Search

In view of the modal aliasing phenomenon that occurs when EMD is used to deal with the vibration signals of hydropower units which are coupled by multiple signals and noise, a fault signal preprocessing method is proposed based on differential empirical mode decomposition and energy characteristics.The signal is decomposed by differential empirical mode algorithm and then the energy matrix of each order differential signal is constructed. The simulation results show that the differential empirical mode decomposition can effectively avoid the mode mixing effect. The support vector machine is used to verify that the identification accuracy can reach 98% in the preprocessing of the guide bearing signal of the hydropower unit. This method has a good engineering application value in the application of complex coupling signal pretreatment of hydroelectric units. [ABSTRACT FROM AUTHOR]

The waste dump of an iron mine in Anhui Province has been abandoned for several decades. Pyrite in the exposed waste ores is oxidized by acidophiles and large amounts of metal ions and H 2 SO 4 are released, resulting in the formation of an acid mine drainage (AMD) lake since 1970s. Besides the lake, there are also some small-scale AMD adjacent to the newly deposited waste ore. In order to study the acid generation potential of the waste ore and the related microbial communities, soil samples were taken from beside the AMD lake (1LL) and small-scale AMD (5J, 5Y, 6-1, 6-2, 6-3) and the physicochemical properties and microbial community of these samples were analyzed. The results reveal that all of samples were highly acidic and the pH of the 1LL sample was 2.77, while the other samples were even more acidic, at less than 2.6. The electrical conductivity (EC) (0.32 mS·cm -1 ) of the 1LL sample was obviously lower than the other samples (2.25-7.08 mS·cm -1 ), which indicates that the newly deposited waste ore contains higher ion concentrations. The Fe 2+ concentration of the 1LL sample was only 0.80 mg·kg -1 but the other five samples were as high as 2.91-33.40 mg·kg -1 . This suggests that most of the Fe 2+ in the 1LL sample has been converted to Fe 3+ after long-term oxidization. High-throughput sequencing results showed that most acidophiles in 1LL sample were Actinobacteria, Acidobacteria and Chloroflexi but the microbes in the remaining five samples were γ -Proteobacteria, Firmicutes and Nitrospira. The iron-sulfur oxidizing bacteria, such as Sulfobacillus , Leptospirillum , Acidithiobacillus, were scarce in the 1LL sample, while they highly abundant in the other five samples, which proves that the acid-generation process of the newly deposited waste ore is strong. However, the reduced iron and sulfur in the 1LL sample has nearly been depleted. Statistical analysis shows that the microbial composition of the 1LL sample is significantly different to that of the five newly deposited samples, illustrating that microbial community composition is remarkably influenced by physicochemical conditions.

Background and Objective: Gastrointestinal stromal tumor (GIST) is one of the most common mesenchymal tumors of the digestive system. Imaging examination plays an important role in preoperative diagnosis and postoperative evaluation for it. This study was to describe the multi-slice spiral computed tomographic (MSCT) findings and pathologic features of GIST, and to analyze their correlation., Methods: MSCT and pathologic reports of 49 patients with 53 pathologically confirmed GIST lesions were reviewed and compared., Results: Of the 53 GIST lesions, 14 were at very low biological risk, 11 at low risk, ten at moderate risk and 18 at high risk; 36 (67.9%) were found in first visit by CT scans. On CT images, the GIST lesions with maximal diameter of > or =50 mm showed irregular shape, invasive growth, presence of cystic area and heterogeneous enhancement, and most of them were at high risk; the lesions with maximal diameter of <50 mm showed regular shape, expansive growth, and homogeneous enhancement, and most of them were at risk of moderate or below. No lymph node metastasis was found. Only three lesions showed S100-positive, which presented infiltration along the gastric wall or bowel ring on CT images., Conclusions: CT examination is helpful in risk prediction for GIST, but it is difficult to detect small lesions (< 2 cm) by CT scans. Due to the infiltrative growth of GIST with neural differentiation (S100-positive), it is difficult to distinguish GIST from gastric cancer on CT images.

