PACS
Desktop Viewer
PACS
Desktop Viewer
Presentation states containing text strings using alternative character sets, ie, not latin-1, or special font characteristics, such as superscript, were presented incorrectly because the viewer’s DCMTK tools did not support alternative character sets. The DCMTK toolkit has been updated.
When searching for commands in the toolbar customization panel, matching values can show up as empty strings or the scroll bar might be missing, until the user forces the GUI to refresh. This has been resolved.
When editing the toolbar configuration, the new line button was improperly disabled for context menus, preventing users from adding new entries to context menus. This has been resolved.
For some scanning devices, the document scanning tool omitted a required property in the command and only the first page would be scanned. This has been resolved.
When initializing the caching service, before the viewer is running and the cache size has been defined by the configuration, the service could delete studies if the total free space on the workstation exceeds the size of the next study being initialized. This has been resolved.
Automatically splitting very large series could result in a viewer hang. This has been resolved.
A notification indicating image and presentation state data was sent to the server was not displayed when the data was a single, post-processed image. This has been resolved.
When deleting post-processed data that are in use in image frames, the user is notified and requires confirmation from the user. This notification appeared for every image, so deleting a series resulted in multiple notifications and confirmations.
When editing the toolbar configuration, editing a group containing a separator by moving an item to a position failed to handle the separator object correctly, resulting in a viewer crash. This has been resolved.
After adding a command to a group in the toolbar manager and saving it, the tool appeared in the wrong location by default because its position was based on the applied filter, not the complete list. This has been resolved.
If the viewer was unable to connect to the configured prefetch server, it ignored the error and as a result, tried again, which might exceed the retry count and eventually freeze. This has been resolved.
Selecting a VOI LUT function from the selection list cleared the VOI LUT presets in the list, resulting in an empty preset list. This has been resolved.
After clearing a manually-defined spatial registration mapping between two series in which one of them shares a frame of reference with a third series, the viewer could crash because it improperly released the frame of reference relationship data. This has been resolved.
When a user is logged into multiple servers at the same time and the streaming connection to one of them is unavailable (for example, the port is blocked), the viewer released the unavailable connection but crashed as it mishandled the resource when unlocking it. This has been resolved.
The length of the thumbnail toolbar had an unnecessary maximum size limiting the number of series displayed. This limit has been eliminated.
CAD SR objects missing a Modality field value caused the viewer to crash because it didn’t check for an empty value. This has been resolved.
The manual registration tool could incorrectly apply the cross correlation marker (ie, Magic X) because it might use the wrong position conversion function. This has been resolved.
Scanning a color image failed to set the default window and level values, allowing the viewer to apply zero values and display a white image. This has been resolved.
After a recent refactoring, the ROI with W/L tool failed to update the handling the deletion of some data, resulting in a viewer crash. This has been resolved.
When deleting an image or series the user created (exported) from a plugin module, the viewer did not clear the image/series from image and plugin frames, resulting in a viewer crash when the user selected a frame containing the purged data. The viewer detects frames containing the image/series and notifies the user to clear them before invoking the delete function. This has been resolved.
Images containing embedded overlays exported from the 3D or Fusion plugin modules would include artifacts (white lines) in the images because the overlay data was mishandled in the generated image. This has been resolved.
Multiple current studies were misidentified because the viewer failed to read the primary status information for all but the first study, labeling the first as current but the others as priors. This has been resolved.
Attempting to append a list of prefetch servers to the Settings Apps’ settings overwrote the existing list because the append code was disabled. This has been resolved.
When a very small rescale slope is applied to a pixel value and results in a value less than the minimum window size (i.e., less than 1), the image’s initial view was washed out and adjusting the window and level values had no effect. This has been resolved.
Closing the DICOM information panel failed to release the panel’s resources. This has been resolved.
When the DICOM information panel displays sequence data containing multiple items, the data from the first item was displayed for each item. This has been resolved.
When the DICOM information panel displays sequence data containing two or more levels, the contents of the sub-sequence items were empty because the viewer miscalculated the first item in the sequence. This condition existed for and affected all multi-level sub-lists, including drop down lists, options menus, etc. This has been resolved.
