Computer scientists and electrical engineers are devising algorithms that look for useful new patterns in data produced by medical sensors.
Medical radiography is basically a species of photography. Historically, the patient’s limb would be positioned between an x-ray source and a photographic plate. The plate would be exposed and developed, and the result was an image of the limb’s interior.
Today, most larger x-ray machines use digital sensors rather than photographic film, but otherwise, little has changed. The raw data captured by the sensors is easily interpretable as a visual image.
That’s not true of more recent imaging technologies, however. In magnetic resonance imaging (MRI), for instance, different types of electromagnetic signals are individually analyzed and then stitched into a composite image. The algorithms that produce MRI scans are just as remarkable as the hardware.
Interpretation of the data produced by medical sensors remains one of the most fruitful applications of computer-science and signal-processing techniques in medicine, and it’s one that MIT researchers are pursuing down a number of different avenues.