Background & Objective: It is difficult to discriminate progressive gastric cancer and gastric lymphoma by CT imaging, because incrassate gastric wall, lump in gastric cavity, confined gastric cavity, intumescent lymph node, and distant metastasis can be displayed in both of them. This study was to compare the CT findings between gastric cancer and gastric lymphoma to improve diagnosis of gastric tumors, especially for gastric lymphoma., Methods: CT images of 27 patients with pathologically proved progressive gastric cancer and 25 patients with pathologically proved gastric lymphoma were reviewed. Tumor location, appearance, scope of involvement, gastric wall thickness, mucous membrane, mucosal fold, serosa membrane, necrosis, enhancement degree and uniformity, involvement of other organs, and abdominal lymph nodes were observed., Results: White line sign was observed in 23 cases (85.2%) of gastric cancer, but not in the 25 cases of gastric lymphoma. The extent of white line sign in gastric cancers was larger in portal vein phase than in arterial phase. Enhancement degree outside the white line was higher in portal vein phase than in arterial phase in 13 cases (48.1%) of gastric cancer. The extent of involved gastric wall was smaller than 50% of the whole gastric wall in all the 27 cases of gastric cancer, while it was larger than 75% in 23 cases (85.2%) of gastric lymphoma. Gastric mucous membrane ulcer was found in all of the 27 cases (100%) of gastric cancer, while it was found in only 1 case (4.0%) of gastric lymphoma. Intumescent lymph nodes in two or more areas were found in 11 cases (40.0%) of gastric lymphoma, but not in gastric cancer. Intumescent lymph nodes in the retroperitoneal space below renal hilum were found in 8 cases (32%) of gastric lymphoma, but not in gastric cancer., Conclusion: There are some different CT features between gastric cancer and gastric lymphoma, such as white line sign, gastric mucous membrane ulcer, extent of involved gastric wall, location of intumescent lymph nodes surrounding the stomach and in retroperitoneal space below renal hilum, and so on, which could be helpful in differential diagnosis of these two diseases.

Background & Objective: Most parotid tumors grow slowly. It is difficult to evaluate their malignancy. This study was to analyze the CT features of parotid tumors., Methods: CT images of 133 patients with 157 lesions of parotid tumors, confirmed by histopathology, were reviewed. The location, contour, margin, enhancing type, necrosis and liquefaction, and adjacent structure involvement were analyzed., Results: Among the 157 lesions, 110 (70.1%) were benign tumors, 18 (11.5%) were borderline tumors, and 29 (18.5%) were malignant tumors; 80 (72.7%) benign lesions and 14 (77.8%) borderline lesions located in the superficial lobe, whereas 7 (24.1%) malignant lesions located in the deep lobe and 10 (34.5%) in both the superficial lobe and the deep lobe; 99 (90.0%) benign lesions had sharp margins, 8 (44.4%) borderline lesions had sharp margins and 10 (55.6%) had partly unsharp margins, 10 (34.5%) malignant lesions had partly unsharp margins and 11 (37.9%) had unsharp margins. Most benign lesions were round (68/110, 61.8%) or oval (23/110, 20.9%), while malignant lesions were often irregular (14/29, 48.3%). Eleven lesions, including 2 borderline tumors and 9 malignant tumors, involved the adjacent subcutaneous fat. All the lesions which involved the bed of parotid gland were malignant. No benign tumor involved adjacent structures., Conclusion: On CT images, a parotid tumor located in the superficial lobe, with a round or oval contour and sharp margin, is more likely to be a benign tumor; otherwise, it might be a malignant tumor.

