Imaging in the palm of your hand: Phone-based DICOM viewers

Here are two really capable DICOM viewers, iOviyam and DWV, either of which can probably run on the phone you’re carrying right now, with no installs or software downloads.  They’re both up and live on our demo server (skip to the end of each review if you’re impatient to try them).

Both these programs let a standard smartphone query a PACS, and display the images that are retrieved.  How on earth does a cell phone handle the complexity of the heavyweight DICOM protocol?  It doesn’t.  As far as the phone is aware, it’s talking to a web server, which just happens to be serving medical-looking images.  These two programs both are deployed as a module within the industry-leading DCM4CHEE PACS server, and in fact installing each involves copying only one or two files, if you have such an installation running (no trivial task, but our wiki notes might help).

So for most people, this is imaging as a web service: something you connect to using whatever device is handy.  It provides a convenient mobile phone interface to a PACS, allowing querying by patient details and dates.  The DICOM images are retrieved from the PACS by the server module, then converted to standard web-technology images and sent to any HTML5-capable browser (anything from the last few years).  Your phone never knows it’s talking to a PACS.


iOviyam

IMG_3893
iOviyam2
 doesn’t take much longer to launch than it does to say.  It is the smartphone/tablet companion to the HTML5-based Oviyam2, and comes from the same productive team at Raster Images.  It’s a highly capable 2D viewer, with many of the features you’d look for in a desktop viewer.  Security is provided by the hosting PACS, so you need an account to connect to the server and view images.  Once logged in, there’s a nicely mobile-optimized search window, leading on to the image viewer. There are technical notes online in the dcm4che.org forum, which go in to more detail about the back-end architecture of the server module (there’s nothing at the client end).  iOviyam can use the web-oriented WADO protocol to talk to the PACS server, but can also act as a standard PACS client, and is configured into the server as such.  But the user is protected from all this, and just needs a login.

Try it right now!  The link below will route to iOviyam2 if you’re on a mobile device, and to the desktop-browser version of Oviyam2 if you’re on a computer.  You’ll need a login: it’s guest and the password is guest.  Easy.

Click here to launch the iOviyam viewer.


DWV – DICOM Web Viewer

DWV on iPhone

DWV is a new HTML image viewer from Yves Martelli.  It’s implemented in Javascript and HTML5, so it too will run on anything with a modern web browser.  It is deployed from a growing list of PACS servers, including ConQuest, DCM4CHEE, and the innovative up-and-comer, Orthanc.  DWV is launched from within the PACS search interface, rather than having its own search functionality as with iOviyam, and it can also be launched directly from a URL (the program’s website has some examples of this, as does our wiki entry for DWV).

The browser itself has the now-familiar subset of functions for the limited space available on a mobile device screen: pan, zoom, scroll, windowing and colour mapping.  DWV also offers a limited range of image filters, and the display of DICOM header values with real-time searching.  That’s a lot of functionality to fit into a small screen.

DWV is installed on our PACS server, and there are two ways of accessing it.  The first is to log in to the DCM4CHEE PACS server (username guest, password guest), and click ‘Search’ with all search fields empty (don’t worry about the warning, there are very few images to search).  Then, click the ‘eye’ icon to open the web viewer.  You can then choose between the installed web viewers (currently, DWV and Weasis).  Click on DWV, and you’ll get a new tab with the image volume loading.

dwv_in_dcm4chee_0

dwv_in_dcm4chee_1
                                       
The second way to open the DWV demo is just to click on one of the link images below.  Simpler, but less choice.


Click the images below to launch the DWV image viewer in your browser

1040_cardiac_mr_2d_150

Knee axial MR 23 sl

1010_04_ax_2d_0
                     
DWV’s chief developer is an imaging software professional and has provided all the resources that power users and developers would look for.  Source code is available at the project’s GitHub repository, along with extensive documentation of the code and the technical aspects of DICOM integration.  There’s a comprehensive wiki and a complete description of the development environment.  Yves has generously released this project under the GNU GPL and welcomes collaborators.  This would be an excellent opportunity to get involved in a cutting-edge imaging project that already has a solid background and a working product.

Web DICOM viewers are still a new technology, but one that is  evolving rapidly.  As we move from ‘computers’ to ‘devices’, the advantages of a zero-install workstation are so numerous and compelling that it seems clear that, for many needs, this is the immediate future of imaging.

ORS Viewer – the first FDA-approved free clinical workstation

ORSVisualSI_bug_largeFrom Object Research Systems in Monteal comes ORS Viewer Lite, a slightly reduced version of their flagship medical imaging workstation.  This is a serious workstation from a commercial company, and is the first free imaging workstation to be approved by the US FDA, Health Canada, and CE Mark – a significant step forward in free medical imaging software.  Gaining these approvals is a lengthy and expensive project, and Object Research Systems should be applauded for releasing this high-level workstation to the free  software movement.  Naturally, they support themselves by selling and supporting the full version of this program, which includes 3D reconstruction, volume rendering, endoscopy, and advanced modality-specific plugins.  But the Lite version lacks only these advanced features, it is not restricted in any way (though you’ll have to buy the full version to do the cool 3D beetle rendering).

