L2M NSERC - AI4Path for the Cancer Marker Ki-67

一种称为Ki-67的核蛋白已被确定为在癌症相关活检的预测和鉴定中发挥作用的最重要因素之一。作为细胞增殖的结果，具有高Ki-67指数的肿瘤更可能具有增殖细胞，其由于细胞增殖而具有更快生长的更大趋势。当涉及到确定Ki-67时，病理学家使用许多不同的方法，最广泛使用的方法是手动计数和视觉估计（目测），这两种方法都被广泛用作确定组织样本中Ki-67的方法。尽管如此，这是一个非常耗时和费力的过程，对人类的一部分，以实现这一目标。我们的目标是开发一种通用但完全自动化的方法，使用深度学习技术分割细胞体的显微镜Ki-67图像，并开发一个可以从世界任何地方访问的门户网站（AI 4Path），并能够用于对捕获的图像进行进一步分析。除了减少人力投入，AI 4Path还可以为我们的用户提供许多好处，例如能够以更高的可扩展性处理大型数据集，提高再现性，并允许我们以更高的再现性处理大型数据集。

A nuclear protein called Ki-67 has been identified as one of the most important factors that play a role in the prediction and prognostication of cancer-related biopsies. As a consequence of the proliferation of cells, tumors with high Ki-67 indices are more likely to have proliferating cells, which have a greater tendency to grow more rapidly as a result of cell proliferation. When it comes to determining Ki-67, there are many different methods that pathologists use, and the most widely used methods are manual counting and visual estimation (eyeballing), both of which are widely used as methods to determine Ki-67 within a tissue sample. In spite of this, it is a very time consuming and laborious process on the part of the human being to achieve this goal. Our goal is to develop a generalist yet fully automated method of segmenting microscopy Ki-67 images for cell bodies using deep learning techniques, as well as to develop a web portal (AI4Path) that can be accessed from anywhere in the world and be able to be used for further analysis of the captured images. In addition to reducing human effort, AI4Path can also provide a number of benefits to our users, such as being able to handle large datasets with a greater degree of scalability, improving reproducibility, and allowing us to handle large datasets with a higher degree of reproducibility.