match the formatin witch the instances are stored in the archive.
Ideally moving the instances to the client should not require any transformation of the instances (both application supports both transfer syntaxes) but I'm not sure how to achieve this.The client, instead of offering one presentation context with two transfer syntaxes for the MRImageStorage, should split it into two separate presentation contexts, one with implicit VR endian and one with JPEG lossless so that the archive can accept
Are the conversion of instances unavoidable in this scenario or I'm missing something related to the DICOM protocol in this case?
Regards,
Istvan
On Thursday, January 14, 2021 at 12:23:47 AM UTC+1, istvan wrote:performed.
Hi,
I implemented C-Get SCP in a DICOM Archive application but I'm not sure how to handle the following situation:
The DICOM Archive contains MR images , some are in uncompressed format and some are stored in JPEG format. The client DICOM Application (viewer) is capable of handling both uncompressed and JPEG format for these instances.
When the Client DICOM Viewer issues a C-Get request for some series with MR instances, offers the following Presentation Contexts:
1 classUID: GETStudyRootQueryRetrieveInformationModel transferSyntaxUID: Implicit VR Endian
2 classUID: MRImageStorage transferSyntaxUID: Implicit VR Endian, JPEG Lossless
At the association negotiation phase the DICOM Archive need to pick (select) a single transfer syntax for MRImageStorage. The Presentation Contexts negotiation take place before the C-GET query is resolved and the resulting C-Store operations are
in witch the instances are stored in the archive.When the transfer syntax selection is performed we do not know the resulting instances. This will lead to possible unnecessary transformations (from uncompressed to JPEG or JPEG to uncompressed) if the selected transfer syntax do not match the format
both.Ideally moving the instances to the client should not require any transformation of the instances (both application supports both transfer syntaxes) but I'm not sure how to achieve this.
Are the conversion of instances unavoidable in this scenario or I'm missing something related to the DICOM protocol in this case?
Regards,The client, instead of offering one presentation context with two transfer syntaxes for the MRImageStorage, should split it into two separate presentation contexts, one with implicit VR endian and one with JPEG lossless so that the archive can accept
Istvan
Best,
Paolo
Sysop: | Keyop |
---|---|
Location: | Huddersfield, West Yorkshire, UK |
Users: | 113 |
Nodes: | 8 (1 / 7) |
Uptime: | 55:59:21 |
Calls: | 2,499 |
Files: | 8,661 |
Messages: | 1,910,423 |