5
DICOM Search in Medical Image Archive Solution e-Sushrut Chhavi B K Sahu Health Informatics Division Centre for Development of Advanced Computing Noida,India [email protected] Raghvendra Verma Health Informatics Division Centre for Development of Advanced Computing Noida,India [email protected] Abstract—Digital Imaging and Communication in Medicine (DICOM) standard promotes the communication of digital image in the medical imaging environment. A picture archiving and communication system(PACS) is an electronic and ideally film less information system for acquiring, transporting, storing and electronically displaying medical images. It plays a key role in provision of clinical information in the form of images and reports to informed decision making and provision of care. e-Sushrut Chhavi is such an enterprise PACS Solution developed by CDAC,Noida that goes beyond the Radiology department which caters to the need of any department of the hospital that deals with medical images, DICOM or Non-DICOM. DICOM search is the frequently used feature available in e-Sushrut Chhavi and the DICOM Standard for providing Query/Retrieve facility in the image archiving solutions. In e-Sushrut Chhavi this search mechanism is provided for DICOM images and also for converted DICOM images. By using this search facility any user can quickly access Patient images through a multilevel query User Interface. To achieve a better DICOM search in the archive storage the archive server extracts Patient, Study, Series and image level information from the received image and puts in the archive database. So any search related with user given values with respect to search keywords webPACS system only searches the archive database and generate the results depending on the available image information at that time. So the addition any Patient’s images in the archive storage can be checked at any point of time. And also any addition of a Series of images or Study of a patient in the archive storage will reflect in the search result. The method of extracting information from the DICOM image is to look for the desired tag and its value. In e-Sushrut Chhavi all the images converted to DICOM have similar Study Instance UID, So when user want to list images generated by e-Chhavi modality or application it actually leads to a search based on Study UID. The DICOM search facility can be extended further for the need of any department that generates patient images by getting monthly, quarterly or annual reports for any modality or patient age or any other information field and also benefits hospital as a whole for viewing a specific image or the associated report of the Patient through a secured user profile under a category of users. Keywords-DICOM; DICOM Tag; PACS; SOP Class; C- FIND; C-MOVE; DICOM Information Model; WebPACS I. INTRODUCTION In medical image archiving solutions DICOM image standard is used extensively as the images generated by different modalities such as Computed Tomography(CT), Magnetic Resonance(MR), UltraSound(US), Nuclear Medicine(NM) and Digital X-Ray are of DICOM type. The standard has a unique feature of associating the patient information, acquired date, time, modality type, physician’s name, institute name and other relevant information with each DICOM image generated by these modalities. Each of this information has a DICOM tag specified by Group and Element Number [3]. This DICOM image information model has four levels which are Patient, Study, Series and Image Level. Patient Level is the highest level where all the information for a single patient is taken into account [1]. Patient details contain Patient Name, ID, Date of Birth, Sex and insert time stamp information. Study level is to store information handled by various modalities for a single examination request. A single patient can have multiple studies as the result of previous examination requests. The study details contain Study ID, Study Instance UID, Study Description, Accession No or Bill No and referring physician name information. Series are always a collection of related images coming from a single modality. A single Study can have multiple series of images. Series Details contain Series Number, Series Instance UID, Modality Name, Series Description, Body part examined information, insert time stamp and study parent information. Image level contains acquisition and positioning information besides image data itself [1]. A single series can have multiple images of a patient. Image details contain Image Number, SOP instance UID, image width, height, insert time stamp and series parent information. For Non-DICOM images e-Sushrut chhavi converts the images into DICOM images by associating all four level of patient, study, series and image information with each image. Chhavi provides the facility to edit these values before saving into a DICOM image. DICOM Search in e-Sushrut Chhavi uses Query/Retrieve DICOM model for search and transfer of images from archive storage in the image viewer application. In the WebPACS system one can use the DICOM Search to view the images or SR reports of a patient in the browser. In the browser one can view the list of Patient IDs by typing few initial characters. One can also search all images, series, studies or reports of a particular modality such as CT/MR/NM/US/X-Ray/Non-Dicom. For Non-Dicom images 256 ___________________________________ 978-1- 4244 -8679-3/ 11/$26.00 ©2011 IEEE

[IEEE 2011 3rd International Conference on Electronics Computer Technology (ICECT) - Kanyakumari, India (2011.04.8-2011.04.10)] 2011 3rd International Conference on Electronics Computer

Embed Size (px)

