Model Number 8.2 |
Device Problems
Application Program Problem: Parameter Calculation Error (1449); Defective Component (2292)
|
Patient Problem
No Clinical Signs, Symptoms or Conditions (4582)
|
Event Date 02/09/2022 |
Event Type
malfunction
|
Manufacturer Narrative
|
Merge healthcare investigated the issue identified by a merge pacs software component supplier, (b)(4).(b)(4) requirement to check for square or equal x and y pixels, prior to rendering dicom compliant 3d image datasets, was not fully implemented in their software toolkit.(b)(4) performed their own investigation and did not find any customer/user complaints for non-square pixels in 3d rendered images over the past 18 years.(b)(4) searched their image database, amassed over 18 years, and did not find a single 3d renderable dataset with non-square pixels.Merge healthcare analyzed over 12 years of service records and complaints and was unable to find a single instance of a customer or user with an issue related to 3d images with non-square pixels.Merge healthcare also analyzed internal imaging databases and could not find any 3d renderable datasets with non-square pixels.(b)(4) complainant identified the missing requirement by using a non-square pixel 3d renderable dataset that was mathematically manipulated.Between (b)(4) and merge healthcare, neither company could find any real-world clinical/medical 3d renderable imaging datasets with non-square pixels.While there is a potential of non-square 3d renderable images existing in the field for ct, mr, pt and a few other imaging dataset types, no actual real-world evidence was found.Research into medical imaging design and standards also indicated that non-square x,y pixels in 3d dicom compliant image datasets is atypical.If real-world medical imaging 3d datasets with non-square pixels were generated in the viewer, there is a potential that diagnostic or treatment decisions may be inaccurate due to image distortion or incorrect measurements.Neither (b)(4) or merge healthcare have had any allegations from users indicating or alleging that patients were injured or harmed due to diagnostic or treatment decisions being made in the 3d viewer.Clinicians may render images in 3d for improved visualization of the patient's anatomy or a specific region of interest (roi).However, standard clinical practice is to measure a patient's anatomy or regions of interest in 2d.Merge healthcare's resolution for the missing (b)(4) requirement is to add a screening check to merge pacs.This check will prevent dicom compliant 3d image datasets with non-square x,y pixels from being rendered for viewing in (b)(4) tooklit.No further action is indicated.
|
|
Event Description
|
Merge pacs is a medical image and information management system that is used for viewing, selection, processing, printing, telecommunications, and media interchange of medical images from a variety of diagnostic imaging systems.On (b)(6) 2022, merge healthcare received an email notification from an approved supplier, (b)(4).(b)(4) provides a 3d medical imaging software toolkit that is used with merge pacs to render dicom compliant 3d images for viewing.The email provided information regarding a requirement that was not fully implemented within (b)(4) 3d medical imaging toolkit.The requirement is that: x and y pixel spacing must be equal for data loaded into 3d image rendering engines.Per (b)(4), "although cross sectional 3d imaging data almost always has equal x & y spacing as additional assurance, and to protect from the rare events they are not equal, our next release of the sdk will now not allow this data to load into 3d render engines (2d render engines will continue to support it)." merge healthcare has received no complaints from customers or users of merge pacs regarding the rendering of non-square x,y pixel 3d images.Merge healthcare, in conversations with (b)(6) medical, learned that the missing requirement was identified by a customer/oem that uses (b)(4) 3d imaging toolkit.This customer was doing research and development (r&d) on a feature unrelated to 3d image rendering.To perform their r&d activities, (b)(4) customer created a derived mathematical phantom dataset.When this mathematically manipulated image dataset was rendered in (b)(4) 3d viewer, the image was observed to be out of proportion.There is a potential that diagnostic or treatment decisions made from 3d images with unequal or non-square x,y pixels may be incorrect due to image distortion & inaccurate measurements.There are no reports of patient injury or harm as a result of this issue.Reference complaint (b)(4).
|
|
Search Alerts/Recalls
|
|