| The study concentration in this paper was biomedical informatics, and the research object was heart. In this paper, we developed a method of multi-parameters to explore hidden information extraction of the cardiac state. As was well known, cardiac was a very complex organ. Although the cardiac study could be traced back to the early of last century, the previous study generally concentrated on static analysis such as electrocardiograph (ECG) figuration analysis and single parameter detection, which was unsatisfactorily with characterizing the various heart states precisely and overall. In fact, the cardiac activity implied abundant information about heart states, some of which remained unemployed even until the present and far from being explored. On consideration of this, we give emphasis to the study of method on characterization of multi-parameters which are local and global parameters made up a parameter set for use. including high frequency electrocadialogical (HF-ECG) indices, linear global parameters of heart rate variability (HRV), and non-linear indices containing Poincare disserted-dot plot and its quantitative parameters, fractal dimension (FD), as well as mutual information (MI). On the basis of this, a series of pathologic heart states was simulated anatomically on animals in vivo, and tested the role of correlative parameters in heart states changing process and their realistic application value with dynamic detection of multi-parameters. Besides, the sensitivity and effects of various parameters under particular conditions were studied in detail, and their differential properties specified respectively in order to consummate the parameter set characterization of the physiological and pathological heart states, and study of the multi-parameter method in hidden information extraction of heart state changes carried out as well. They will be expected to provide proof for further complex interpretation and for clinical application. It was one of the principal innovations in this paper.To investigate parameters' changing properties when neural control of the heart was inhibited by global extrinsic control, we first made pathologic models with animals in vivo by means of anesthesia and surgery. Evidences were found that inhibiting global extrinsic control, to a certain extent, would influence global parameters greatly, rather than take much effect on the extrinsic control of the heart autoregulation. It could be inferred that the global parameters were special sensitive to extrinsic control from both sympathetic nerve and parasympathetic nerve.Both pathological models of acute myocardial ischemia and injury on dogs heart were made, and the method of multi-parameter detection was used to study dynamical changes of the heart under its various states resulting from acute injury. As a result, the method did reveal the changing properties of heart states satisfactorily with a number of parameter referring to each other. For example, MI behaved most sensitive to heart state when spontaneous respiration inhibited with respiration machine; FD could manifestheart healthy level, which reduced sharply sooner after acute injury but remained almost steady afterwards; the parameters of Poincare dissert-dot plot and HRV could accurately distinguish the state change before and after ligaturing; the HF-ECG excelled in not only revealing injury degree but also locating injury position. Nevertheless, no one of the parameters could singly characterize the integrated changing processes of heart states under experiment conditions, that showed the multi-parameter method to here obvious advantage.In this dissertation, the multi-parameter method in chronic experiments was also applied in which heart states were dynamically measured in two months from an acute operation started to the end. In this series of experiments, remarkable advantages were demonstrated of dynamic detection method. About cardiac status at the first week after acute ligaturing, the heart states changing phenomena could be observed a whole and dis... |
