Quantitative phase microscopy of thick objects

显微镜对厚标本的成像一直存在问题。问题是，当显微镜工作者试图聚焦到这类物体深处的平面时，图像会被样本其余部分产生的散焦背景所淹没--要求的分辨率越高，问题就越严重。最简单的解决方案是将样本切成非常薄的物理切片，但这既耗时又具有破坏性，会破坏样本的原始状态。这项研究的目的是研究在不需要物理改变标本的情况下实现相同切片效果的计算方法。我们假设，这可以使用一种相对较新的称为层析成像的技术，结合样品旋转的思想和层析成像的算法来实现。通过解开照明源与厚样品的复杂、多次散射相互作用，我们希望将这种混合方法的3D成像能力扩展到对这两种技术的传统形式施加的限制。这项研究的潜在应用跨越了显微镜模式的范围。在红外波长检测半导体晶片中的微小缺陷是一个令人兴奋的应用，目前超出了现有的基于光的检测设备的范围-最终结果将是更好地理解光刻工艺和更高的成品率。对于可见光，一个主要的例子是厚组织样本的成像，这是生物学家长期以来的一个难题，他们将从以微米级分辨率进行无斑点活体成像的能力中受益匪浅。在光谱的X射线端，应用包括骨碎片的腔隙-管状网络的成像和软X射线全细胞成像，如果能够超越当前成像方法固有的厚度限制，则精度和可实现的分辨率都将大大提高。对于电子，我们的研究可以改善对半导体关键材料性质的成像，如掺杂浓度和磁场，对这些性质的测量对未来电子器件的发展至关重要。

Microscopes have always had trouble imaging thick specimens. The problem is that when the microscopist tries to focus on a plane deep within such objects, the image is swamped by out-of-focus background generated by the rest of the sample - and the higher the resolution required the worse this problem gets. The simplest solution is to physically slice the specimen very thinly, but this is time consuming and disruptive, destroying the sample's native state. The aim of this research is to investigate computational methods for achieving the same slicing effect, without the need to physically alter the specimen. We hypothesise that this can be achieved using a relatively new technique called ptychography, in combination with the sample rotation idea and algorithms from tomography. By unravelling the complex, multiple-scattering interaction of an illumination source with a thick specimen, we hope to stretch the 3D imaging capabilities of this hybrid approach beyond the limits imposed upon the conventional forms of these two techniques. Potential applications for the research stretch across the spectrum of microscope modalities. At infrared wavelengths detecting small defects in semiconductor wafers is an exciting application, currently beyond the reach of existing light-based inspection devices - the end result would be improved understanding of the lithographic process and better yields. For visible light, a primary example is the imaging of thick tissue specimens, a long-standing bugbear for biologists, who would benefit greatly from the ability to perform stain-free in vivo imaging at micron-scale resolutions. At the X-ray end of the spectrum, applications include the imaging of the lacuna-canalicular network of bone fragments and soft X-ray whole cell imaging, where both accuracy and achievable resolution will be greatly improved if the thickness limit inherent to the current imaging methods can be surpassed. For electrons, our research could improve the imaging of key material properties of semiconductors, such as doping concentrations and magnetic fields, measurement of which is vital to the development of future electronic devices.