A learning method for the class imbalance problem with medical data sets

In medical data sets, data are predominately composed of “normal” samples with only a small percentage of “abnormal” ones, leading to the so-called class imbalance problems. In class imbalance problems, inputting all the data into the classifier to build up the learning model will usually lead a learning bias to the majority class. To deal with this, this paper uses a strategy which over-samples the minority class and under-samples the majority one to balance the data sets. For the majority class, this paper builds up the Gaussian type fuzzy membership function and α-cut to reduce the data size; for the minority class, we use the mega-trend diffusion membership function to generate virtual samples for the class.Furthermore, after balancing the data size of classes, this paper extends the data attribute dimension into a higher dimension space using classification related information to enhance the classification accuracy. Two medical data sets, Pima Indians’ diabetes and the BUPA liver disorders, are employed to illustrate the approach presented in this paper. The results indicate that the proposed method has better classification performance than SVM, C4.5 decision tree and two other studies.     

前列腺癌等位基因失衡(英文)

目的 了解前列腺癌克隆演变(clonal evolution)过程中遗传学机制. 方法 采用激光显微切割技术从保存的石蜡包埋组织中获取基因组DNA;利用6个位于染色体8p12-21、8p22、17q21上的具有多态性的微卫星标记,对25例患者原发癌及相应转移灶中等位基因的缺失或保留进行分析. 结果 在24例可供信息的病例中,14例(58%)在原发癌及相应转移灶中所有位点均表现为相同的等位基因缺失或保留模式,而另外10例(42%)则显示不一致的等位基因缺失.这10例中有5例原发癌表现为等位基因保留而在相应转移灶则为缺失,另外5例在一个或一个以上的位点表现为原发癌等位基因缺失而在相应转移灶保留. 结论 前列腺癌在原发癌及相应转移灶遗传组成上的差异可能与其内在异质性、整体遗传不稳定性及克隆差异有关.     

严重烧伤后不同时段酸碱失衡类型变化的探讨

目的 :探讨严重烧伤病人在伤后不同时段 (伤后 1～ 3天、 4～ 1 0天、 >1 0天 )发生酸碱失衡类型的变化 ,分析其发生的原因及机理。方法 :利用新的酸碱失衡四步判断法对我科 1 990年 1月至 2 0 0 1年 1 2月住院的 1 5 3例严重烧伤病人同步检测的血气分析和电解质结果进行分析判断。结果 :(1 )全组患者除 1 6例酸碱失衡类型属正常外 ,其余 1 37例中 ,单纯性酸碱失衡类型 4 4例 ,其中以呼吸性碱中毒 (呼碱 )最多 ,2 3例 ;双重性 5 7例 ,以呼吸性碱中毒并代谢性酸中毒 (代酸 )为主 ,2 3例 ;三重性 36例 ,以呼吸性碱中毒并高阴离子间隙 (AG)代酸并代谢性碱中毒 (代碱 )为主 ,共 2 3例。 (2 )在休克期 (伤后 1～ 3天 )以单纯性酸碱失衡类型为主 ,最多见的类型是代谢性酸中毒 (1 4例 ) ,在水肿回收期 (伤后 4～ 1 0天 )与修复期 (伤后 1 0天以后 )均以双重性酸碱失衡类型为主 ,但最多见的类型分别是呼碱 (9例 )与呼碱并代碱 (5例 )。结论 :严重烧伤患者在伤后不同时段的酸碱失衡类型比较复杂 ,烧伤休克期的酸碱失衡类型最多 ,修复期最少。发生率最多的是呼碱、呼碱并代酸以及呼碱并高阴离子间隙代酸并代碱。在各时段中 ,三重性酸碱失衡并不少见 ,以呼碱并高AG代酸并代碱为主。     

心肌梗死与心包经双侧经络电阻抗失衡的相关研究

目的　探讨心肌梗死与心包经双侧经络电阻抗的相互关系 ,寻求无创伤性临床疾病的早期诊断。方法测量左右侧心包经上的内关穴与三阴交穴位之间的电流波形 ,计算其电阻抗值。结果　观察组显示两侧电阻抗差别有显著性意义 (P <0 0 1) ;对照组两侧差别无显著性意义 (P >0 0 5 )。结论　正常人双侧经络电阻抗的对称性、经络循行线的低阻性及双侧经络电阻抗失衡与其络属脏器病变有着密切的相关性 ,此项测定可能为心脏疾病提供一种新的早期诊断的方法。     

BOTULINUM TOXIN IN OPHTHALMOLOGY

Alan B. Scott selected, researched and developed Type A Botulinum toxin for clinical use in ophthalmology. This unique drug has proved invaluable for treatment of a number of conditions which are difficult to treat in ophthalmology and in a variety of other disciplines. The indications, methods and problems of its use are described and the results of treatment of 133 patients are discussed. Up to December 1986 over 13 000 patients have been treated in a multicentre international trial without significant complications.     

