Improved Automated Urinalysis

描述（由申请人提供）：目前的自动尿液分析仪器无法识别尿液沉积物样本中的晶体、铸型及其亚型。本提案旨在开发一种利用双折射现象对铸件和晶体的类型和亚型进行自动分类的技术。由此产生的成像和分类技术将允许在目前可用的自动化尿液分析仪器的量子改进。尿中的铸型和结晶是临床异常的重要指标。目前的自动化技术无法对铸件和晶体的子类型进行分类。因此，需要人工检查可疑尿样，从而浪费了时间和资源。解决方案是改进这类仪器使用的分类技术。该项目旨在开发新的光学和模式识别技术，用于尿液沉积物中铸件和晶体的微观颗粒分析。由于目前使用流式细胞术和图像分析的尿液分析仪器在分类铸型和晶体类型和亚型方面的能力有限，因此本项目特别打算增加对尿液中的铸型和晶体进行亚型分型的功能。我们的方法是使用偏振显微镜来获得新的特征，以帮助分类晶体，铸件和其他颗粒。另一个感兴趣的领域是探索这些晶体的磁性，并使用新技术通过磁场对其双折射特性的影响来检测和识别晶体。第一阶段项目的目的是证明双折射作为一种独特的“特征”的效用，可以产生对尿液颗粒进行分类的新特征。如果第一阶段的可行性标准得到满足，那么在第二阶段，我们将建立一个自动尿液颗粒分析仪，结合偏振光学、电流显微镜和数字图像处理技术。由此产生的自动化仪器将提供更快、更准确的结果，因此将成为临床诊断中尿液筛查的更重要工具。目前的自动尿液分析仪器无法识别尿液沉积物样本中的结晶、铸型及其亚型。本提案旨在开发一种利用双折射现象对铸件和晶体的类型和亚型进行自动分类的技术。由此产生的成像和分类技术将允许在目前可用的自动化尿液分析仪器的量子改进。

DESCRIPTION (provided by applicant): Current automated urinalysis instruments are unable to recognize crystals, casts, and their subtypes in urine sediment samples. This proposal seeks to develop a technique for automated classification of the types and subtypes of casts and crystals by exploiting birefringence phenomena. The resulting imaging and classifying technique will permit a quantum improvement in currently available automated urinalysis instruments. The casts and crystals in the urine are clinically important indicators of abnormalities. Current automated techniques are unable to classify the sub-types of casts and crystals. As a result, manual examination of suspect urine samples is necessary, thus wasting time and resources. The solution is to improve the classification techniques that such instruments use. This project aims to develop novel optical and pattern recognition techniques for microscopic particle analysis of casts and crystals in urine sediment. Since current urinalysis instruments using flow cytometry and image analysis are limited in their ability to classify the types and subtypes of casts and crystals, this project specifically intends to add functionality for sub-typing casts and crystals in urine. Our approach is to use polarization microscopy to obtain new features to aid classification of crystals, casts and other particles. Another area of interest is to explore the magnetic properties of these crystals and to use novel techniques to detect and identify the crystals by the influence that magnetic fields have on their birefringence properties. The aim of the Phase I project is to demonstrate the utility of birefringence as a unique 'signature' that can give rise to new features for classifying urine particles. If the Phase I feasibility criteria are met, then in Phase II we will build an automated urine particle analyzer incorporating polarization optics combined with current flow microscopy and digital image processing techniques. The resulting automated instrument will provide quicker and more accurate results, and will thus become an even more important tool for urine screening in clinical diagnostics. Current automated urinalysis instruments are unable to recognize crystals, casts, and their subtypes in urine sediment samples. This proposal seeks to develop a technique for automated classification of the types and subtypes of casts and crystals by exploiting birefringence phenomena. The resulting imaging and classifying technique will permit a quantum improvement in currently available automated urinalysis instruments.