• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

MAUDE Adverse Event Report: SYNAPTIVE MEDICAL INC. CLEARCANVAS RIS/PACS; SYSTEM, IMAGE PROCESSING, RADIOLOGICAL

  • Print
  • Share
  • E-mail
-
Super Search Devices@FDA
510(k) | DeNovo | Registration & Listing | Adverse Events | Recalls | PMA | HDE | Classification | Standards
CFR Title 21 | Radiation-Emitting Products | X-Ray Assembler | Medsun Reports | CLIA | TPLC
 

SYNAPTIVE MEDICAL INC. CLEARCANVAS RIS/PACS; SYSTEM, IMAGE PROCESSING, RADIOLOGICAL Back to Search Results
Model Number PERSONAL EDITION WORKSTATION 13.1
Device Problem Image Display Error/Artifact (1304)
Patient Problem No Patient Involvement (2645)
Event Date 03/28/2017
Event Type  malfunction  
Manufacturer Narrative
On (b)(6) 2017, the customer sent sample anonymized images to synpative rep.Older images appeared correct (jpeg lossless).The problematic ones come from this year and are jpeg 2000 compressed.The customer found there are no data luts (look up tables) encoded into the images.All images have window center ((b)(4)) and window width ((b)(4)) values set to 2047 and 4096 respectively.The probe tool revealed the following values: black pixels have values of 0.0 (as expected), white/grey pixels have values that range from: 1 to ~250 on study #1 (problematic image), 1 to ~3800 on study #2 ("good" image) the customer determined the window center/width values are in an appropriate range for the "good" images (i.E.Study 2 and 3), but are completely out of range for the "bad" image (study 1).The customer also noted: "clearcanvas's analysis is incorrect as the images have data beyond 256.The customer concluded there is a bug decompressing jpeg 2000 lossless images and so their probe analysis is incorrect" (note: data beyond 256" is referring to pixel depth).At the time, further analysis from synaptive found the following: the bits stored ((b)(4)) value is set to 12-bit in the dicom header; however, it appears the jpeg 2000 compressed pixel data is actually 16-bit; when the tag value is edited to 16, the image appears correct (i.E.Not dark) on load.Therefore, the dicom data is incongruent.A workaround was provided to the customer at the time: correct the dicom header to reflect the pixel data correctly.This should be fixed at the image source.Otherwise, manually change the bits stored ((b)(4)) value to 16.On 04-feb-2020, a software defect was discovered on while investigating an issue regarding a research device.This led to a retrospective analysis of feedback tickets.The research device uses the same jpeg 2000 codec as the device described in this report.While the likelihood of potential serious health consequences is remote, the use of the defective software associated could result in misdiagnosis, potentially causing significant indirect harm necessitating medical intervention that is serious but temporary.This defect is only encountered in select scenarios, where a modality produces images that are compressed using jpeg 2000, and the image pixel data is less than 16-bit, and the "bitness" of the compressed data stream does not match that of the image pixel data.The outcome of the above is a loss of precision in the decompressed pixel data, which causes the symptoms that were reported by the customer.The root cause is that a software defect found in the device that is encountered when it is used with non-dicom compliant jpeg 2000 compressed images.The software improperly uses the dicom bit depth (i.E.The bits stored tag) to decompress the compressed pixel data stream for display, instead of the bit depth that is encoded in the compressed pixel data stream itself.In non-dicom compliant images where the dicom and compressed pixel data stream bit depths do not match, the software outputs an image with some loss of precision in the decompressed pixel data.In these scenarios, the modality producing the image is itself also not compliant to dicom ps3.5 8.2.4.Note: the customer has two device licenses, (b)(4) the complaint describes issues encountered while using one device but did not disclose which device was being used.As the complaint is regarding one device, a single mdr will be reported.Mdr reports: 3012075008-2020-00001, 3012075008-2020-00003 and 3012075008-2020-00004 have the same root cause.A capa has been opened and recall initiated to address the issue in affected devices.
 
Event Description
The customer contacted synaptive medical inc on 2017-03-08 and reported that some mammography images appear very dark on image load.No adverse effects to the patient have been reported as occurring.
 
Search Alerts/Recalls

  New Search  |  Submit an Adverse Event Report

Brand Name
CLEARCANVAS RIS/PACS
Type of Device
SYSTEM, IMAGE PROCESSING, RADIOLOGICAL
Manufacturer (Section D)
SYNAPTIVE MEDICAL INC.
555 richmond street w.
suite 800
toronto, on M5V3B 1
CA  M5V3B1
Manufacturer Contact
maham ansari
555 richmond street w.
suite 800
toronto, on M5V 3-B1
CA   M5V 3B1
MDR Report Key9777544
MDR Text Key224676166
Report Number3012075008-2020-00002
Device Sequence Number1
Product Code LLZ
Combination Product (y/n)N
Reporter Country CodeUS
PMA/PMN Number
K110332
Number of Events Reported1
Summary Report (Y/N)N
Report Source Manufacturer
Source Type other
Reporter Occupation Non-Healthcare Professional
Remedial Action Recall
Type of Report Initial
Report Date 03/02/2020
1 Device was Involved in the Event
1 Patient was Involved in the Event
Date FDA Received03/02/2020
Is this an Adverse Event Report? No
Is this a Product Problem Report? Yes
Device Operator Other
Device Model NumberPERSONAL EDITION WORKSTATION 13.1
Device Catalogue NumberSYN-0524
Was Device Available for Evaluation? No
Date Manufacturer Received02/04/2020
Was Device Evaluated by Manufacturer? Yes
Date Device Manufactured12/15/2014
Is the Device Single Use? No
Is This a Reprocessed and Reused Single-Use Device? No
Type of Device Usage Reuse
Patient Sequence Number1
-
-