After exporting an image/series from a plugin module, loading it into a new plugin frame, exporting it again and then deleting the image/series, the viewer could hang because the first exported instance was not properly removed from the viewer. This has been resolved.
Support for multiple detector data with different image orientation vectors was missing, resulting in the orientation information from the first detector being applied to images from all detectors. This has been resolved.
An error message appears when the user attempts to restore a bookmarked study after the study was deleted from the system.
If the image count exceeded 1,000, the stack ruler’s position number value was too big to fit in the text box. This has been resolved.
DICOM media that does not include a CD viewer does not include the herpa data file. CD viewers attempting to open studies from local or removable disk without the herpa data file would not transition to the state needed to load DICOM files directly, resulting in a failure to load any images. This has been resolved.
If the user closes the settings application before it completes the request to clear the cache, the service could crash in the callback. This has been resolved.
Users were unable to turn Dragon recording off if they used the Record toolbar button, rather than the Speechmike, to turn it on. This has been resolved.
When closing the viewer’s context panel after selecting an item from a dropdown list on the context panel, Windows didn’t always select the viewer as the foreground application, which might move the viewer to the background. This has been resolved.
Importing service settings created an extra, uninitialized prefetch server entry which caused the service application to crash when it attempted to parse it. This has been resolved.
A change when adding the modality-specific context menus caused the toolbar manager to save the command group list before it applied the updated order. As a result, the order could not be modified. This has been resolved.
If one Windows user started a viewer session, it locked access to the crash log. If another Windows user attempted to start a viewer session on the same workstation, it stopped loading waiting for the first session to release access to the crash log. This has been resolved.
For very noisy images in which the starting point value change is high, the skin line detection algorithm failed to find a good initialization point, causing the skin line to fail. This has been resolved.
An image refresh issue caused annotations and other renderings applied to part of an inverted grayscale image to appear with a black, empty background rather than the image data itself. This has been resolved.
Users were able to activate the viewer from the system tray when the viewer’s About panel was displayed. Since the About panel is modal, activating the viewer on top of it prevented the user from closing it, making the viewer appear frozen. This has been resolved.
When a mammography image was missing a laterality value, which is uncommon but can happen in an edited image object, the viewer miscalculated an offset, causing the viewer to access invalid memory and crash. This has been resolved.
Splitting a series could cause thumbnail grouping to reference an uninitialized structure, causing the viewer to crash. This has been resolved.
ROI annotations were mishandled, resulting in duplicate mouse handlers and missed updates when the image frame changed (eg, a resized frame). As a result, ROI annotations were not always updated and sometimes caused the viewer to crash. This has been resolved.
If the server name is long, it can obscure the override port displayed in the streaming server status panel. The panel is now sized based on its content.
The control for displaying image orientation information has been extended to three states: hide orientation information; show orientation information based on modality; and always show orientation information.
The keyboard shortcut control panel didn’t recognize all key-down events on the Alt key, preventing users from assigning the key to a shortcut. An alternate method for recognizing the Alt key’s state has been applied.
The system attempts to sort all series based on the body part but when the images in a series are defined to be in the same position and orientation, as is the case with some ultrasound series, the sort is unnecessary and could result in unintended re-ordering. These series or subsets of a series are left in their original order.
A free space calculation error caused the cache service to purge some data prematurely after workstation reboot. This has been resolved.
When the cache service starts, it could pack the message buffer causing some messages to fail. These failed messages were not retried, causing the viewer to report a lost connection to the cache manager. This has been resolved.
Restoring a bookmarked session containing an updated study (e.g., it received a new image) after it was saved might reload the pixel spacing data in the wrong order, resulting in erroneous measurement values. This has been resolved.
Images using a SIGMOID VOI LUT function and defined with a MONOCHROME1 photometric interpretation value defaulted to an inverted grayscale image. This has been resolved.
When handling a routine network error, the viewer didn’t check for the instance of the network class. When it didn’t exist, the viewer crashed. This has been resolved.
If the viewer’s streaming setting changed after the server started creating the gw3 files, the viewer might request the other file format causing file size inconsistencies that could lead to a crash. This has been resolved.
The load progression indicator in the study header failed to update because the viewer was checking for the wrong state. This has been resolved.
When no frame of reference was defined, drawing a 3D annotation across multiple images resulted in a crash. This has been resolved.