Citation preview

DICOM Search in Medical Image Archive Solution e-Sushrut Chhavi

B K Sahu Health Informatics Division

Centre for Development of Advanced Computing Noida,India

[email protected]

Raghvendra Verma Health Informatics Division

Centre for Development of Advanced Computing Noida,India

[email protected]

Abstract—Digital Imaging and Communication in Medicine (DICOM) standard promotes the communication of digital image in the medical imaging environment. A picture archiving and communication system(PACS) is an electronic and ideally film less information system for acquiring, transporting, storing and electronically displaying medical images. It plays a key role in provision of clinical information in the form of images and reports to informed decision making and provision of care. e-Sushrut Chhavi is such an enterprise PACS Solution developed by CDAC,Noida that goes beyond the Radiology department which caters to the need of any department of the hospital that deals with medical images, DICOM or Non-DICOM. DICOM search is the frequently used feature available in e-Sushrut Chhavi and the DICOM Standard for providing Query/Retrieve facility in the image archiving solutions. In e-Sushrut Chhavi this search mechanism is provided for DICOM images and also for converted DICOM images. By using this search facility any user can quickly access Patient images through a multilevel query User Interface. To achieve a better DICOM search in the archive storage the archive server extracts Patient, Study, Series and image level information from the received image and puts in the archive database. So any search related with user given values with respect to search keywords webPACS system only searches the archive database and generate the results depending on the available image information at that time. So the addition any Patient’s images in the archive storage can be checked at any point of time. And also any addition of a Series of images or Study of a patient in the archive storage will reflect in the search result. The method of extracting information from the DICOM image is to look for the desired tag and its value. In e-Sushrut Chhavi all the images converted to DICOM have similar Study Instance UID, So when user want to list images generated by e-Chhavi modality or application it actually leads to a search based on Study UID. The DICOM search facility can be extended further for the need of any department that generates patient images by getting monthly, quarterly or annual reports for any modality or patient age or any other information field and also benefits hospital as a whole for viewing a specific image or the associated report of the Patient through a secured user profile under a category of users.

Keywords-DICOM; DICOM Tag; PACS; SOP Class; C-FIND; C-MOVE; DICOM Information Model; WebPACS

I. INTRODUCTION In medical image archiving solutions DICOM image

standard is used extensively as the images generated by different modalities such as Computed Tomography(CT), Magnetic Resonance(MR), UltraSound(US), Nuclear Medicine(NM) and Digital X-Ray are of DICOM type. The standard has a unique feature of associating the patient information, acquired date, time, modality type, physician’s name, institute name and other relevant information with each DICOM image generated by these modalities. Each of this information has a DICOM tag specified by Group and Element Number [3]. This DICOM image information model has four levels which are Patient, Study, Series and Image Level. Patient Level is the highest level where all the information for a single patient is taken into account [1]. Patient details contain Patient Name, ID, Date of Birth, Sex and insert time stamp information. Study level is to store information handled by various modalities for a single examination request. A single patient can have multiple studies as the result of previous examination requests. The study details contain Study ID, Study Instance UID, Study Description, Accession No or Bill No and referring physician name information. Series are always a collection of related images coming from a single modality. A single Study can have multiple series of images. Series Details contain Series Number, Series Instance UID, Modality Name, Series Description, Body part examined information, insert time stamp and study parent information. Image level contains acquisition and positioning information besides image data itself [1]. A single series can have multiple images of a patient. Image details contain Image Number, SOP instance UID, image width, height, insert time stamp and series parent information. For Non-DICOM images e-Sushrut chhavi converts the images into DICOM images by associating all four level of patient, study, series and image information with each image. Chhavi provides the facility to edit these values before saving into a DICOM image.

DICOM Search in e-Sushrut Chhavi uses Query/Retrieve DICOM model for search and transfer of images from archive storage in the image viewer application. In the WebPACS system one can use the DICOM Search to view the images or SR reports of a patient in the browser. In the browser one can view the list of Patient IDs by typing few initial characters. One can also search all images, series, studies or reports of a particular modality such as CT/MR/NM/US/X-Ray/Non-Dicom. For Non-Dicom images

256

___________________________________ 978-1-4244 -8679-3/11/$26.00 ©2011 IEEE

one can specify any modality name while saving the image to DICOM, but the study Instance UID is always derived from a particular root which is used to search all converted DICOM images. During image archiving e-Sushrut Chhavi stores all four level of DICOM information model in the database. So for faster search operation in webPACS search module the database is searched giving the same result as that of a DICOM Search in the image viewer application. The user level security also authenticates the access to the image archive storage and database through a security layer.