Outer Breast Quadrants Demonstrate Increased Levels of Genomic Instability

Background: Theory holds that the upper outer quadrant of the breast develops more malignancies because of increased tissue volume. This study evaluated genomic patterns of loss of heterozygosity (LOH) and allelic imbalance (AI) in non-neoplastic tissues from quadrants of diseased breasts following mastectomy to characterize relationships between genomic instability and the propensity for tumor development.Methods: Tissues from breast quadrants were collected from 21 patients with various stages of breast carcinoma. DNA was isolated from non-neoplastic tissues using standard methods and 26 chromosomal regions commonly deleted in breast cancer were examined to assess genomic instability.Results: Genomic instability was observed in breast quadrants from patients with ductal carcinomas in situ and advanced carcinomas. Levels of instability by quadrant were not predictive of primary tumor location (P = .363), but outer quadrants demonstrated significantly higher levels of genomic instability than did inner quadrants (P = .017). Marker D8S511 on chromosome 8p22– 21.3, one of the most frequently altered chromosomal regions in breast cancer, showed a significantly higher level of instability (P = .039) in outer compared with inner quadrants.Conclusions: Non-neoplastic breast tissues often harbor genetic changes that can be important to understanding the local breast environment within which cancer develops. Greater genomic instability in outer quadrants can partially explain the propensity for breast cancers to develop there, rather than simple volume-related concepts. Patterns of field cancerization in the breast appear to be complex and are not a simple function of distance from a developing tumor.     

Genomic patterns of allelic imbalance in disease free tissue adjacent to primary breast carcinomas

Mammary stroma plays an important role in facilitating the neoplastic transformation of epithelial cells, modulating integrity of the extracellular matrix, and maintaining genomic stability, but molecular mechanisms by which stroma affects epithelial structure and function are not well-defined. We used laser-assisted microdissection of paraffin-embedded breast tissues from 30 patients with breast disease and a panel of 52 microsatellite markers defining 26 chromosomal regions to characterize genomic patterns of allelic imbalance (AI) in disease-free tissue adjacent to sites of breast disease and to define genomic regions that may contain genes associated with early carcinogenic processes. The mean frequency of AI in histologically normal tissue adjacent to the primary carcinomas (15.4) was significantly higher than that in distant tissue from the same breast (3.7). The pattern of AI across all chromosomal regions differed between the adjacent tissue and primary tumor in every case. Unique AI events, observed only in tumor (15 of informative markers) or only in adjacent cells (10 of informative markers), were far more common than AI events shared between tumor and adjacent cells (~ 4). Levels of AI characteristic of advanced invasive carcinomas were already present in non-invasive ductal carcinomas in situ, and appreciable levels of AI were observed in adjacent non-neoplastic tissue at all pathological stages. Chromosome 11p15.1 showed significantly higher levels of AI in adjacent cells (p < 0.01), suggesting that this region may harbor genes involved in breast cancer development and progression. Our data indicate that genomic instability may be inherently greater in disease-free tissue close to developing tumors, which may have important implications for defining surgical margins and predicting recurrence.     

Temporary confusion: a fearful experience

Older people admitted to hospital often develop acute temporary confusion. Earlier studies have focused on problems in providing care to the confused, causes of the confusion and the key aspects of caring. The aims of this study were to describe, from a nursing perspective, how older people experienced the phenomenon of temporary confusion and to describe the older peoples' reasoning when they described their experiences. Interviews with five older informants who had recovered from their confusional state were conducted and analysed using the phenomenological method. Two phenomena were studied, 'being temporarily confused' and 'reasoning about experiences of temporary confusion', each was found to be represented by four inter-related constituents. The phenomena can be understood as aspects of suffering and power imbalance between staff and older people.     

Allelic imbalance in hereditary and sporadic prostate cancer

BACKGROUND: In this study, we evaluate the pattern of allelic imbalance (AI) in both sporadic prostate cancer (SPC) and hereditary prostate cancer (HPC) at loci that frequently show allelic imbalance in sporadic prostate cancer, or are believed to have a putative role in the disease. METHODS: DNA obtained from 35 sporadic tumors and 46 hereditary tumors were tested for AI, by using a panel of 35 microsatellite markers. RESULTS: Chromosomal regions that display high frequencies of AI (>or=30%) in HPC include 1q, 5q, 7q, 8p, 13q, 16q, 17q, 18q, and 20q. In SPC, high frequencies of AI were found at 5q, 7q, 8p, 10q, 13q. Main differences (delta >or= 20%) in AI between HPC and SPC were at 1q, 10q, 17q, 18q, and 20q. CONCLUSION: AI at the prostate cancer susceptibility loci HPC1, PCaP, and HPC20 was seen more often in HPC compared with SPC. It appears that there are marked differences in the pattern of AI between sporadic and hereditary PCa.