Background & Objective: Malignant epithelial parotid tumors have various histopathologic types. Computed tomography (CT) has been widely used to detect parotid masses. This study was to analyze CT features of malignant epithelial parotid tumors to evaluate the diagnostic value of CT in the characterization of malignant epithelial parotid tumors., Methods: CT reports of 29 patients with malignant epithelial parotid tumors (8 at low grade and 21 at intermediate or high grade), confirmed by histopathology in Cancer Center of Sun Yat-sen University, were reviewed. The localization, contour, margin, enhancing type, homogeneity, infiltration into the adjacent structures, and lymph node metastasis detected by CT scan were compared with relevant pathologic results., Results: Of the 8 cases at low grade, 4 were well-differentiated mucoepidermoid carcinoma, 3 were acinic cell carcinoma, and 1 was epithelial-myoepithelial carcinoma; 3 had sharp margin, 3 had partly unsharp margin, and 2 had unsharp margin; 5 had regular contour, and 3 had irregular contour; all were enhanced obviously; 2 showed homogeneous appearance, and 6 showed inhomogeneous appearance with low-density areas; none had adjacent infiltration; 3 had lymph node metastasis. Of the 21 cases at intermediate or high grade, 5 were poorly-differentiated squamous cell carcinoma, 8 were malignant pleomorphic adenoma, 2 were adenocarcinoma, 1 was lymph-epithelial carcinoma, 4 were mucoepidermoid carcinoma, and 1 was adenoidcystic carcinoma; 5 had sharp margin, 7 had partly unsharp margin, and 9 had unsharp margin; 10 had regular contour, and 11 had irregular contour; 17 were enhanced obviously, and the 4 cases of malignant pleomorphic adenoma were enhanced slightly; 9 showed homogeneous appearance, and 12 showed inhomogeneous appearance with low-density areas; 8 had adjacent infiltration (the fat space between tumor and the parotid bed disappeared); 6 had lymph node metastasis., Conclusions: To some extent, CT features of malignant epithelial parotid tumors are correlated to the differentiation of tumor cells. Moreover, tumor size and histologic subtype should be considered to make a more accurate imaging diagnosis.

Background & Objective: The helical double-phase CT scan imaging features of hepatocellular carcinoma (HCC) overlap those of other hepatic lesions. This study was to investigate the helical double-phase CT scan imaging features of HCC to improve diagnosis accuracy., Methods: Double-phase CT data and pathologic data of 52 HCC patients, received resection in Cancer Center of Sun Yat-sen University from Dec. 2000 to Dec. 2002, were analyzed. The double-phase CT features of HCC lesions were summarized. The pathology of false-positive lesions was analyzed., Results: CT scan showed 56 lesions in the 52 patients: 51 were cancer lesions, including 49 HCC lesions and 2 mixed lesions of HCC and cholangioma, 5 were false-positive lesions. Arterial phase of these HCC lesions showed obvious heterogeneous enhancement, and the portal vein phase showed heterogeneous low dense. Necrosis was seen in all massive lesions, but was seldom seen in nodular and small lesions. Most lesions had clear borders and amicula. The pathologic diagnoses of the 5 false-positive lesions were hepatic cirrhosis with hepatocellular nodular hyperplasia, regenerative nodule, hepatic cirrhosis, bile duct calculus companied with inflammatory reaction, and fibrosis hyperplasia., Conclusions: Helical double-phase CT scan can be used to diagnose typical HCC lesions. There are no obvious differences in helical double-phase CT scan between HCC lesions and false-positive lesions. The diagnosis of HCC must be based on clinical information, follow-up or biopsy.

