IEEE International Conference on Signal and Image Processing (ICSIP), Beijing, China, 13 - 15 August 2016, pp.462-465
Electrogastrography (EGG) is a noninvasive technique to record the gastric electrical activity. In recent years, researchers have proved that EGG has great clinical potentials to evaluate the gastric disorders. However, it has still not been used in clinic to diagnoses of diabetic gastroparesis or some other disorders. One of the reasons of this can be related to methods using for EGG signal analyzing. Normal gastric myoelectrical activity consists of a slow wave or Electrical Control Activity (ECA) and spikes or Electrical Response Activity (ERA) potentials. Although ERA could be responsible for some changes in EGG signals, it is ignored and neither First furrier transform (FFT) based running spectral analysis (RSA) nor can discrete wavelet transform (DWT) reflect the ERA effect. Line length is an amplitude-and frequency-dependent linear feature therefore it reflects the waveform dimensionality changes. The aim of this study was to investigate the line length feature of the EGG whether it has potential for discriminate the healthy subjects (CT) from diabetic gastroparesis (GP) patients and discuss the line length capability to reflect the ERA effect on EGG. For this aim EGG signals were recorded from 20 CT and 20 GP patients who had delayed gastric emptying. The recordings were analyzed and compared in both groups' pre- and post-prandial states. The results we obtained from line length analysis of EGG signals proved that there are significant differences between the GP and CT groups is statistically significant for pre and postprandial condition (p = 0,0298* for pre-prandial and p = 0,0032* for post-prandial). The proposed method confirmed that line length feature is useful for detecting delayed gastric emptying and using with some other features more successful computer-aided detection system can design.