Infrared Microspectroscopy for Cervical Cancer Screening

描述（由申请人提供）：本提案旨在建立红外显微光谱（IR-MSP）作为一种比目前使用的方法更快，更便宜，更客观和更准确的检测宫颈dvplasia的技术。目前通过细胞学检查（Papanicolaou或“巴氏”试验）进行宫颈疾病筛查，该检查于1940年代末首次提出[Papanicolaou， 1948年]。尽管定期使用巴氏试验进行妇科筛查大大降低了宫颈癌的发病率和死亡率，但仍然存在相对较高的不准确或不确定诊断率[美国卫生与公众服务部，1989年]。经典细胞学是一种基于定位和识别具有改变形态的细胞作为疾病指标的视觉检查方法。形态变化是细胞突变的结果。虽然这种方法已经建立，但由于形态变化的解释是基于人类的，因此在本质上是主观的，这一事实使这种方法受到影响。此外，操作人员的疲劳、培训和经验影响细胞学诊断的质量、可靠性和可重复性。为了减轻视觉检测的固有局限性，计算机成像检测技术在过去的十年中得到了发展和应用。然而，基于计算机的图像分析并没有显著提高诊断的可靠性，也没有证明具有成本效益[Hutchinson， 1996]。IR-MSP是一种通过检测细胞组成及其变异来识别异常的新技术，而不是通过形态学特征的变化来识别异常。事实上，IR-MSP检测到形态变化的前兆。在IR-MSP中梳理出的实际测量是客观的、可量化的；因此，该诊断方法比现有方法更敏感、特异、可重复性和可靠性高。在本提案中，将开发子宫颈涂片的IR-MSP检测方法，包括纯化脱落细胞，在合适的底物上制备细胞单层，通过配备焦平面阵列探测器的红外微光谱仪对约104个细胞进行光谱成像，并通过多元统计方法分析收集的数据以达到诊断。这项工作有可能进一步减少假阳性和假阴性诊断，并引入快速和即时的常规筛查试验，从而改善宫颈癌筛查。

DESCRIPTION (provided by applicant): This proposal is aimed at establishing Infrared Microspectroscopy (IR-MSP) as a faster, less expensive, more objective and more accurate technology for the detection of cervical dvsplasia than presently used methods. Screening for cervical disease is presently carried out via a cytological procedure (the Papanicolaou, or "Pap" test), that was first proposed in the late 1940's [Papanicolaou, 1948]. Although regularly scheduled gynecological screening using the Pap test has reduced the incidence of, and mortality from, cervical cancer enormously, there remains a relatively high rate of inaccurate or inconclusive diagnoses [US Department of Health and Human Services, 1989]. Classical cytology is a visual inspection method based on locating and recognizing cells with altered morphology as an indicator of disease. The morphological changes are a consequence of cellular mutations. Although well established, this approach suffers from the fact that the interpretation of morphological change is human-based and therefore, subjective in nature. Furthermore, operator fatigue, training and experience influence the quality, reliability and reproducibility of cytological diagnoses. To alleviate the inherent limitation of visual inspection, computer based inspection by imaging technology has been developed and utilized over the past ten years. However, computer based image analysis has not significantly improved the reliability of the diagnoses, and has not proved cost effective [Hutchinson, 1996]. IR-MSP is a novel technique that recognizes abnormality by detecting cellular composition and variations therein, rather than by changes in morphological features. In fact, IR-MSP detects the precursor of the morphological change. The actual measurement carded out in IR-MSP is objective and quantifiable; consequently, the diagnostic method is more sensitive, specific, reproducible and reliable than existing methods. In this proposal, the methodology for IR-MSP testing of cervical smears will be developed, including the purification of the exfoliated cells, preparation of a cell monolayer on a suitable substrate, spectral imaging of about 104 individual cells via an infrared microspectrometer equipped with a focal plane array detector, and analysis of the collected data via multivariate statistical methods to arrive at a diagnosis. This work has the potential to improve cervical cancer screening by further reducing false positive and false negative diagnoses, and introducing fast and point-of-care routine screening tests.