RSNA 2012 - Next Generation Infrastructure for Medical Imaging

Interoperability and Integration—from HL7, DICOM, IHE, to SOA

Modern radiology workflow requires consumption, choreography, and orchestration of content from multiple disparate information systems that do not natively “talk to each other.” Without optimal integration and interoperability amongst these systems, humans are required to serve as “integrating agents:” this frequently results in inefficiency and error because humans are not good as a integrating agent.
Examples of practical integration strategies that have been used successfully are web viewer EHR integration, single sign-on, RIS vs PACS driven workflow).
Advanced integration strategies, include using vendor APIs, state aggregation, SOA, and IHE.
Interopability is not just about single logon but should address the integration of the different software tools on a granular scale.
IT should move from choreography to orchestration. Instead of just providing a static edge environment, a more versatile system should be available which is loosely coupled using a middle (bus) layer and move towards a service oriented architecture. A component based architecture that supports composite applications and orchestrates these applications into complex workflows.
SOA is, according to dr. Chang, not going to work in healthcare although everyone else is doing it since healthcare IT groups will not make the effort to go for SOA. However, applications can be implemented to mimic a similar environment collecting and combining all available data in the hospital that can go into a dashboard. In this action, normalization of medical data is essential.

Image Sharing—A Fond Farewell to CDs

We are currently moving from sharing data on CDs to Sharing healthcare information in the cloud. Although it is solving part of the problems with CDs a major concern becomes security and confidentiality when moving to network sharing. Some proprietary enviroments are available but the industry should move to the available standards. The IHE XD profiles should be used when applicable.

RSNA 2012 - Leveraging Imaging Informatics to Improve RadiologyEducation: Beyond the Teaching File

Simulation is prominent in many specialties and jobs and should also play a more important role in radiology.

To make a Simulation two steps are fundamental:
1. Make a model of the system
2. Run the model over time

Simulation is done to practice situations that in real are difficult, expensive or dangerous to perform or to train and model new situations.
In the future Simulation has a potential in educational, training, evaluation and quality improvement settings within radiology. Many opportunities of simulation are apparent in radiology ranging from low to high fidelity simulation.

A lot of information is collected in training and daily work in radiology using Google. More dedicated search engines are much less frequently used (e.g. Yottalook or GoldMiner).
Radiology education could also leverage the PACS and RIS and thus use the material available in the hospital. RIS and PACS are tremendous resources of data to be used for systems like:

1. Report comparator (providing a list of reports showing both temporary and final report with tracked changes)
2. Quantitative reporting skills evaluation (computation of changes percentage in each report, should go down in time)
3. Discrepancy logger (the trainee will receive a mail from a attending radiologist based on a small form filled out after evaluation of the report)
4. call simulator (residents can look into the cases reported by the discrepancy logger as a learning tool)
5. Resident educational dashboard (a full access to different teaching tools with the ability to compare the own performance with peers)

In summary this sessions showed that IT technology can be implemented to improve radioogy education. This ranges from simulation in different ways to implementation of data mining and use of interactive devices such as clickers during classroom sessions. Interesting tools were presented to provide residents with information about thei performance and to support them in the evaluation of their own reports. Furthermore, the use of simulation in many different ways in radiology to increasae knowledge or awareness is also demonstrated as being an important topic for the future. The final presenter talked about how to introduce clickers and how you can increase the interaction by having group based learning by having students discuss their answers with their neighbors.

RSNA 2012 - Rapid Application Development with XIP™ - the eXtensible Imaging Platform

The eXtensible Imaging Platform (XIP™) is an open source framework supporting rapid development of imaging and visualization applications. Imaging software developers can utilize XIP's visual ‘drag-and-drop’ programming tool (the XIP Builder™) and associated libraries (the XIP Libraries™) in creating applications. In addition to functions from the popular ITK and VTK libraries, the XIP Libraries include modules tailored for medical imaging, many of which are hardware accelerated via GPU programming (e.g., OpenGL® GLSL or OpenCL or CUDA C). Applications created with XIP can either run standalone, or as DICOM Hosted Applications. Through the DICOM Application Hosting interfaces (DICOM WG-23), a Hosting System, such as the XIP Host™, relieves the application developer from the need to re-implement infrastructure common to all applications (e.g.DICOM network connectivity, database, etc.). Users can execute Hosted Applications, such as those created with the XIP Libraries, via the XIP Host.

URL's http://www.OpenXIP.org

RSNA 2012 - IT should be embraced by radiology

As stated by the cover story of today's RSNA daily bulletin, IT is required to become a value innovator in radiology. In future radiology, IT influence should not be restricted to RIS and PACS, but extend to embracing new developments in mobile computing, social networking and virtual collaboration.

3Gear kinect gesture based interaction system

Check out 3Gear systems for their approach to use two kinect cameras to track motion.

