Value of MR diffusion weighted imaging in the diagnosis of prostate cancer / 中华泌尿外科杂志

Objective To investigate the value of magnetic resonance (MR) diffusion weighted imaging(DWI) in the diagnosis of prostate cancer(PCa). Methods Fifty-seven patients with suspected prostate cancer underwent DWI and T_2-weighted imaging (T_2W). These images and apparent diffusion coefficient (ADC) maps results were compared with histopathologic findings. Receiver operating characteristic(ROC) analysis was used to compare the cancer detection performance of them. The results were rated on a scale of scores Ⅰ (benign) to Ⅴ (malignant) on the basis of ADC maps. Abnormal voxels were overlaid on the corresponding transverse TRUS images and used to perform voxel-guided biopsy. Results DWI had a sensitivity of 85%, specificity of 82%, positive predictive value of 80%, negative predictive value of 86% , and accuracy of 83%. T2WI had a sensitivity of 77%, specificity of 71%, positive predictive value of 69%, negative predictive value of 79%, and accuracy of 74%. The areas under the ROC curves for DWI and T_2WI were 0. 830 and 0. 742, respectively. The performance of DWI in PCa detection was significantly better than of T_2WI (P<0. 05). 6 of 30 patients with negative DWI results also had negative biopsy findings. PCa was detected in 17(85%) of 24 men findings with voxel score Ⅳ , with a sensitivity of 100%, specificity of 46%, positive predictive value of 71 %, negative predictive value of 100% , and accuracy of 77%. Conclusions The performance of DWI in PCa detection was better than of T_2 WI. ADC maps can be transferred to TRUS images and used to sample regions of cancer in men with rising PSA levels and negative findings at prior biopsy with good accuracy. DWI appears to be a robust and reliable method to examine the whole prostate within an acceptable scan time in clinical settings.

The application of gene expression programming in the diagnosis of heart disease / 生物医学工程学杂志

GEP (Gene expression programming) is a new genetic algorithm, and it has been proved to be excellent in function finding. In this paper, for the purpose of setting up a diagnostic model, GEP is used to deal with the data of heart disease. Eight variables, Sex, Chest pain, Blood pressure, Angina, Peak, Slope, Colored vessels and Thal, are picked out of thirteen variables to form a classified function. This function is used to predict a forecasting set of 100 samples, and the accuracy is 87%. Other algorithms such as SVM (Support vector machine) are applied to the same data and the forecasting results show that GEP is better than other algorithms.