When loading a hanging protocol, a missing class instance was not checked and when it didn’t exist, the viewer crashed. This has been resolved.
Refactoring applied for 3D annotations caused the viewer to require a spatial registration license for permitting cross correlation of studies declaring multiple frames of reference. This has been resolved.
Due to a misinterpretation of DICOM, the viewer applied the referenced image sequence of spatial registration objects inappropriately causing a failure to correlate frames of reference. This has been resolved.
When drawing a measurement annotation close to the image frame border, the value’s display location is forced to fall inside the image frame.
If data has been downloaded to the workstation cache using cw1 file format and then updated in the cache with cw3 file format, the viewer failed to reallocate the cache space, which might cause the viewer to crash. This has been resolved.
The certificate checker didn’t ignore non-certificate errors, such as time outs, and when one occurred, reported warnings with no details. This has been resolved.
A change to add support for a specific VOI LUT function caused pixel data to be interpreted literally rather than normalized for display. When the pixel data had been encoded using an inverted scaling factor, the image appeared incorrectly on the screen with inverted W/L values. This has been resolved.
If an object declared the image displays using a LINEAR_EXACT VOI LUT function, window and leveling failed to work because the viewer already overwrote the user-specified values with the full range values. This has been resolved.
In some conditions, rotating a mammography image 90 degrees inverted the stack ruler’s direction. This has been resolved.
A change in build 70 introduced a refresh problem which failed to display images exported from a plugin module in the thumbnail panel. This has been resolved.
Image reconstruction failed an internal consistency check when the resampled image had a dimension less than that used by the thumbnail images (64 pixels). This has been resolved.
A timing issue could cause the viewer to read the next or previous study URL before it was ready, resulting in an invalid study error message. This has been resolved.
If the primary monitor number is greater than the number of available monitors, which can happen after disconnecting monitors after assigning the primary monitor, no primary monitor is defined and some of the viewer’s layout functions fail. In this state, the viewer automatically reassigned the primary number to the highest available monitor number.
When magic glass linking is active and multiple magic glass panels appear over each linked image frame, unlinking an image frame left the linked magic glass panel visible and unresponsive.
The shortcut manager panel labeled the Delete button (for user-defined macros) incorrectly. This has been resolved.
When configured to use hyper streaming, the timer used to refresh the IQ image download percentage was not stopped once the download completed.
An image arithmetic optimization applied to lossy images was incorrectly omitted when the software was merged with the server’s equivalent code, causing slower decompression speeds. This has been resolved.
Some special characters used in names and rendered by the viewer, primarily on tabs and in the DICOM information panel, were displayed incorrectly. This has been resolved.
The configuration panel listing the available reporting options did not identify which option was currently applied. The configuration panel now lists all available options, their applied trigger event(s), if any, and other details.
If the print panel is displayed before the study starts to load, the viewer would crash. This has been resolved.
If the viewer restores a bookmarked session after the current session had been minimized and then maximized, the restored session would initialize in a minimized state. This has been resolved.
If the viewer downloaded a report object from the server between the time the user started dragging a presentation state from the thumbnail panel and dropped it in an image frame, the viewer could crash. This has been resolved.
The viewer could mismanage the handling of the patient folder thread because it was not made thread-safe. This has been resolved.
If the user closes the viewer and starts exiting Windows before it completes, the viewer could crash because Windows might have already terminated the frame. This has been resolved.
If the text used to populate the thumbnail header is long, it might not be properly truncated and marked as such if the full study information has not been downloaded yet. This has been resolved.
When a profile setting was deleted, the viewer might use the wrong setting ID which would remove the wrong setting and if the setting was later added cause the viewer to crash. This has been resolved.
Keep-alive status messages were processed by a transient thread rather than the main thread. Depending on the timing, this thread might not exist, causing the viewer to crash when it attempted to process the status message. This has been resolved.
The size of the tool to resize the magic glass panel was unintentionally modified to a very small area. It has been restored. Additionally, a failure to catch a right click event displayed an invalid context panel that would not clear when the user attempted to close it. This has been resolved.
Based on timing, if the viewer attempts to initialize the patient folder’s tab title while the session is closing, it might access an empty structure and crash. This has been resolved.