According to their website their technology enables the Kinect to reconstruct a finger-precise representation of what the hands are doing. This allows us to build simple and intuitive interactions that leverage small, comfortable gestures: pinching and small wrist movements instead of sweeping arm motions, for example.

RSNA 2012 - Radiology Informatics Series: Mobile Computing Devices

This early morning a whole session on Mobile Computing Devices. Apperent from the attention mobile devices get both from the industry and the RSNA itself, this is a very important topic that will greatly influence the future of radiology.


Mobile lessons from other failures are:

1. Make sure your app works very well

2. Native apps win hands-on over web-apps


Both healthcare professionals and patients will be demanding the use of mobile devices. Although ipads and iphones are used prominently in industry and hospitals, the consumer market is dominated by Android because of the much larger market share in smartphones of Android.

To keep up to pace, healthcare institutions should pay attention to mHealth and start developing apps for both their employees and customers.

VSIN31-03 • Comparative Efficacy of Intra-operative Interfaces for Interventional Radiologists: The Microsoft Kinect Device, Hillcrest Labs Loop Pointer and the Apple iPad

Cherng Chao et al


Intra opative image review is challenging and standard setups using keyboard and mouse are not optimal. This group compared the used of Microsoft Kinect, hillcrest labs loopp pointer, apple iPad for image interaction all using mimcloud software. Time, measurement and easy of use was determined for a prototypical task measuring a liver lesion. The Kinect software used a button approach to switch tasks. 29 radiologists participated in this study. The iPad was the fastest in performing the task while the Kinect was the slowest. However, sterility requirements hampers the use of the loop and the iPad. All three devices were believed to be usefull in the future. Windowing and measuring proved to be the most challenging tasks. Users with video game experience and experience with a particular interaction method showed a better performance.


Main shortcoming of this study was that the users using the Loop and iPad did not use gloves.

VSIN31-04 • Modular Design of a Mobile Web-app Clinical Decision Support and Communication Tool for Radiologists and Ordering Clinicians in a Multi-Facility Academic Medical Center

Loyrirk Temiyakarn , MD et al


This group used iwebkit 5 to build a modular webapp that provided all kind of information beneficial to different healthcare providers such as guidelines, phone lists, protocols, etc.

VSIN31-05 • Enhancing Utilization of Mobile Imaging: Assessing Variation in Diagnostic Accuracy of iPad with the Nature of Radiological Findings on Chest Radiographs

Supriya Gupta et al


30 cases following tube placement at ICU were evaluated on iPad and (after one week) on a PACS workstation. The conclusion is that the iPad still has to be evaluated for image quality and to determine that it can be used for a certain task. Image quality heavily depends on the app. DICOM calibration of iPad screens is becoming a critical issue and methods have to be developed specifically for tablets.

RSNA 2012 - The Use of Business Analytics for Improving Radiology Operations, Quality, and Clinical Performance

Data needed for business analytics for radiology is distributed over many different source systems (RIS, PACS, IHE, etc, etc). The data in those systems can provide vital information. However, to use this information we have to be able to gather and combine this information in an easy and fast manner.
Leveraging informatics tools such as business analytics can help a practice transform its service delivery to improve performance, productivity and quality. BA tools can be used for quality assurance, maintenance of certification (MOC) and pracitce quality improvement (PQI).
The general steps involved in business analytics, include extract, transform, load (ETL) and key performance indicators (KPI). BI involves: Aggregation, Integration (ETL), Storage (data warehouse), Analysis(OLAP, data mining), and Presentation. Relational databases are concerted to an OLAP configuration to allow fast data searching and analytics.
Before starting with Business Analytics, the goals and metrics needed to achieve those goals need to be defined.

Business intelligence maturity from least to most mature:
1. Standard reports
2. Ad hoc reports
3. Drilldown
4. Alerts
5. Statistical analysis
6. Forecasting
7. Pre modeling

- Business intelligence tools can leverage radiology systems' utility and empower executive decisions.
- The selection of the most appropriate KPIs are critical.

RNSA 2012 - Decision Support in Clinical Practice

Decision support systems use knowledge -- ranging from books, to web sites, to real-time artificial intelligence systems -- to help physicians improve their decision making.
Passive decision support leav responsibility to search the data with the radiologist while active decision support uses artificial intelligence and computer software to actively provide relevant information and guidance to the radiologist.
In the US a lot of the work in decision support in order entry is stimulated by the appropriateness criteria for radiological examinations.
Decision support systems have to be real time and integrated to be able to be a success in clinical practice.
Recipe for success of decision support depends on the system being:
- available
- easy to use
- integrated
- collaborative
- good medicine
- aware of limits

Predicting Diagnosis and Outcome
Currently there is a clearly identified role for informatics in predicting diagnosis and outcome based on variables derived from imaging. However, there are important trade-offs that exist when developing or using predictive models.
The goal of decision suppprt is to take predictive information and assist the physician in quantifying risk of disease or probability.
An important trade-off is that In most cases the correctness of the outcome of decision support heavily depends on the dataset used to construct or train the decision support system.

