应用数据挖掘技术获取周围型肺癌临床和CT诊断规则的初步研究

目的探讨数据挖掘技术对周围型肺癌影像诊断规则提取的价值。方法收集58例经过临床病理证实的周围型肺癌病例,对其临床及CT表现属性进行标准化认定,输入数据库,分别采用自主开发的基于关联规则知识发现程序与通用数据分析工具ROSETTA中的粗糙集约简算法和遗传分类算法对58例周围型肺癌临床及影像学数据进行挖掘对比研究。结果由Johnson’s Algorithm粗糙集约简算法产生诊断规则51条,由ROSETTA遗传分类算法所产生的诊断规则有5千多条,基于关联规则的挖掘算法所产生的诊断规则有123条。这3种不同的数据挖掘方法产生的最重要的诊断规则基本上都将性别、年龄、位置、毛刺、形状、毛玻璃样密度等属性作为诊断周围型肺癌的主要依据。结论数据挖掘技术在医学影像诊断和鉴别诊断中具有潜在的应用价值。     

Fine mapping of X-linked clasped thumb and mental retardation (MASA syndrome) in Xq28

The MASA syndrome is an X-linked disorder with mental retardation, spastic paraparesis, and adducted thumbs as the most characteristic features. We performed linkage analysis, using Xq28 markers, on a large MASA syndrome family. The maximum lodscore was 6.37 at 0 recombination for DXS52 and 5.99 at 0 recombination for DXS305. Crossovers were demonstrated between the disorder and DXS455. Clinical and linkage data from this family further support the hypothesis that the MASA syndrome and X-linked hydrocephalus are allelic disorders.     

生理信号检测中的同步工频陷波器

市电50Hz干扰一直是生理信号检测过程中最难以克服的问题之一。本文叙述了利用同步滤波器技术克服50Hz干扰的方法。同步滤波器可以自动跟踪市电频率的变化且具有极高Q值和理想的梳状滤波特性,而元件的灵敏度却很低,因而很容易获得40dB以上的陷波效果,可以从根本上克服50Hz干扰而几乎对有用信号毫无影响。     

The use of 3D surface fitting for robust polyp detection and classification in CT colonography

In this paper we describe the development of a computationally efficient computer-aided detection (CAD) algorithm based on the evaluation of the surface morphology that is employed for the detection of colonic polyps in computed tomography (CT) colonography. Initial polyp candidate voxels were detected using the surface normal intersection values. These candidate voxels were clustered using the normal direction, convexity test, region growing and Gaussian distribution. The local colonic surface was classified as polyp or fold using a feature normalized nearest neighborhood classifier. The main merit of this paper is the methodology applied to select the robust features derived from the colon surface that have a high discriminative power for polyp/fold classification. The devised polyp detection scheme entails a low computational overhead (typically takes 2.20 min per dataset) and shows 100% sensitivity for phantom polyps greater than 5 mm. It also shows 100% sensitivity for real polyps larger than 10 mm and 91.67% sensitivity for polyps between 5 to 10 mm with an average of 4.5 false positives per dataset. The experimental data indicates that the proposed CAD polyp detection scheme outperforms other techniques that identify the polyps using features that sample the colon surface curvature especially when applied to low-dose datasets.     

Cell mediated immunity (CMI) and post transplant viral infections — Role of a functional immune assay to titrate immunosuppression

Cell mediated immunity (CMI) was assessed by the ImmuKnow assay in 12 patients after kidney transplantation, who presented with viral infection. Treatment included lowering of immunosuppression in all cases and antiviral treatment if indicated. The assay was repeated during the follow up. The ImmuKnow assay at time of presentation of viral infections was 56.8 ± 58.2 (range 3–178; median 22) ATP ng/ml. With the clearance of viral infection and lowering of immunosuppression, the assay showed an increase in the level of CMI at 194.5 ± 118.9 (range 53–409; median 150) ATP ng/ml. There was viral clearance or stabilization in all cases and there was no incidence of allograft rejection. The ImmuKnow assay of CMI can be used to titrate initial immunosuppression reduction and its subsequent increase, in patients with viral infection after transplantation.     

In vivo fluorescence spectroscopy and imaging of human skin tumours

The feasibility of using in vivo autofluorescence for the diagnosis of skin cancer was evaluated. In vivo fluorescence measurements were performed on healthy human volunteers, and patients with different types of benign and malignant skin tumours. Fluorescence spectra as well as fluorescence images were acquired. The excitation-emission matrix of normal skin (n=3) showed a broad peak at the shortest excitation wavelength (365 nm) and at 440 nm fluorescence wavelength, smoothly decreasing towards longer excitation and fluorescence wavelengths. Non-melanoma skin tumours (n=31) and control skin excited with 375 nm showed a broad fluorescence band from 400 to 700 nm, peaking around 436 nm. No significant differences in measurements between tumours and the corresponding control sites were found. A large spatial variation in the fluorescence intensity was observed both in the tumours and in the control sites. Standard deviations found ranged from 0.15 to 1.5 times the mean fluorescence. Fluorescence images, excited with 375 nm and taken with an image intensified CCD camera, on eight malignant melanomas and eight benign pigmented lesions did not indicate any fluorescence intensity distribution specific to the malignancy of the lesion. Neither the shape of the fluorescence spectra, nor the spatial distribution of the fluorescence intensity showed any signature specific to the histopathological nature of the lesions investigated. Optical diagnostics of skin tumours using the autofluorescence does not seem to be feasible at the present time.     

Comparison of redox and D29 phage methods for detection of isoniazid and rifampicin resistance in Mycobacterium tuberculosis

Rapid, accurate and inexpensive methods are essential to detect drug-resistant Mycobacterium tuberculosis and allow timely application of effective treatment and precautions to prevent transmission. The proportion method, the MTT and Alamar Blue redox methods, and the D29 mycobacteriophage assay, were compared for their ability to detect resistance to isoniazid and rifampicin. When tested against a panel of known M. tuberculosis strains, the redox methods and the D29 assay showed good sensitivity and specificity compared to the proportion method, suggesting that they could be useful alternatives for identifying multidrug resistance in M. tuberculosis.     

Breast cancer surveillance — a cost-effective strategy

Summary The effectiveness of a surveillance program for breast cancer recurrence in extending survival is predicated on two assumptions: 1) most recurrences are detected at an early stage at surveillance visits; and 2) the early treatment of recurrence offers a better chance of cure or longer survival. However, the data suggest that neither of these two assumptions is correct, and that postoperative follow-up of patients with breast cancer is expensive and does not significantly extend survival.This minisymposium was presented December 8, 1992, at the annual San Antonio Breast Cancer Symposium, and was sponsored by educational grants from Amgen and from Bristol-Myers Oncology Division.