Background & Objective: Blood flow patterns with CT perfusion imaging (CTPI) had been innovated to delineate abnormal hemodynamic lesions in the liver, brain, and kidney. This study was to evaluate the blood flow patterns of pulmonary masses by CTPI, and determine the value of CTPI in differential diagnosis among benign, malignant, and inflammatory masses., Methods: Fifty-two patients with previously diagnosed pulmonary masses (37 with malignant masses, 7 with benign masses, and 8 with inflammatory masses) underwent dynamic CTPI. Time-density curves (TDC) of artery, vein, and pulmonary masses as well as mass perfusion images and parameters including perfusion volume (PV), peak height (PH), mean transit time (MTT) and blood volume (BV) were obtained by Phlips CT-perfusion software., Results: The values of PV, PH, and BV were significantly higher in malignant masses and active inflammatory masses than in benign masses [(27.63+/-15.06) ml.min(-1).ml(-1) and (30.80+/-20.33) ml.min(-1).ml(-1) vs. (11.81+/-3.74) ml.min(-1).ml(-1), (28.46+/-12.07) Hu and (32.15+/-15.89) Hu vs. (10.41+/-3.77) Hu, (21.64+/-10.97) ml/100 g and (28.38+/-14.55) ml/100 g vs. (10.61+/-5.33) ml/100 g, P<0.01]. However, the differences of MTT among malignant, inflammatory, and benign masses were not significant [(28.39+/-21.66) s, (25.91+/-14.57) s, and (29.86+/-13.57) s, P=0.928]. No significant differences in the 4 parameters were found between malignant and inflammatory masses. When PV >20 ml.min(-1).ml(-1) and pH >15 Hu were set as a diagnostic threshold (excluded active inflammatory masses), the sensitivity, specificity, and accuracy were 91.9%, 100%, and 84.1%, respectively., Conclusion: CT perfusion imaging provides quantitative information about blood flow patterns of pulmonary masses and is an applicable diagnostic method for differentiating pulmonary masses.

Background & Objective: Based on biological behavior and histopathologic characteristics, ameloblastomas in the maxillomandibular region can be divided into several subtypes with different prognosis. This study was to explore the clinical and CT imaging features of solid or multicystic ameloblastoma (SMA), unicystic ameloblastoma (UA), and malignant ameloblastoma (MA), and thus to improve the CT diagnostic accuracy for these diseases., Methods: Clinical and CT features of 25 histologically proven ameloblastomas in the maxillomandibular region, including 14 SMAs, 6 UAs, and 5 MAs were analyzed retrospectively., Results: The median age of the patients initially diagnosed as SMA, UA, and MA were 33.5, 20.5, and 56 years, respectively. The mean maximal transverse diameters on CT images were 33 mm for SMA, 46 mm for UA, and 59 mm for MA. Of the 14 SMAs, 11 (79%) showed low to intermediate density as compared with muscles, 3 (21%) showed homogenous and intermediate density; 12 (86%) appeared as lobulate lesions; 10 (71%) had incomplete bone septa; 9 (64%) showed root resorption of the neighboring teeth; 13 (93%) showed cortical perforation; 6 (43%) had soft tissue masses; all SMAs showed well-defined margins. Of the 11 SMAs examined by enhanced studies, 9 (82%) showed strong enhancement. CT images of the 6 UAs showed homogenous and low density in comparison with muscles. Of the 6 UAs, 2 (33%) were lobulate, 4 (67%) were regularly round or oval; 1 (17%) had bone septa, 2 (33%) had root resorption of the neighboring teeth, 3 (50%) had cortical perforation. All UAs showed well-circumscribed margins without soft tissue masses. Three UAs were examined by enhanced studies and showed mild to moderate enhancement of the wall. Of the 5 MAs, 3 (60%) showed low to intermediate density as compared with muscles, 2 (40%) showed homogenous and intermediate density; 3 (60%) had bone septa; 3 (60%) showed root resorption or destruction of the neighboring teeth. All 5 MAs were lobulate with cortical perforation, soft tissue masses, and ill-defined margins. All 3 MAs examined by enhanced studies showed strong enhancement., Conclusions: SMAs are usually seen in middle-aged patients. CT images of SMAs frequently show lobulate lesions with strongly enhanced solid and cystic components. UA occurs mostly in young patients, and typically appears as cystic lesion with mild to moderate enhancement of the wall on CT images. MA often appears in old patients, and shows extensive bone destruction, soft tissue mass, ill-defined margin, and strong enhancement on CT images. Combined analysis of CT images and clinical data is helpful in differentiating these 3 subtypes of ameloblastoma.