Feature based retrieval of echocardiographic images using DICOM structured reporting

Computers in Cardiology, 2003

Olinic²,

Popovici³

et al. 2003

Self Cite

This paper presents our contributions in the design IntroductionUntil recently, medical reports created using the findings in medical images contained only free text, which made unfeasible their semantic processing because of both lack of structure and the large number of possibilities to express the same medical aspects. To solve this problem, DICOM standard has introduced a solution for defining structures of medical documents [1] and controlling their content. The DICOM solution is based on the exclusive use of controlled and encoded terminologies of medical concepts. In order to impose precise requirements on the structure and content of reports, the standard introduced two basic elements that rely on the use of concepts. They are context groups and templates. Contexts are concerned with the information content of a report, while templates refer to the structural aspect. These elements facilitate the systematic description of medical images and the oriented retrieval of reports.As part of our long-term attempt to build a software environment for echocardiography image acquisition, processing, storing, communication, and reporting [2,3], an advanced DICOM compliant structured reporting environment was designed and has been implemented. Structuring methodologyOur approach tries to extend the DICOM echocardiography procedure report, dealing not only with numerical findings, but also with the non-numerical medical diagnosis knowledge. Templates and context groups presented by DICOM supplements cover the numerical findings. We proposed a structuring methodology involving systematically investigation of the heart diseases aiming to extend the set of contexts and templates with those corresponding to non-numerical findings. In order to achieve this, the following approach was adopted:• exhaustive description of the diseases investigated by echocardiography, in terms of anatomical and functional findings, etiological and pathogenic findings, and disease severity assessment • comparison of the disease descriptions for identifying the common contexts and templates • proposal of a consistent set of concepts, context groups, and templates. The experimentation of this methodology required the development of a suite of software tools for coded concepts, context groups, templates and reports editing. 3.Report structureThe structure of DICOM medical reports is specified through templates. They define the items of information a report can contain and the possible relationships between these items. An item of information or content item is a concept name-value pair. Concept names are always coded medical terms. In order to exclude semantic ambiguity that can be generated by the use of a coded concept in different contexts, DICOM standard introduced the context group element. Actual values of a content item can be of various types (e.g. text, numerical value with measurement units, coded term, date, time, image reference). Figure 1 contains an extract from a DICOM report showing the content items and some of the seven...

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

DICOM compliant environment for structured reporting in echocardiography

Computers in Cardiology, 2003

Olinic²,

Popovici³

et al. 2003

Self Cite

“…The environment model is similar to the one in [9] and consists of a 3D lane, driving area delimiter, and 3D objects. We briefly describe in this section the 3D systems of coordinates and the 3D lane model.…”

Section: Environment Modelmentioning

confidence: 99%

“…The search is simple but accuracy is lost because of image re-sampling. In [8], [9] two possible methods were presented, one using image rectification and the second one without rectification. A better 3D accuracy for objects is obtained when the general geometry is used.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Increased Accuracy Stereo Approach for 3D Lane Detection

2006 IEEE Intelligent Vehicles Symposium

Oniga

Danescu

et al.

Self Cite

Curb Detection Based on Elevation Maps from Dense Stereo

Oniga

2007 IEEE International Conference on Intelligent Computer Communication and Processing

Meinecke

2007

A new approach for the detection of curbs (borders of relevant traffic isles, sidewalks, etc) is presented, in the context of urban driving assistance systems. The 3D data from dense stereo is transformed into a rectangular elevation map. The 3D resolution and uncertainty of the stereo sensor are evaluated and used in the detection process. Edge detection is applied on the elevation map in order to highlight height variations. Relevant lines are extracted using a Hough accumulator. Each relevant line is validated as a curb by applying a set of proposed conditions. The algorithm was evaluated in an urban scenario. It works in real time and provides robust detection of curbs.