Quantitative Image Analysis for Image Retrieval, Decision Support, and Knowledge Discovery
Quantitative image analysis is to characterize images and or parts of images with rich features that can be accessed by computers for comparison or decision support.
Image characterization can be done by annotation and computing of image features. This characterization (among others based on quantitative image analysis) can be used for content based image retrieval by defining a vector of different features. The acknowledged features in this vector should be weighted to determine their relative importance. This weighting can either be done by human definition or by computer learning or training and depends on what you are interested in. In radiology this can be used to retrieve similar images from the PACS or, more challenging, find images with similar lesions. Standardization using e.g. RadLex and identifying exact locations is vital to obtain usefull desciptions to use as image features.

Key points
1) Evolving technologies provide new ways to integrate advanced decision support into routine clinical practice, and decision support systems can improve outcomes in patient care.
2) Decision support gives radiology a chance to transform to a more proactive role in the managing of patients and examinations.

RSNA 2012 - New and Improved Features and Functionality of the RSNAMIRC Teaching File Software(LL-INE1201)

New and Improved Features and Functionality of the RSNA MIRC Teaching File Software
William Weadock, MD , Sarah Abate, BS

Over the past few years, the RSNA MIRC system has undergone significant improvements in functionality. Installation, creation and display of teaching file, and conferencing have been significantly improved. The MIRC system has become the standard format for teaching file software. The software is free to download and use and is supported by the RSNA. This exhibit will show how the system can be incorporated in daily workflow.

Key inprovements in functionality shown at the demonstration at RSNA are the ability to further anonymize or de-identify the DICOM data not only by de-identification of DICOM fields but also by removing parts of the image that could possibly contain patient related information such as in ultrasound captures or secondary captures. Another feature is the ability to export directly to powerpoint of one or more cases to construct a presentation on a certain topic.

RSNA 2012 - ISP: Informatics (Image Sharing, Management and Display)

ISP: Informatics (Image Sharing, Management and Display)

SSA11-01 • 10:45 AM
Informatics Keynote Speaker: Image Sharing

Network based sharing, probably using cloud services, is mainly the replacement of the sneaker net (CD transfer) and comes with the same challenges. One has to be careful not to use tools that are used daily like Dropbox for patient related data sharing. Standards like XDR and XDM should be used to ensure safe transfer of patient data. Cloud services have to be standards based (IHE XDS.i) as implemented in the RSNA image sharing project. The exchange is based in a money transfer system where the patient gets an eight digit code and password. With these codes the patient can access the data and grant access to healthcare providers.

SSA11-02 • 10:55 AM
The RSNA’s Image Sharing Project: Initial Challenges from the Patients’ Perspective
Yousaf Awan et al

Traditionally patients are not involved in data sharing. The push to patient centered healthcare lead to the implementation of the RSNA image sharing project where the patient has control over his/her own data. Inability to access their images was the biggest issue with the current system looking at the helpdesk calls. In general, not many problems occurred.

SSA11-03 • 11:05 AM
Outcomes of an Automated System of Importation of Outside Imaging Studies into the Radiology Information System
Jonelle Marie Petscavage et al 

Patient data from CDs is uploaded decentrally and recorded in RIS and PACS. Sending AE title is set to 'OUTSIDE CD'. The error rate was less than 1%. interesting report, check our publications by van Ooijen et al in the Journal of Digital Imaging where we demonstrated something similar quite some years ago (free fulltext available)

SSA11-04 • 11:15 AM
Introducing Intuitive Simplicity in a New Type of Browser-based Dicom-Viewer for Non-Radiologists' Daily Clinical Routine
Sebastian Bickelhaupt et al

Many users are presented with non intuitive viewers that either have too much or too little functionality. The new viewer presented by this group is based on the body regions on an anatomic picture of a human. A model of human anatomy is given with colored dots on the model where the color indicates the modality and the datasets are shown next to it. After selecting the dataset of interest the DICOM data is shown and direclt connected to a 3D anatomical atlas. They showed that the new prototype speeds up and simplifies DICOM image viewing.

SSA11-05 • 11:25 AM
Influence of Medical Display System on Productivity and Eye-strain of Radiologists
Mony Weschler et al

This group compared setups with two 3MP screens to one 6MP screen both from Barco. Factors like eyestrain are evaluated using a questionnaire. It is show by this study that 6MP provides significantly less eyestrain and higher reading speeds. Although this is only based on manual registration of findings by the radiologists on a questionnaire.