The session information data might contain information about prior studies that don’t actually exist in the pbs file. If the viewer attempted to handle these studies, it would crash. This has been resolved.
If the viewer attempted to reload a study, due to a refresh command or from a bookmarked state, after the session has some terminating effect, such as a timeout, the viewer might crash if the study was, technically, no longer available for loading. This might also result in a message indicating the report was empty. This has been resolved.
When the window/level tool was applied to an image frame in tile mode, the starting image was not the first one in the tiled frame, and the user scrolled to the end of the images so an empty tile frame appeared under the cursor, the viewer could crash. This has been resolved.
The viewer didn’t check the state of an image when it received a download progress update. If the image was no longer available, eg, the study was being closed, the viewer might crash. This has been resolved.
Some modal (ie, blocking) popup panels could be closed through unsupported keystrokes or mouse clicks without clearing the modal state, leaving the viewer disabled and greyed-out. This has been resolved.
Some WMI log entries contained the wrong transaction ID, specifically calls that resulted in a dialog panel. This has been resolved.
If the viewer session times out while a print panel is open, the viewer would crash when the user returns and logs in again. This has been resolved.
Transient prefetch errors resulted in a popup error message, interrupting the user’s work. These errors are logged instead and not displayed in popup windows.
In a multi-server dotcom environment in which the viewer uses streaming to download data from multiple servers at the same time, a closed thread releases a shared handle but in some cases it was done in the wrong order, causing the viewer to hang. This has been resolved.
If additional compressed data arrives at the same time the last data decompression task finishes, the viewer could mishandle the release of the decompression tasks causing the viewer to crash. This has been resolved.
A change in viewer-8.0.30 created a second session ID cookie but failed to recognize more than one. When using HTTPS, if the user logged out of the server without closing the viewer session, the viewer reported a session error when opening the next study. This has been resolved.
Calculating the correlating position used the wrong frame of reference direction and in cases where the matrix was different, the correlating positions were incorrect. This has been resolved.
When hybrid streaming from multiple servers, the stream scheduler mismanaged the task list and used the wrong connection to complete a specific task, causing the viewer to crash. This has been released.
Details about specific SSL errors are saved in the log file to aid in debug investigations. These details have been removed from the message box displayed on the user interface and replaced with the specific error code.
The viewer could crash when closing if the structure identifying a concurrent session was NULL. This has been resolved.
When starting a progression with the previous view tool, progressing with the next view and then again going back using the previous view, the progression state was mishandled and the previous progression view was not displayed. This has been resolved.
If the web viewer receives a command to close the session while it is still initializing the patient folder panel, it might crash. This has been resolved.
If a report is received after the viewer starts terminating the session, the viewer might crash. This has been resolved.
If a second web control request to open the patient folder panel is issued before the first completes, the viewer could crash. This has been resolved.
The viewer defaults to the last-used location when uploading data from DICOM media. If that location was on a removable drive, and the current upload is not using the same drive, the viewer would hang until Windows timed out mounting the same drive. Now, the viewer will not save a removable drive as the last-used location, causing the viewer to pop up a browse window instead.
The thumbnail panel slider would not move across monitors because of a missing recalculation after updating the thumbnail panel. This has been resolved.
After a downed or slow network delayed the delivery of a response to append additional data to the viewer session, the user terminated the session. When the next session started, the viewer received the request’s results but didn’t verify the session identifier, causing the viewer to crash when it attempted to apply the data. This has been resolved.
A retired javascript call was not completely removed, allowing the viewer to make two javascript calls to the patient folder at the same time, resulting in a viewer crash. This has been resolved.
The viewer mishandled the task message reference counter, allowing a second task to interrupt a running task, which isn’t supported, resulting in a viewer crash. This has been resolved.
Multiple close requests were processed at the same time because the viewer didn’t check the state before processing them. Releasing resources that had already been released caused the viewer to crash. This has been resolved.
A request for information about a view state arrived after the view was closed. The viewer failed to check the view’s state and crashed when attempting to access information that no longer existed. This has been resolved.
When applying an overlay, the viewer failed to verify the resulting bitmap. If it was invalid, drawing it resulted in a viewer crash. This has been resolved.
Multiple login requests issued at the same time from the same WMI client interrupt the viewer’s login request to the server. The viewer mishandled deleting the interrupted request, resulting in a viewer crash. This has been resolved.
