Method for positioning scalp electroencephalogram epilepsy region based on artificial intelligence

Abstract

The invention relates to the field of artificial intelligence, in particular to a scalp electroencephalogram epilepsy region positioning method based on artificial intelligence. Comprising the steps that historical electroencephalogram waveform images are collected and compressed, and a sparse coding image of each historical waveform image is obtained; a neural network is used for training, and the LOSS function historical value of each sparse coding image is obtained; calculating a mean value of historical values of the LOSS function of the sparse coding image, and obtaining an oscillation sparse coding image; calculating the similarity between the oscillation sparse coding image and each sparse coding image, and obtaining an associated sparse coding image; and reconstructing an LOSS function of the neural network, and outputting a labeled image. According to the technical means provided by the invention, the LOSS function of the neural network is re-constructed, so that the recognition accuracy of the neural network can be effectively improved, and the influence of a relatively poor neural network training result caused by inaccurate manual annotation is effectively avoided.

Description

technical field [0001] The invention relates to the field of artificial intelligence, in particular to an artificial intelligence-based method for locating epileptogenic regions of scalp electroencephalography. Background technique [0002] Epilepsy is a short-term brain dysfunction caused by abnormal discharge of neurons in the brain. It is a very common chronic neurological disease. The clinical symptoms are loss of consciousness, convulsions, and even fainting, and sometimes repeated attacks. Patients hope for surgical treatment to get rid of the impact of the disease on work and life. However, there is still an important problem in the surgical resection of epileptogenic foci—how to preserve the corresponding brain functional areas without being destroyed. Under the influence of many limitations, doctors can only perform palliative resection, which will lead to recurrence of the disease or damage to brain function. Therefore, it is very important for patients to accurate...

AI Technical Summary

Problems solved by technology

[0005] The invention provides a method for locating the epileptogenic zone of the scalp EEG based on artificial intelligence to solve the existing problems, including: collecting historical waveform images for compression, and obtaining sparsely coded images of each historical waveform image; using a neural network Perform training to obtain the historical value of the LOSS function of each sparsely encoded image; calculate the mean value of the historical value of the LOSS function of the sparsely encoded image to obtain an oscillating sparsely encoded image; calculate the similarity between the oscillating sparsely encoded image and each sparsely encoded image to obtain associated sparse encoding Image; rebuild the LOSS function of the neural network, and output the labeled image

Images