Browse Prior Art Database

Central Repository for Medical Images and Corresponding Reports

IP.com Disclosure Number: IPCOM000011908D
Publication Date: 2003-Mar-25
Document File: 5 page(s) / 298K

Publishing Venue

The IP.com Prior Art Database

Related Documents

125722: OTHER

Abstract

The use of hospital acquired medical images is becoming increasingly important in the research and development of medical technology. To facilitate new development, patient images will need to be shared between hospitals and various R&D facilities internal and external to the hospital. The immediate need for patient images along with corresponding physician reports is multifold - · For image-based algorithm development, it is essential to have a database of images including a comprehensive representation of examples of non-diseased patients as well as patients with confirmed pathologies. The confirmation of these pathologies or lack thereof is achieved through Biopsy, other single or multi-modality images and/or other tests. The database needs to include the data from multi-modalities (CT, MR, PET, Nuclear, X-ray, etc) and Biopsy Reports. · To compare performance of different algorithms, a database of images and reports as well as a systematic method to store and track the results of various algorithms is needed. · When algorithms are transitioned from the feasibility stage to the product development stage, a database of images along with the “gold standard” results must be provided so that algorithms re-developed by product teams on different hardware/software platforms can be validated. An organized way to store the images and track results provided by various versions of hardware/software is needed. · For teaching purposes at hospitals, a set of good images with well-documented case histories is needed.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 42% of the total text.

Title:      Central Repository  for  Medical Images and Corresponding Reports

BackgroundThe use of hospital acquired medical images is becoming increasingly important in the research and development (R&D) of medical technology.  To facilitate new development, patient images will need to be shared between hospitals and various R&D facilities internal and external to the hospital.  The immediate need for patient images along with corresponding physician reports is multifold -

·         For image-based algorithm development, it is essential to have a database of images including a comprehensive representation of examples of non-diseased patients as well as patients with confirmed pathologies.  The confirmation of these pathologies or lack thereof is achieved through Biopsy, other single or multi-modality images and/or other tests.  The database needs to include the data from multi-modalities (CT, MR, PET, Nuclear, X-ray, etc) and Biopsy Reports.

·         To compare performance of different algorithms, a database of images and reports as well as a systematic method to store and track the results of various algorithms is needed.

·         When algorithms are transitioned from the feasibility stage to the product development stage, a database of images along with the “gold standard” results must be provided so that algorithms re-developed by product teams on different hardware/software platforms can be validated. An organized way to store the images and track results provided by various versions of hardware/software is needed.

·          For teaching purposes at hospitals, a set of good images with well-documented case histories is needed.

Description:    Medical images are usually encoded using the DICOM (Digital Image Communication in Medicine) standard. DICOM images have a header section followed by the image data. The header has several fields, which contain information such as patient name, patient identification (ID), birth date, hospital name, date of acquisition, techniques used for the acquisition, etc.  As dictated by the DICOM Standards’ “Supplement 55”, key fields such as patient name and patient ID are rendered anonyms before the images can be shared with research facilities.  The same requirement applies for medical images encoded using other standards and medical reports encoded with structured reporting.

The process of including images into the database involves reading the DICOM header and extracting key fields from the header as well as key information from physician reports to be placed in appropriate field within the structure of the database.  A user interface for the database will be generated, and the database will be made available internally to personnel needed to access the information via appropriate communication media, such as internet, intranet, data-encrypted secure line, etc.

This invention retrieves images and reports from hospitals, automatically reads the DICOM header fields of the images and enters all relevant informat...