Stratification of Pulmonary Nodules in Incidental Findings and Lung Cancer Screening

肺癌是最常见的癌症之一，在英国和世界范围内死亡率最高。2010年，英国诊断出约42，026例新病例，记录了34，859例肺癌死亡。2008年，全世界记录了160万新发肺癌病例和140万例死亡。在此背景下，该项目解决了在胸部计算机断层扫描（CT）扫描中对肺结节（PN）患者进行分层的巨大挑战和未满足的需求。这种发现通常可能发生在两种情况之一：作为与肺癌无关的扫描的偶然发现，例如肺栓塞检查或肺癌筛查。在这两种情况下，问题是相同的：结节在胸部CT中非常常见，因此如果足够可疑，则需要进一步检查，否则在6个月，12个月和18个月后进行随访成像。然而，大多数结节不是癌症，因此，本项目将为胸部CT结节患者开发新的基于图像的分层技术。该项目有两大目标。首先，开发新的图像处理技术，使其更有效地阅读多个胸部研究。其次，开发新的图像处理技术，并在必要时结合新的协议，以从根本上提高肺部成像的灵敏度和特异性。

Lung cancer is one of the most common cancers with the highest mortality rate both in the UK and Worldwide. In 2010, some 42,026 new cases were diagnosed in the UK and 34,859 lung cancer deaths recorded. In 2008, 1.6 million new lung cancer cases and 1.4 million deaths were recorded worldwide. Against this background, this project addresses a hugely challenging and unmet need in stratifying patients with Pulmonary Nodules (PNs), small masses in the lung, in Chest Computed Tomography (CT) scans. Such findings might typically occur in one of two situations: as incidental findings on scans unrelated to lung cancer, e.g. investigations for pulmonary embolism, or in lung cancer screening. In either situation, the problem is the same: nodules are very common in Chest CT and so either require further investigation, if sufficiently suspicious, otherwise follow-up imaging after 6, 12 and 18 months. However, most nodule are not cancers.Therefore, this project will develop new image-based stratification techniques for patients with Chest CT nodules. The project has two broad objectives. First, to develop new image processing technologies to make it much more efficient to read multiple Chest studies. Second, develop new image processing techniques coupled with new protocols where necessary to radically improve both the sensitivity and especially specificity of imaging of the lung.