Volume 3 Issue 6
Authors: R.S. Houmsou; E.U., Amuta; B.E., Wama; T.D., Hile
Abstract: Proteinuria is an indirect marker of urinary schistosomiasis in endemic areas but cases of misdiagnosis could occur in areas where other diseases that produce proteinuria overlap with urinary schistosomiasis. The study was conducted to assess whether proteinuria measured by urine reagent strips could be used as a rapid screening technique of urinary schistosomiasis in onchocerciasis endemic areas. Urine reagent strips and filtration technique were used to screen proteinuria and to determine Schistosoma haematobium eggs in 1,124 school children respectively. Our findings demonstrate that proteinuria positively correlated with intensity of infection (rho=0.71, p<0.01), highly sensitive (95.7%) and moderately specific (67.7%). We suggest that measuring proteinuria using urine reagent strips may be useful for diagnostic purposes, monitoring and evaluating urinary schistosomiasis in onchocerciasis endemic areas. Furthermore, we recommend that additional research should be done to further elucidate the relationship between proteinuria and pathological changes of the urinary tract that may be attributed to infection with Schistosoma haematobium in onchocerciasis endemic areas.
Keywords: Proteinuria; Onchocerciasis; Schistosomiasis; Filtration; Strips
Authors: Lilia Lazli; Mounir Boukadoum
Abstract: When the feature space undergoes changes, owing to different operating and environmental conditions, multi-aspect classification is almost a necessity in order to maintain the performance of the pattern recognition system and improve robustness and reliability in decision making. This is an important issue being investigated in ANN research, in many cases, the problems can be solved more effectively by combining one or two other techniques rather than implementing ANN exclusively. New learning methods, especially multiple classifier systems, are now actively studied and applied in pattern recognition. So, the main goal of this paper is to propose two hybrid models and compare your performance in complex pattern recognition problem: speech recognition and biomedical diagnosis. This paper compare, the performance obtained with (1) Multi-network RBF/LVQ structure, we use involves Learning Vector Quantization (LVQ) as a competitive decision processor and Radial Basis Function (RBF) as a classifier. (2) Hybrid HMM/MLP model using a Multi Layer-Perceptron (MLP) to estimate the Hidden Markov Models (HMM) emission probabilities. For pre-classification, the k-means clustering algorithm is proposed to obtain optimum information for the biomedical and speech training data for the proposed hybrid models.
Keywords: Medical Diagnosis; Speech Recognition; K-means Clustering; Hybrid RBF/LVQ Model; Hybrid HMM/MLP Model
Authors: Baosen Zhou; Zhihua Yin; Zhongxing Chen; Wei Wu; Ling Zhou; Yong Cui; Junqiao Guo
Abstract: The study was to determine the prevalence, types and epidemiologic patterns of birth defects in the vicinity of Hongyanhe nuclear power plant between 1996 to 2009. Hongyanhe nuclear power plant is the first nuclear power plant in northeast China and the first one receiving central government approval to build four units at the same time. We used the database of the Liaoning Birth Defects Monitoring Network to investigate birth outcomes for infants with defects from 1996 to 2009, in Wafangdian. The infant must have been born in there with a birth diagnosed within 7 days after delivery, and the mothers living in Wafangdian for more than one year were enrolled. During 1996 to 2009, a total of 783 birth defects were identified among 95,632 births, which yielded a prevalence of 81.88 per 10,000 births. The highest prevalences of birth defects was observed in 2007 (128.98 per 10,000 births), and the lowest one was in 1996 (42.24 per 10,000 births). There was an upward trend detected in the time span. The major types of birth defects in our study were: congenital heart disease, hydrocephalus, cleft lip and cleft palate, polydactyly, spina bifida and cleft lip. All of the prevalences were lower than the national prevalence and other studies. The major types of birth defects in our study are similar with the types of national. The prevalences of NTDs and trisomy 21 thought to be caused by ionizing radiation are lower than other studies of China. To our knowledge, few studies about prevalence of birth defects have the large intervals of time, from 1996 to 2009 in northeast China.
Keywords: Nuclear Power Plant; Birth Defects; Prevalence
Authors: Antonis I. Christopoulos
Abstract: Coexistent lung cancer (LC) and pulmonary tuberculosis (TB) were sources of clinical and pathological interest and speculation for many years. Nowadays, there is compelling evidence that TB increases the risk for LC. Furthermore, active TB has been reported to complicate LC. It is well documented that lung inflammation and fibrosis could induce genetic damage, increasing the risk for LC. The patients usually have already suffered from TB when cancer was diagnosed. In such a case, the disease was confined to the upper lobes, with or without cavities, or was spread to the lymph nodes. Whereas, when TB appeared later in the course of the cancer, after anticancer chemotherapy, TB was extended and disseminated. Clinical/radiological characteristics of TB and LC resembled each other. It is extremely difficult to verify the diagnosis of active TB from clinical and/or and radiological features in LC patients. Definite diagnosis requires pathological confirmation from biopsies and microbiologic studies. TB treatment in an immune compromised host requires individualized therapeutic decisions. There are no significant differences in TB treatment responses and/or toxicity of the anti-tuberculosis medication when combined with anticancer therapy. As poor survival and high mortality rates have been reported, interactions between anti-tuberculosis and anticancer medication, affecting the efficacy of the latter, deserve further studying.
Keywords: Tuberculosis; Lung Cancer; Chemotherapy; Treatment