II. METHODOLOGY

A. DICOM Image Information Model Dicom Image Information Model is based on

assumptions about the way in which information from different modalities is related; see Figure 2.1. The four levels of this information model are Patient, Study, Series and Image.

Each of the DICOM image have Patient Level information such as Patient Name, Patient ID,i.e, unique id generated for a hospital, Patient Date Of Birth, Sex, Age and insert time stamp information in e-Sushrut Chhavi. All these information are used for DICOM Search operation. When any DICOM image is archived in e-Sushrut Chhavi , it extracts using DICOM tag and stores in the database for the search operation in the WebPACS system. The list of standard DICOM tags for the patient level information is provided in TABLE 1 [4]. The associated database table for the patient level of information uses unique Patient ID as a primary key.

Each of the DICOM image also have Study Level information such as Study ID, Study Instance UID, description, Accession No or Bill No, Referring Physician’s Name and insert time stamp information in e-Sushrut Chhavi. All these information are used for DICOM Search operation. The list of standard DICOM tags for the study level information is mentioned in TABLE 1 [4]. The associated database table for the study level of information uses unique Study Instance UID as a primary key.

Each of the DICOM image have Series Level information such as Series Number, Series instance UID, Modality Name, Series Description, Body Part examined info and the insert time stamp information in e-Sushrut Chhavi. All these information are used for DICOM Search operation. The list of standard DICOM tags for the Series level information is provided in TABLE 1 [4]. The associated database table for the series level of information uses unique Series Instance U ID as a primary key.

Each DICOM image contains acquisition and positioning information of the image in the image level. It is the lowest level of the information in the DICOM information model. In case of multiframe images image level contains data for multiple images. One series contains many images, so each image is linked to a series instance UID. Image level information contains image number, SOP(Service-Object Pair) Instance UID, Image width, height and acquisition time.

The list of standard DICOM tags for the Image level is mentioned in TABLE 1 [4]. The associated database table for the image level of information uses unique SOP Instance UID as a primary key.

Figure 2.1 Mapping Real World Examination to Information Model

TABLE 1. LIST OF DICOM TAGS

Patient Level Info DICOM Tag (Group No:Element No)

Patient ID 0010:0010 Patient Name 0010:0020 Patient Birth Date 0010:0030 Patient Sex 0010:0040 Patient Weight 0010:1030

Study Level Info DICOM Tag Study ID 0020:0010 Study Instance UID 0020:000D Study Description 0008:1030 Accession No 0008:0050 Referring Physician Name 0008:0090 Study Date 0008:0020 Study Time 0008:0030

Series Level Info DICOM Tag Series Number 0020:0011 Series Instance UID 0020:000E Series Description 0008:103E Modality 0008:0060 Body Part examined 0018:0015 Series Date 0008:0021 Series Time 0008:0031

Image Level Info DICOM Tag Image Number 0020:0013 SOP Instance UID 0008:0018 Image Height 0028:0010 Image Width 0028:0011 Acquisition Date 0008:0022 Acquisition Time 0008:0032

257

B. Query/Retrieve Information Model The DICOM standard Query/Retrieve information model

is associated with FIND, MOVE and GET SOP Classes [2]. The FIND can be used to query for a collection of images. The Move and GET can be used to initiate the transfer of DICOM images. The actual transfer is done by using the Storage Service Class. DICOM Message Service Element (DIMSE) operation services used for this purpose are C-FIND, C-MOVE and C-GET [3]. Applications may conform to SOP Classes as either a Service Class User (SCU) or Service Class Provider (SCP).

The following three hierarchical Query/Retrieve information models are defined in the DICOM Standard.

1. Patient Root

2. Study Root

3. Patient/Study Root

In Patient Root query model user can specify the attributes as per the four level hierarchies (i.e., Patient, Study, Series and Image Level). Patient Level is the top level and have Attributes associated with the Patient Information. Study Level is below the Patient Level and have Attributes associated with the Study Level. To specify Study level attribute user has to provide at least one patient level attribute. Both Patient and Study level attributes are modality independent. Series level is below the study level and have Attributes associated with the series. A Series belongs to a single study, so to specify any series level attributes user has to provide at least one study level attribute. Image level is below the series level and have Attributes associated with image object. Most of the series and image level attributes are modality dependent.

The Study Root query model is identical to Patient Root query model except that the top level is the study level. Attributes of patients are considered to be the attributes of studies. So the user can start with study level attributes for the search and also can provide series and image level attributes for narrowing the search results.

