Neural network development helps dentofacial surgeons to diagnose.
Neural network means software that works similar to human nervous system. When confronted with unfamiliar objects, it analyzes their properties, draws conclusions, and uses the information obtained for further training.
One of the ways to develop a neural network is to analyze unstructured information in the form of pictures, images, photos, and in particular, medical documents.
Scientists from South Korea have confirmed the success of the technology. According to the journal “Nature” the neural network analyzed photos of 411 patients who needed surgery for bite correction and those of 411 people who didn’t not need the procedure.
In 90% of cases the neural network correctly identified patients who needed surgery and those who didn’t.
Epol Soft has similar experience – developing a neural network to create a 3D model of the jaw and individual teeth. The service uses an ordinary X-ray image and:
- displays images of jaws, teeth in 2D and 3D formats;
- divides the photos of the jaw into individual teeth;
- recreates multiple “layers” of the tooth with a color display of the tooth structure by tissue.
The resulting data is downloaded to a device that simulates the dental treatment process and:
The software recreates the dental treatment process as realistically as possible and facilitates dental education for students by allowing to remotely monitor and analyze their compliance with treatment procedures.
Development of the neural network for dentistry is one of the things Epol Soft does. Read about other medical solutions here.