It’s possible, under some undetermined sequence of events, the same magic glass object can be created twice. When the viewer session terminates, releasing it twice causes the viewer to crash. This has been resolved.
When advancing the displayed study using the next/previous study tool, the viewer failed to check for images. If the next/previous study contained only attachments, reports or other non-image objects, the viewer would crash. This has been resolved.
When loading multiple studies from a multi-level dotcom into a full workstation cache, causing the makespace function to clear data, and some objects in the last study were already in cache but others not, the viewer could crash. This has been resolved.
The viewer’s system tray icon failed to indicate it was active or downloading data after the icon object’s ID was changed in a receive viewer update. This has been resolved.
A recent change to the location of the net dump log files was not applied to the upload function, and the viewer was unaware log files existed and needed to be sent to the server. Additionally, net dump logs are created when a network issue occurs, even when net dump log creation is disabled.
Snapshot and crash logs that failed to upload to a server were marked as sent, preventing them from being uploaded when the server could receive them. Additionally, all service snapshot logs are uploaded to the server.
The Hologic ImageChecker entry in the list of supported CAD objects could not be moved from its priority position due to list mismanagement. This has been resolved.
The viewer failed to display the notice informing a user that they lost the study lock because of a formatting error in the message the server sent to the viewer. This has been resolved.
Viewer macros consisting of other user macros could cause the viewer to hang when closing due to the order in which these macro resources are released. This has been resolved.
The error check used in the iterative process detecting anatomy from background was updated to increase the success rate of skin line detection. Also, an experimental algorithm detecting a point on the skin line is available. It is disabled by default. Enable it from the Enhanced Smart Fit Detection setting. See Zoom
Parsing a CAD SR object’s sequences could miss sequence items when they are unexpected, causing the viewer to ignore the data. This has been resolved.
The visible part of images whose height far exceeds their width were misrepresented in the thumbnail panel due to a calculation error involving the padding area applied in the image frame. This has been resolved.
Converting stream block sizes from alphanumeric strings to integer values used the incorrect integer type size, although block sizes don’t exceed the type’s limit. This has been resolved.
When settings on the Options panel are locked, some numerical fields appeared enabled for editing when the mouse hovered over them, resulting in confusion. This has been resolved.
When terminating a viewer session, the Speechmike control window could be closed by the viewer or Windows. If closed by Windows first, the viewer would attempt to close it again, causing the viewer to crash. This has been resolved.
When the task that manages rendering data to the screen was issued a refresh request, but the image frame being rendered is destroyed before the refresh starts, the viewer could crash. This has been resolved.
Old-format user profiles used as the default profile on DICOM media omitted the auto-orientation default settings for CT studies. This has been resolved.
Some messages issued between the patient folder and main viewer threads were not handled well. Under certain conditions, it is believed these messages could cause the viewer to crash. This has been resolved.
When streaming compressed data to the viewer, updating the priority of a decompression task failed to lock the task, allowing the task to release some data before the priority was set, causing the viewer to crash. This has been resolved.
When terminating a viewer session, a viewer configured to notify a third party application by writing an XML file to the disk could crash if the report was released before the file was written. This has been resolved.
When a partially downloaded study is stopped and then restarted using a different streaming method, the viewer could mismanage the decompression pump, causing the viewer to crash. This has been resolved.
If the user clicks to launch a viewer session faster than the request can be communicated from server to viewer, the viewer can replicate a control file’s name causing one file used by two sessions, resulting in a viewer crash. This has been resolved.
If the user bookmarks a study before the session initializes by switching tabs in the RIS worklist, the viewer would hang. This has been resolved.
When terminating streaming tasks, the viewer failed to check for batched task operations. Those tasks could attempt to run after the command to stop has been issued, causing the viewer to crash. This has been resolved.
When the viewer found the cache file size didn’t match the real file size, it attempted to resize the cache file but failed to lock the resource, resulting in a deadlock. This has been resolved.
A series of issues related to a change intended to improve streaming performance caused messages (reporting the download completion status) to get processed in the wrong order. As a result, image displaying would hang or the viewer would crash. This has been resolved.
