From IEEE Spectrum:
Neural networks are artificial intelligence systems that excel at interpreting images. … But they are typically power hungry, which has limited their use so far. Vision processor company Movidius of San Mateo, Calif., hopes to change that with a low-power chip designed to run neural networks.
Aside from being amazing technology this development could have direct applications in dentistry and medicine. The article states that neural nets are good at interpreting images. That means they could be taught to read a radiograph or evaluate the photo of a lesion.
In fact a whole range of diagnostic evaluations could be performed by a neural net system using scans of tissue with a device like S-Ray or Canary or even traditional x-rays. The scan could be taken by a para professional or even by the patient themselves, vastly reducing costs.
Just another step to the Tricorder.