As a 2D workstation, ORS Visual Lite has plenty of capabilities and can handle multiple sequences with image fusion and linked cursors.  The configuration options are particularly extensive, and include advanced features such as customisable key shortcuts for many functions, and saveable custom anonymization rule groups.  The program is fully PACS-capable, adding remote nodes is straightforward, as is query/retrieve, and send.  Processed scans can be exported to a removable media or remote PACS destination.  The clinical emphasis of this application is apparent with its support of user accounts and login, MPR and MIP views, text and graphical annotations, and DICOM print and report.

ORS Visual Lite is an excellent product, and its release is a good day for the free software community.  In a crowded PACS workstation marketplace, Object Research Systems deserve to be rewarded for being the first to make this bold step.

Web-based DICOM viewers – read about them, try them!

Medical imaging is a leader in the trend of applications moving to the browser. Browser-based imaging applications use a variety of web technologies, both new (such as HTML5, just a few years old) and longer-established (Java, which has been around since the mid 90s). The availability of new web technologies and  high-level development tools has made it increasingly easy to build and deploy web apps that can approach the power and versatility of installed applications, with little or no client preparation required. Here, we’ll look at two image viewers affiliated with the dcm4che project, and deployed on top of the DCM4CHEE server, that take different approaches to getting images from the confines of a complex PACS server, and into a common web browser.

Weasis

Weasis is a Java Web Start application developed by Nicolas Roduit. Java Web Start comes as part of the free Java Runtime Environment that most personal computers have installed. It is activated when a JNLP file is downloaded in your browser (usually by clicking on a link), and it downloads and installs the Java application linked to this file. The application is stored on the computer, so subsequent launches can omit the download step unless the application is out of date. One of the key terms here is ‘computer’ – downloaded Java apps can’t be run on tablets (Android or iOS) for legal and technical reasons. JWS traces its history to around 2002 and accordingly, is built around computers rather than portable devices.

Weasis does not require any preparation by the user beyond the presence of Java. It’s deployed as part of a PACS installation, primarily the widely-used open source DCM4CHEE. Communications between Weasis and the PACS server is by means of  WADO, a DICOM standard designed to facilitate image viewing over the Web. The installation of Weasis on the server is not difficult, but does depend upon a rather complex PACS server, so we’ll concentrate on the viewer itself.

Clicking on a Weasis launch icon downloads the JNLP file.  The downloaded application may start automatically, but increasingly, this association is disabled for security reasons, so the downloaded file must be located and clicked to start the application (see the wiki notes).  Also, running the application for the first time will present a series of popup windows to authorise the application (a familiar part of life with Java applications).

Weasis launches as an application independent of the web browser and in fact from this point may subsequently be run as an installed program. Data comes along with the download or may be read from disk like any other imaging application. It’s a fairly standard 2D image viewer with no reslicing or volume rendering, but it runs quickly and provides comprehensive controls for adjusting brightness, contrast, scale, and colour map. There are tools for measuring distances and areas, and multiple studies can be displayed in a split screen. Weasis doesn’t provide all the functionality of a heavier viewer, but from a correctly-configured site, you can be viewing images within a minute of clicking a link, no installation required.

Try it out!  Click on one of the images below to download Weasis along with the data set.  The second and subsequent downloads are faster, as they don’t have to download the viewer.

1010_06_ax_2d_1_150

1010_02_ax_2d_0_150

1030_knee_mr_2d_150

1040_cardiac_mr_2d_150

There are other links to images and the Weasis viewer from the Sample Data page on our technical wiki.

There are extensive installation notes on the dcm4che site, which also has an active forum.  If you’d like to install Weasis on an image server, take advantage of I Do Imaging’s wiki notes, based upon the experience of installing this software several times on different platforms. They may save you hours of work.

Oviyam

Here’s another approach to browser-based viewing. Oviyam is entirely developed in HTML5, a modern Web technology that greatly enhances the capabilities of a web browser. In this approach, Oviyam is installed as a component of a PACS server and performs all the DICOM communications within the server, to which Oviyam appears as a standard PACS node. On the network side of the image server, Oviyam acts similarly to web server, sending image data to the browser in standard web formats. So a modern HTML5-capable browser can display the images without requiring any DICOM or PACS functionality.

Installing Oviyam, then, is strictly a server-side operation. And on a properly-configured DCM4CHee installation, it’s as simple as copying a .war file into the server’s directory structure. The client then visits a web location (port 8080, the default port of the JBoss framework underlying DCM4CHEE), and never knows it’s accessing a PACS. Oviyam provides all the basic functionality of a 2D browser, with no installation of any kind on the client end. It even works on portable devices: iOviyam provides pure HTML5 to a compatible web browser.

Try it out! Oviyam2 is installed on our public PACS server. Login is ‘guest’ with the password ‘guest’.  Click on the ‘idoimaging’ tab, then on ‘Sample Data’ at the top (or just ‘Search’), to bring up a list of studies.

Click here…
oviyam_data_annotated
…to get this.
oviyam_viewer