A Disease Classification System Based on Eye Movement Information

Abstract

The invention relates to the technical field of disease diagnosis systems, and discloses a disease classification system based on eye movement information. The disease classification system comprisesan eye movement feature extraction system, a weak classification system and a strong classification system; the eye movement feature extraction system is used to extract eye movement feature vectors from eye movement videos of eye movement tests of a testee; each time the required eye movement test is completed, m eye movement feature vectors are extracted in total; the weak classification systemincludes pre-trained LSTM weak classifiers in one-to-one correspondence to the m eye movement feature vectors, the LSTM weak classifiers are used to calculate attribute values of individual eye movement feature vectors belonging to various diseases; the strong classification system is used to jointly classify the m eye movement feature vectors, and calculate the joint classification probability ofvarious diseases. The disease classification system solves the technical problem that the prior art needs to rely on thorough prior medical knowledge for disease diagnosis, classifies various diseases, has no single function, is more flexible diagnosis and non-invasive, and has no harm to the human body, strong practicability and low cost .

Description

technical field [0001] The invention relates to the technical field of disease diagnosis systems, in particular to a disease classification system based on eye movement information. Background technique [0002] Medical research shows that eye movement involves 6 pairs of cranial nerves, including the optic nerve, oculomotor nerve, and trochlear nerve. Abnormal eye movement is related to various mental activities, mental disorders, and physical diseases. [0003] From domestic and foreign studies, relevant research on eye movement images based on video acquisition has been carried out and some progress has been made, but the mining of the rich physiological and psychological information contained in human eye movements is far from enough. Limitations include: [0004] (1) The main research stays in the intuitive eye tracking performance improvement and human-computer interaction application; [0005] (2) Due to the unclear mechanism, it is difficult to directly correlate e...

AI Technical Summary

Problems solved by technology

[0005] (2) Due to the unclear mechanism, it is difficult to directly correlate eye movement characteristics with disease discrimination, resulting in relatively single product functions, and it is difficult to distinguish multiple diseases with traditional statistical methods, and it is mainly for vestibular diseases, vertigo, etc. ;

[0006] (3) The automatic analysis function is weak, and most applications only present intuitive information to users, and researchers or doctors make judgments, which are highly subjective and poor in practicability;

[0007] (4) The equipment is expensive and difficult to promote

Images