DEPENDENCY NOTICE: Desktop Apps 8.0.27.1 (in EPConnect.exe 8.0.14) is required.
If a local cached object was compressed, the viewer starts a task to decompress it. If the file was re- downloaded with a different compression format before the decompression task started, the task would attempt to access the original object file and crash. This has been resolved.
The skin line detection algorithm attempted to identify the image orientation based on the image data. When paddles are present, the orientation determination could fail. The algorithm now uses the DICOM data to determine the breast orientation.
CAD objects defining polygons with more than 500 points were displayed but not in their entirety. This has been resolved.
Some special characters in patient demographic values were improperly encoded in commands sent to the server. The necessary encoding is now applied.
The viewer was not releasing memory in the following cases: when a CAD object is applied to multiple tomo images; when a compressed image download was interrupted during decompression; when releasing thumbnail panel resources after exiting the viewer. This has been resolved.
After initiating the shutdown procedure but before completing, Windows could terminate the viewer causing services (e.g., the cache manager) to get out-of-sync. This has been resolved.
Improper thread handling in the thumbnail panel manager could result in a viewer crash when loading studies into a viewer session. This has been resolved.
The viewer could fail to release the mapping object, preventing the cache manager from reallocating mapped memory when necessary and causing it to crash. This has been resolved.
Launching a study into the viewer from DICOM media processed the PbR file twice, resulting in two panels for each note entry, if present, and could prevent the viewer from closing. This has been resolved.
Acknowledging the prompt to save a presentation state put the viewer window in the background. This has been resolved.
Hybrid streaming, which is optimized for compressed data transfers, was deferring to the auto-detection setting, resulting in the less-efficient uncompressed data transfer when downloading images to the viewer. This has been resolved.
CAD headers were not displayed on cross-referenced 2D mammography images if no findings existed in the corresponding 3D (tomo) CAD object. This has been resolved.
Multiple threads attempted to manage the patient folder window, resulting in an exception causing the viewer to crash when updating the patient folder. This has been resolved.
The scaling factor used to size text, icons, etc., stopped working after a change to apply profile settings to the CD viewer. This has been resolved.
To speed image loading in the CD viewer, processed data is stored on the CD so the viewer doesn’t have to process it on-demand. This processed data contained an incorrect byte count for some image objects. As a result, the image(s) failed to load in the CD viewer. This has been resolved.
The cross-referenced CAD marker tool had priority over annotation tools, making it impossible to start an annotation on top of a (cross-referenced) CAD marker. This has been resolved.
A server side enhancement to reprocess images allows the viewer to notify the server when images should be reprocessed. The command is sent from the viewer. This setting was mistakenly added to the presentation state and key image descriptions, causing the annotation references to not match and be missing from PS and key image views. This has been resolved.
Newly created hanging protocols forced the Auto Load Localizer setting to disabled for all pages except the first page. This has been resolved.
When disconnecting an active stream, the viewer attempted to access an object after the thread released it, resulting in a viewer crash. This has been resolved.
The task responsible for freeing up disk space when available disk space dipped below its low mark restarted itself unnecessarily. This has been resolved.
Some residual cleanup was needed to remove a retired monitor calibration feature.
Toggling the CAD’s show/hide state did not activate/deactivate the cross-correlation tool between 2D and tomographic mammography images. This has been resolved.
The reordering of configuration parsing applied in a recent plugin enhancement resulted in the viewer ignoring plugin activation shortcuts. This has been resolved.
The cross referencing feature providing a shortcut from 2D (FFDM, synthetic) images to BTO series only applies when all images are defined with the same size and aspect ratio.
User profiles configured with specific viewer settings (specifically, those associated with hybrid streaming) and copied to DICOM media would cause the CD viewer to crash when loading a study. This has been resolved.
The CD viewer failed to ignore the print layouts in the user profile copied onto the DICOM media. When initializing the print panel in the CD viewer, only the Windows print device is recognized and offered to the user.
Some user interface settings defined in the user profile copied to DICOM media were applied after initializing the viewer, and therefore did not take effect until opening and closing the settings configuration panel. This has been resolved.
Older Hologic/R2 CAD objects contained incomplete Malc data that the viewer attempted to render on images. Malc data information is now limited to Hologic Genius AI CAD objects.
