WASHINGTON, D.C. -- While the HITECH Act focuses primarily on making medical records interoperable for the benefit of the patient, one key looming technical Goliath is medical image storage -- that is, the proliferation of pixels and where to put them.
Here at the First AMA-IEEE Medical Technology Conference on Individualized Healthcare, engineers
New compression technologies might ease the burden of transmitting large files en route to achieving interoperability, said Zohara Cohen, IEEE member and program director at the National Institute of Biomedical Imaging and Bioengineering. So might computer intelligence that can help hospitals zero in on the relevant parts of a medical image (or video studies, an emerging technology) so health care providers won’t necessarily have to capture -- or retain -- as many pixels as they do now.
Data overload in medical imaging is a problem in health care for both technical and cultural reasons, Cohen said. Imaging equipment sometimes captures more information than is relevant. In addition, a patient might get multiple scans of the same area if he goes from one practitioner to another, because there could be barriers to accessing the first image. On top of that, while telemedicine’s new technologies might be improving patient care, the sensor data and video it transmits piles more data onto the clinician, who needs to sift through it all to find a few relevant pieces to make a treatment decision.
“We’re collecting a lot of data,” Cohen said in an interview with SearchHealthIT.com. “From a technology perspective, we need to look at algorithms that know how to detect the interesting information in an image and store that. I’m a fan of intelligent compression, where the compression of an image might compress different parts of an image to different extents -- as a function of where the important content is.”
Cohen is managing National Institutes of Health grants to help carry out that research, as well as to help develop tools that, in the long run, aim to enable such individual practitioners as radiologists and physicians to winnow a patient’s data in order to get to that informed treatment decision. Software can help, with more user-friendly graphics, as well as enabling a practitioner to group images from a number of patients and get an idea of what a particular health problem typically looks like.
PHR services as a cure to medical image storing ills?
Compression will help transmit imaging data more quickly and efficiently. For IT managers seeking relief from medical image storage woes, however, it can only do so much until the health care world can figure out just how much it can and cannot store in the light of federal rules and other motivators -- such as the pressure to keep every shred of data for defense against possible future litigation.
We’re collecting a lot of data. From a technology perspective, we need to look at algorithms that know how to detect the interesting information in an image and store that.
Zohara Cohen, program director, National Institute of Biomedical Imaging and Bioengineering
“I think that there are policy issues,” Cohen said, adding that they will require human decision making to sort out. Once an organization is past that, technology can help ease the medical image storage burden. “If you have a CT scan with 100 images, well, maybe you really only need to store 30 of them. But right now, no one has the time to really look through, manage the data and throw out what they don’t need. Certainly, that’s something that computers could be good at, and I see that happening.”
As for breaking down barriers to access to images among providers, Cohen does not see the model of patients carrying around thumb drives or DVDs from appointment to appointment. Instead, she imagines large, back-end databases with secure Web portals that physicians can log into and get what they need to administer a patient’s care. While Google Health and Microsoft HealthVault might be newcomers to the market for personal health record services -- and some consumers might have reservations trusting those large companies with their data -- PHRs might take off.
“I think that people will become more comfortable with that,” Cohen said. “The challenge, from the technology perspective, is for the researchers in this field to show the benefit. Once we have those classic killer apps where patients recognize how much value they can gain from allowing their data to be shared … I think they’ll warm up to it.”
Let us know what you think about the story; email Don Fluckinger, Features Writer