The Patient/Study only query model is identical to Patient Root except the series and image level attributes are not supported. Even though this model does not include image object attributes, user can retrieve all the images of one Study or more studies of a patient root.

III. RESULTS OF DICOM SEARCH

A. DICOM Search in the e-Sushrut Chhavi Image Viewer Application In e-Sushrut Chhavi DICOM image search

Query/Retrieve DICOM model is being used for search and transfer of images from image archive storage to the image viewer application. User can search the storage by specifying Patient ID or Patient Name using Patient Root Query model as shown in the figure 3.1. User can also feed regular expressions for different attributes to generalize DICOM

search, for example one can feed “2011*” for all patient ids starting with 2011. Besides the IDs user can select the modality name and/or between two Study Dates to further reduce the no of search results. Once user finds the required Patient images, user can move the required Patient or Study or Series images to the image viewer from the archive storage. To use this feature the viewer application requires archive server details and also the archive server to be configured to recognize a valid viewing station and its DICOM messages. Using viewer application the Studies, Series and images can be moved from the archive storage.

Fig 3.1 (a) Patient Root Query Result

Figure 3.1 (b) Study Root Query Result

B. DICOM Search in the e-Sushrut Chhavi WebPACS System In e-Sushrut Chhavi WebPACS DICOM image search

uses DICOM Information model Attributes for performing an image or SR Report search in the archive storage. User has to feed atleast one of any three keywords Patient ID(Patient Level information), Patient Name(Patient Level information) and Referring Physician’s Name(Study Level information) for finding all Studies/Series/Images/SR Reports in the image archive storage. Figure 3.2 shows a typical user session accessing WebPACS System. The user session starts with a user login, remains until logout in between user authentication succeeds for accessing each web

List of Study Ids of a Patient with Id as 253*, Name as sudhakar*

List of Study Ids of all modalities available in the archive Storage

258

pages. Figure 3.3 shows few of DICOM search results in the WebPACS system.

The PHP based WebPACS System uses a set of DICOM utilities for extracting DICOM information which are not available in the database, also uses DICOM Image Viewer activex control for displaying images and related image properties and have a user level security layer for accessing within the hospital. The administrative user can create, delete, modify the user names and roles for a particular department of the hospital. The user name and encrypted password is saved in the MYSQL database for user authentication. This module can use for viewing purpose in a hospital.

Figure 3.3 (a) All Studies of a Patient using Search

Figure 3.2 One User Session in WebPACS

Figure 3.3 (b) Images of a Particular Series

Figure 3.3(c) Image View with Patient Details

Figure 3.3(d) List Patient IDs in the archive storage

IV. CONCLUSION e-Sushrut Chhavi image archiving solution provides the

DICOM Search facility using DICOM image information model. By using key attributes in the Patient, Study, Series and image level user can narrow down the number of search results in the image archiving storage. Once the relevant Studies, Series, Images or Reports are available in the search result, the relevant images or SR Reports can be reached and viewed in the WebPACS System or transferred using image viewer application. The developed browser based search system provides following facilities: (1) View available Patient Ids by typing few initial characters. (2) View modality (CT/MR/NM/US/X-Ray) specific Studies, Series,

List Images of a series of Patient Id 253/201002094176

Image viewer with Patient, Study and Series Details

Patient Ids Starting with 2

List Study Ids of Patient Id starting with 253& Modality type MR

259

Images or Reports. (3) View all images converted to DICOM using Chhavi. (4) View Today’s Studies, Series, Images or Reports. and (5) Browse through Patient’s images and Reports. The browser based viewing of DICOM images and SR Reports in a hospital without installing any application and without storing the images locally is the major benefit of such a system.

This search facility can further be extended to include searching of Patient IDs between two dates keeping other level of information optional. Other statistical reports can also be generated using this DICOM Information Model for a hospital such as number of patients admitted for MRI Scan in a month/year with their Age Group.

REFERENCES

[1] Bas Revet, “DICOM Cook Book for Implementations in Modalities”,Nederland: Philips Medical Systems, 1997.

[2] NEMA, “Digital Imaging and Communications in Medicine Part 4: Service Class Specifications”, NEMA Standards Publication, 2004

[3] Suapang, P., Dejhan, K., and Yimmun, S., “A Web-based DICOM-Format Image Archive, Medical Image Compression and ICOM Viewer System for Teleradiology Application,” SICE Annual Conference 2010.

[4] NEMA, “Digital Imaging and Communications in Medicine Part 6: Data Dictionary”, NEMA Standards Publication, 2004.

260