DEPENDENCY NOTICE: This fix requires medsrv-8.0.20.
Although the file successfully uploaded, an invalid error message was logged when uploading a document from a worklist server that was redirected to a study on a hub server. This has been resolved.
The timeout period for completing a study (media) upload was too short for large (>2GB) objects, causing the upload process to fail. The timeout has been increased.
The viewer did not support the plugin module’s parameter storage functions, causing it to ignore requests to save plugin settings. This has been resolved.
Converting the media upload function to an independent application (based on the viewer application) in build 32 introduced a bug causing the CD viewer to not recognize the report object when the user tried to display it. This has been resolved.
If the position indicator on the stack ruler crosses the mouse cursor while the user is scrolling through the series using the mouse wheel, scrolling would stop because the mouse wheel event was ignored. This has been resolved.
If the user starts dictating a report before the viewer downloaded or initialized the report object, the viewer notifies the user dictation is unavailable. When the viewer is ready for dictation, the notification disappears and the user may proceed.
Support for new CAD data added an annotation type that was not handled properly by existing annotation tools, including measurements and ROIs. This has been resolved.
Resizing overlapped, mirrored Magic Glass panels can cause the viewer to crash. This has been resolved.
A keyboard shortcut created to open the Magic Glass panel opens the first panel for an image frame but not additional panels. This has been resolved.
Loading an image frame from the thumbnail panel while a Magic Glass panel is visible for that frame would result in a viewer crash when the user closes the Magic Glass panel. This has been resolved.
A recently updated memory function requires a parameter that was previously optional. Instances where this parameter was not provided have been corrected.
Multiframe objects are included in the image count when checking the split series length, but are treated as a single image object when splitting, meaning multiframe objects will not be separated into individual series in the thumbnail panel.
When the thumbnail panel is docked along the left side of the window on a multiple monitor workstation, scrolling the thumbnail panel would eventually change from a vertical scroll to a horizontal scroll, corrupting the screen layout.
When opening a viewer session with multiple studies, at least one of which is in the Final state (ie, a prior study), the primary study's report panel's tabs could be mislabeled until the user manually selects one of the tabs. This has been resolved.
If the viewer session closes before the study pump releases all the thumbnail references, the service component could crash resulting in a lost connection message and no data downloads. This has been resolved.
When the scanning operation completed, the scanning application panel was not closed automatically. This has been resolved.
Spatial registration objects containing references to three or more frames of reference were linked incorrectly, resulting to a viewer crash applied. This has been resolved.
When appending studies with CAD objects to an active viewer session, the viewer appended CAD objects from all studies to the thumbnail panel, duplicating those that already existed. This has been resolved.
Mammography images linked using Smart Link could get out of sync when using the progression tool, especially if one of the linked images failed to identify the anatomical structure in the image. This has been resolved.
Altering a linear or angular measurement on an image whose orientation has been modified from the default (by the auto-orientation feature) could cause the measurement's text box dimensions to become negative, resulting in lines left on the screen and possibly a viewer crash. This has been resolved.
Using a name component (eg, Family Name) multiple times in the name format definition could result in an overrun buffer, causing the viewer to crash. This has been resolved.
Inverting the greyscale on a mammography image before it completely decompresses initiated recursive calls to the skin line detection function, causing a stack overflow and a viewer crash. This has been resolved.
Two CAD finding ROIs that shared a common edge would appear overlapped due to a calculation error when drawing the double-line ROI border. This has been resolved.
If the mouse cursor was positioned directly over an annotation ROI or line graphic, the mouse wheel scroll feature would not activate. This has been resolved.
The spine label plane setting displayed under the Settings tab on the popup panel was not recognized, resulting in the inability to enable or disable the feature when applying spine labels.
After adding center points to ROI annotations, the center point was incorrectly applied to mammo CAD ROI annotations. This has been resolved.
When closing a viewer session containing unsent presentation states intended to be uploaded as secondary capture objects, the close panel includes the option to send them. This setting was not active, meaning changes made in the close panel were ignored. This has been resolved.
When a series is auto-oriented in the axial direction, the order direction could be misassigned,resulting in the wrong sort order. This has been resolved.
If the mouse hovers over the frame border while scrolling via the left-mouse drag tool, scrolling would stop. This has been resolved.