Research On Carotid Hemodynamic Parameters Evaluation Based On Ultrasound Pulsed Wave Doppler

Pulsed Wave Doppler(PWD)is a traditional ultrasound technique used for the diagnosis of cardiovascular disease.This paper firstly introduces the principles and significance of carotid artery vascular ultrasound diagnosis.The spectrum generated by the PWD technique contains information on hemodynamics and hemodynamic parameters have important implications in the diagnosis of cardiovascular diseases.Secondly,this paper design and implement a spectral imaging method including RF signal acquisition,In-phase Quadrature,and frequency spectrum generation.Thirdly,handheld ultrasound has severe background noise in the spectrum because of device integration,and the traditional envelope method is difficult to work effectively.In this paper,we propose a real-time automatic envelope method based on the energy accumulation function.The accuracy of the algorithm is verified using ultrasonically blood-mimicking Doppler phantom as the standard,with an error of about 0.6%against the gold standard in the case of low noise and about 7% in the case of severe background noise.This paper proposes two envelope-based peak and valley detection methods to achieve real-time measurement.The two methods are named the double-step method and the statistical sliding window method.The experimental results show that the false detection rate of the double-step method is 2.15% and the statistical window method is0.72%.Both methods’ errors are lower than the 3.10% of the traditional method.Fourthly,this paper proposes an abnormal parameter detection method based on the Pau Ta Criterion,which screens abnormal parameters of samples based on statistical distribution characteristics.Compared with the traditional method which relies on a large number of positive samples,this method greatly reduces the effect of the specificity of a positive sample and reflects the sample whether is falling in the distribution.Compared with the traditional method which focuses only on individual parameters,this method makes the 13 measurable parameters to curve fit and detects whether the sample is abnormal according to the feature weights and Pau Ta criterion.The experimental results show that the proposed method achieves an accuracy of 87.04%,which is at least 7% higher than the commonly used anomaly detection methods K-Means,Support Vector Machine(SVM),and Gaussian Mixture Model(GMM).This method has the advantages of high accuracy,simple and fast computation,and reflecting distribution characteristics.Fifth,this paper proposes a hemodynamic assessment method based on multiple parameters.Traditional methods rely on a single parameter and tend to classify severe carotid stenosis.In this paper,The proposed method improves the performance of RUSBoost by setting the empirical weight of each sample.Experimental results show that the proposed method achieves 90.1% accuracy,70% sensitivity,and 94% specificity,which is4%,6%,and 2% higher than the original RUSBoost,respectively.Since the empirical weights are calculated based on the distribution without any expert input,the proposed method is objective..In summary,the series of algorithmic procedures proposed in this paper can be effectively applied to the assessment of carotid hemodynamic parameters and are a good guide for the screening and detection of carotid plaque or mild stenosis.