Mobile devices fare well in image display tests, but still face hurdles

November 29, 2010
John C. Hayes

Mobile devices for image display scored well in scientific papers presented Monday but still face significant hurdles before they are ready to become a routine part of image interpretation.

Mobile devices for image display scored well in scientific papers presented Monday but still face significant hurdles before they are ready to become a routine part of image interpretation. 

Papers comparing devices such as iPads and iPhones and, in one case, an Android-based smart phone, to standard LCD displays for diagnosis showed them pretty close to equivalent when viewing low-matrix CT and MR scans.

But the mobile devices lagged in terms of download times, time for diagnosis, and their ability to manipulate images. The iPad in particular was judged good for image display, given its larger size relative to the iPhone, but the interface fell short when it came to such standard processes as window and leveling that occur during interpretation.

Probably a bigger long-term challenge is winning FDA approval for the devices, said Dr. David Hirschorn, director of radiology informatics at Staten Island University and a featured speaker. In January, the FDA served notice that it did not consider mobile devices equivalent to more conventional imaging displays and that they would need FDA approval.

Current mobile devices also fall short in that they can’t be calibrated, and all display devices drift, Hirschorn said. Another issue is the brightness of the displays, although the iPad is bright enough to fall within current ACR guidelines for display brightness.

Finally, there are issues about security, which was not addressed in the session, and network speed, which slowed download speeds for the mobile devices.

Those issues aside, those who tested the mobile devices and presented their data suggested a bright future for them, if not immediately in diagnosis, then in other settings such as rounds, clinician and patient information, and case education.

A presentation by Dr. Kamran Shah of the University of Maryland Medical Center surveyed potential uses of the iPad. In addition to primary interpretation, they included radiographic viewing, speech recognition (with the included microphone and speaker), resident education, and patient/clinician interaction. For resident education, the iPad can deliver electronic books, web content, and the ability to interactively view case-based reports.

Shah displayed images from Japan in which an iPad draped in a sterile covering was used by surgeons to view images of the patient. Last summer, an article in Radiographic was released with interactive images formatted so it could be displayed on an iPad

Italian researchers tested the use of iPhones and an Android-based smart phone for interpreting CT and MR images in a teleradiology setting. When compared with standard LCDs in 93 exams, concurrence numbers were 22 for CT with two partial matches, and 68 for MR with one partial match. Floriana Zennaro, of the Institute of Maternal and Child Health in Trieste, said CT images were the most challenging because of the large number of slices. The investigators eventually limited the phone displays to one slice.

She concluded that, while there was no significant loss of diagnostic information, the phones were not useful for routine interpretations, but may be useful in consultations. Tablets like the iPad may be a better option and will be tried in a second round of the evaluation.

Because of the smaller form size of mobile devices, chest images are a challenge, but a pair of studies found they could work for CT pulmonary embolism scans and for checking chest x-rays for TB.

The PE study, from Massachusetts General Hospital, looked at 20 chest scans on PACS views and an iPhone. Image quality for the PACS images was rated 5 on a 1-to-5 scale, and 4-5 on the iPhone, said Dr. Supriya Gupta. Agreement between the PACS and iPhone images was excellent. The only problem was time of interpretation, 5.8 minutes on the PACS and 10.44 minutes on the iPhone, and transfer speeds.

It’s not likely radiologists will read chest CTs on the iPhone, but a larger and brighter system like the iPad shows potential, she said.

Only one paper assessed radiography on mobile devices. The study at the University of Maryland Medical Center looked at TB diagnosis on an iPad using 10 normal and 20 positive cases. The images were reviewed by two attending radiologists and eight residents. Equal numbers of cases (16) were misdiagnosed with the LCD and the iPad, said Dr. Frederick Weiss.

But when asked which they preferred, three of the radiologists preferred the LCD, two preferred the iPad, and the remaining five rated them equal.