CAD–PACS integration tool kit based on DICOM secondary capture, structured report and IHE workflow profiles
Introduction
Computer aided detection/diagnosis (CAD) can improve the efficiency and accuracy of clinical diagnosis by automatically detecting abnormalities and/or pathologies in medical imaging and performing quantitative analysis. Many CAD applications have been conducted in the imaging community during the last 20 years [1], [2], [3]. Among them, CAD applications in mammography [4], [5], [6], chest [7] and three-dimensional (3D) CT lung imaging [8] have emerged as commercial products over the last 5–6 years. While these commercial CAD companies have devoted most of their time to CAD algorithm development and robustness, little effort has been performed on the system integration of CAD results with a picture archiving and communication system (PACS) [9], [10] where CAD is used. Current clinical practice and workflow usually have CAD results limited to a standalone CAD workstation or a CAD server without integrating the CAD results with a PACS. Fig. 1 shows a typical CAD process flow in current clinical practice. Generally, the RIS (radiology information system) orders the CAD process for an exam when the exam is scheduled. The technician or the radiologist pushes the original image exam from the PACS server or the PACS workstation (WS) to the CAD WS for process. The CAD result is reviewed in the CAD WS. Any physician who wants to review the CAD results has to access the CAD WS, which is usually only available in the radiology department. In order to benefit most clinical physicians, CAD results must be integrated with a PACS where physicians can easily query and retrieve CAD results and the original clinical images to their viewing WS for reviewing.
Digital imaging and communications in medicine (DICOM) [11] standard has defined two structured report (SR) templates for mammography and chest CAD results. The Integrating the Healthcare Enterprise (IHE) [12] has also published a Post-processing workflow profile for integrating CAD applications with clinical PACS workflow. Under the guidance of these initiatives, some commercial vendors [13], [14] are developing the software to integrate their CAD application results with a PACS. However, the system integration software is tightly coupled to their CAD applications and is difficult to be used to integrate third party CAD applications with a PACS.
The ever-increasing field of CAD research and development has produced many applications developed by academic institutions or small companies. Unfortunately, these applications still remain in a standalone CAD workstation, CAD server, or diagnostic workstation. Thus, a toolkit that can facilitate the integration of CAD results of these standalone CAD applications has become an urgent requirement for clinical CAD use.
In this paper, we present a CAD–PACS integration toolkit for this purpose. The remainder of this paper explains the system integration approach of the toolkit and shows some examples of applying the toolkit for CAD–PACS integration.
Section snippets
Methods
The goal of the CAD–PACS toolkit is to store the CAD results in the PACS server and integrate the CAD process flow with a PACS seamlessly. The CAD–PACS toolkit is developed under the premise that the CAD software can be either in a standalone CAD workstation, a CAD server, or integrated within the PACS workstation (WS). The CAD–PACS toolkit is a software package with two versions, the DICOM-SC (secondary capture) version, and the DICOM–IHE version (Fig. 2). The major software package in
Clinical implementation of CAD–PACS with a clinical PACS
Three approaches can be used for clinical implementation of CAD–PACS dependent upon where the CAD software is located, and if the PACS is DICOM- and/or IHE-compliant.
Results
The CAD–PACS toolkit has been integrated with a PACS simulator [15], [16] for evaluation at the IPI Laboratory. Three CAD applications, computer aided bone age assessment [17], [18], computer aided detection of small acute intracranial hemorrhage on CT of brain [19], and computer aided detection of emphysema [20], have been used for evaluation of the toolkit.
Discussion
The integration, implementation, and evaluation of the CAD–PACS i-CAD-SC (secondary capture) version is simple and straight forward and does not require special methodology or additional software modules as described in Section 2.
On the other hand, the integration, implementation, and evaluation of the CAD–PACS DICOM–IHE version which uses structured report are more involved. The previous three approaches in Section 3 have been developed for the clinical implementation of CAD–PACS with a
Summary
We have developed a CAD–PACS integration toolkit to facilitate integrating results of a standalone CAD application with a PACS. The CAD–PACS toolkit contains two different integration methods, DICOM-SC and DICOM–IHE. The former is for quick deployment and the latter for long-term use with emerging DICOM and IHE compliant solutions. Our experimental results shows both methods are effective to integrate CAD results, including quantitative measurements and CAD segmented images, with a PACS
Dr. Zheng Zhou got his PhD from the Department of Biomedical Engineering, University of Southern California. He is currently a postdoctoral research associate at Image Processing and Informatics (IPI) Laboratory, Department of Radiology, University of Southern California (USC). Dr. Zhou has more than 6 years R&D experience in PACS and Imaging Informatics. He has developed a data storage grid that provides a virtual fault-tolerant archive for storing and sharing PACS images, clinical trial
References (20)
- et al.
