The idea of using smart phones as tool for teleradiology (and teleneurosurgery) has existed as long there has been smart phones (from the first greyscale-display Nokia Communicators). Of course the device displays, communication and data processing speeds have not been adequate for competing with real workstations and fixed-line communications.  However, the smart phone features like pocket size dimensions, usage regardless of the physical location and the usage for voice calls at the same time as you review the data have been very appealing for teleradiology as a secondary consultation tool.

The first real system that reached production use (as far as I know it) was the MOMEDA project that I have already bypassed in my previous blogs. With the system developed in that project e.g. neurosurgeons in the Oulu University Hospital in Finland have since year 2000 made their decisions based on CT (Computed Tomography) image data even before entering the hospital with a help of smart phone terminals [1]. As said the first prototypes were done to Symbian predecessor Psion’s MX5 PDA device (running EPOC ER5). Unfortunately MX5 did not have integrated communication capabilities and therefore the actual device that was used was Nokia Communicator 9110 with GeOS 3.0.

In this blog I like to present few suitability issues of current Symbian devices (Series60 & Series80) as tools for medical image (restricted to CT and MR) viewers.

The first obvious issue is the mobile device screen size and resolution. The CT and MRI (Magnetic Resonance Imaging) images are commonly 512x512 pixels with 12bit greyscale dynamics. So theoretically the images can have 4096 different shades of grey. This brings us to the second obvious issue with medical images.

Since it is essential that the original image can be viewed with the original screen resolution without scaling we need to evaluate device screen resolutions. Following picture presents current Symbian device display resolutions versus CT image resolution.

Clearly Series60 displays with 352x416 and Series80 displays with 640x320 resolutions suits best for the viewing. However, to get quick overview of the image(s) scaled images (e.g. 128x128 thumbnails) can be used and then even Series60 device with 176x208 resolution might manage as well.

The other issue is the device capability of displaying pure shades of grey. Since the images might contain theoretically 4096 different shades of grey it is obvious that smart phone displays can not handle that. But even the human eye cannot accurately distinguish between that amounts of different shades of grey. Therefore to allow the observer to interpret the image, only a limited number of greyscales are displayed at once. Clinically useful grey scale is achieved by viewing suitable range of grey scales depending on the tissue being studied. Following picture presents greyscale dynamics with common smart phone displays as well as two examples about different greyscale ranges for different regions of interest.

 

You might wonder how you know what is the greyscale are to display e.g. for brain tissue? To put simple, each pixels represents different x-ray attenuation values. This number is compared to the attenuation value of water and displayed on a scale of arbitrary units named Hounsfield units (HU). So with Hounsfield Units you can distinguish different greyscale ranges e.g. for brain tissue. For those who like to understand better there for example nice document called “Introduction to CT physics”.

Ok, this was quick overview for the subject and I hope this gave you some ideas that smart phones can be used in very different medical applications.