Feature extraction and classification system and method for gastric magnetic signals

A feature extraction and signal extraction technology, applied in the field of bioelectromagnetics, can solve problems such as patient harm and bad eating habits, and achieve the effects of large noise suppression ability, good convergence, and low noise weight.

CN105662391AActive Publication Date: 2016-06-15NORTHEASTERN UNIV
3 Cites 2 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Publication Date
2016-06-15

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention relates to a feature extraction and classification system and method for gastric magnetic signals and belongs to the field of bioelectromagnetics. The feature extraction and classification system comprises a gastric magnetic fast-wave signal extraction module, a gastric magnetic signal filtering module, a gastric magnetic fast-wave feature extraction module and a gastric magnetic signal classification module. The morphological method and the RLS adaptive algorithm are adopted to effectively filter out baseline drift components in signals; low-frequency components and high-frequency components of the gastric magnetic signals are decomposed by wavelet packets, corresponding bands are adaptively selected according to signal features and analysis requirements to be matched with spectrums of the signals, and noise components proximate to fast-wave components in the gastric magnetic signals are filtered out; a matching pursuit and nonparametric basis function combined algorithm not only overcomes the defect that time-frequency atoms need to be built in advance by independently using a matching pursuit algorithm, and ensures that template signals can be adaptively conditioned not to depend too much on prior knowledge; the K-means algorithm is adopted, so that the solving time is short and the classification effect meets requirements.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention belongs to the field of bioelectromagnetics, and in particular relates to a system and method for feature extraction and classification of gastric magnetic signals. Background technique

[0002] Gastrointestinal disease is a common and frequently-occurring disease in today's society, and its incidence exceeds 1% of the world's total population, which has had a serious impact on people's daily work and study; with the accelerated pace of people's lives leading to The diet structure is changing accordingly, and the diversity of diets has also given rise to many bad eating habits; overeating and unscientific dieting are the causes of gastrointestinal diseases; the common way to diagnose gastrointestinal diseases is gastroscopy , radionuclide breath test, gastric emptying scintigraphy, X-ray, etc., but these methods will cause certain harm to patients.

[0003] Compared with the above-mentioned invasive gastrointestinal disease detection met...

Examples

Embodiment Construction

[0065] An embodiment of the present invention will be further described below in conjunction with the accompanying drawings.

[0066] In the embodiment of the present invention, such as figure 1 As shown, the feature extraction and classification system of gastric magnetic signal includes gastric magnetic fast wave signal extraction module, gastric magnetic signal filtering module, gastric magnetic fast wave feature extraction module and gastric magnetic signal classification module, wherein, gastric magnetic fast wave signal extraction module It is used to extract the baseline drift components in gastric magnetic fast wave signals in different physiological states through morphological methods; the gastric magnetic signal filtering module is used to extract gastric magnetic The baseline drift component in the fast wave signal is filtered out; and the noise close to the frequency of the fast wave component in the gastric magnetic signal is filtered out by the wavelet packet me...