Software architecture for a CAD server
CARS
(2004) - et al.
New direction in PACS education and training
J Comput Med Imag Graph
(2003) - et al.
Image feature analysis and computer-aided diagnosis in digital radiography. 3. Automated detection of nodules in peripheral lung fields
Med Phys
(1988) Computer-Aided Diagnosis in Medical Imaging
(1999)Lung nodules and beyond: approaches, challenges and opportunities in thoracic CAD
CARS
(2004)Potential contribution of computer-aided detection to the sensitivity of screening mammography
Radiology
(2000)Computer aided detection (CAD) in screening mammography: sensitivity of commercial system in detecting architecture distortion
AJR
(2003)Image feature analysis and computer-aided diagnosis in digital radiography. 1. Automated detection of micro calcifications in mammography
Med Phys
(1987)A computer aid for radiologists: improved detection of small volume lung cancer on chest radiographs
RSNA
(2003)CAD of actionable lung nodules on multislice CT scans of the lung: sensitivity and false positive marker rates
RSNA
(2002)
Cited by (29)
Design and development of an ethnically-diverse imaging informatics-based eFolder system for multiple sclerosis patients
2015, Computerized Medical Imaging and GraphicsCitation Excerpt :After designing the custom template for lesion quantification results, Fig. 5 shows the workflow steps of converting MATLAB-based quantification results into DICOM-SR objects. The first step is to convert quantitative results to an XML (Extensible Markup Language) document, which then is converted to DICOM-SR via the dcmtk open-source toolkit [18]. In order to generate the correct XML document, sample DICOM SR files obtained and converted to XML.
Bone age assessment of children using a digital hand atlas
2007, Computerized Medical Imaging and GraphicsDesigning and validating a cost effective safe network: Application to a PACS system
2019, International Conference on Advances in Biomedical Engineering, ICABMEDICOM image quantification secondary capture (DICOM IQSC) integrated with numeric results, regions, and curves: Implementation and applications in nuclear medicine
2017, Progress in Biomedical Optics and Imaging - Proceedings of SPIEProposal for an image-based software system for dental implant positioning
2015, 2015 IEEE 15th International Conference on Bioinformatics and Bioengineering, BIBE 2015Integration of CAD and PACS
2015, Computer-Aided Detection and Diagnosis in Medical Imaging
Dr. Zheng Zhou got his PhD from the Department of Biomedical Engineering, University of Southern California. He is currently a postdoctoral research associate at Image Processing and Informatics (IPI) Laboratory, Department of Radiology, University of Southern California (USC). Dr. Zhou has more than 6 years R&D experience in PACS and Imaging Informatics. He has developed a data storage grid that provides a virtual fault-tolerant archive for storing and sharing PACS images, clinical trial images and analysis results. He has also developed a security method that can protect the integrity of two-dimensional and three-dimensional medical images and a system integration method to apply the security method in IHE post-processing workflow and key image note profiles. He was involved in the development of PACS Simulator for PACS trainings, the Internet2 applications for medical images, and a fault-tolerant PACS archive server as well. His current interest is on medical image informatics, data grid, integrating CAD with PACS, image security, HIPAA compliance, and ePR.
Dr. Brent Liu earned a PhD degree from the UCLA Biomedical Physics Graduate Program and performed research as a postdoctorate. He is currently has a joint appointment with the Departments of Radiology, Keck School of Medicine, and Biomedical Engineering, Viterbi School of Engineering. He is also a senior research staff member of the Image Processing and Informatics Laboratory located at Marina del Rey. He has implemented fully filmless PACS in a clinical setting within the Imaging Department of both a high-profile community hospital (Saint John's Health Center, Santa Monica) and a high-profile academic hospital (UCLA) that has multiple campus sites and is currently advising multiple hospitals on their PACS process, including the USC Health Science Campus. His research areas of interest include Medical Imaging Informatics, Picture Archiving and Communication Systems (PACS) clinical uptime and usability, new PACS technology, Disaster Recovery for PACS, design and implementation of high-resolution image display workstations, next generation Internet and its clinical applications, and advances in the area of image processing and information management for healthcare including Security and HIPAA-compliance related issues.
Anh Le is a PhD candidate of Biomedical Engineering Department at University of Southern California. She received her BS degree in Computer Engineering at University of California at Irvine (UCI) in 2005. Currently, she is working as a research assistant at Imaging Processing and Informatics Laboratory (IPI) on PACS–CAD toolkit development project. She has experience on designing and fabricating microfluidic devices when worked on Cell Encapsulation Project at Biomolecular Microsystems and Nano Tranducers Laboratory at UCI. Her interests include image processing, PACS and medical informatics.