Measuring Cancer Prognosis with Self-Supervised Learning

关键字：人工智能，结肠癌，建模，转移预测潜在癌症生长患者的预后是极其困难的。经典的监督式机器学习需要具有预先评分的真实值标签的大型数据集，然而，对于许多癌症和癌前病变（如结肠直肠息肉）来说，这些数据集根本不存在。本项目将使用一套最先进的自监督机器学习技术来研究病理染色中是否存在尚未发现的特征，这些特征可用于预测癌症预后。这些技术需要的数据比有监督的方法少得多，也没有数据标签，使它们免受人类先入为主、错误和偏见的影响。该项目将重点关注癌症，然而，所开发的方法将广泛适用于医学成像的其他领域。进一步的工作将研究多种不同数据模式的整合，特别是组学数据，以及从不同放大倍数的数字病理切片中提取信息的方法。

Studentship strategic priority area: Mathematics, statistics and computationKeywords: Artificial Intelligence, colon cancer, modelling, metastasisPredicting the prognosis of a patient with potentially cancerous growth is extremely difficult. Classical supervised machine learning requires large datasets with pre-scored ground truth labels, however these simply do not exist for many cancers and pre-cancerous growths, such as colorectal polyps. This project will use a set of state-of-the-art self-supervised machine learning techniques to investigate whether there exist as yet undiscovered features in pathology stain which can be used to predict cancer prognosis. These techniques require far less data than supervised methods, and no data labelling, leaving them free from human preconception, error, and bias.The project will focus on cancer, however, the methods developed will be broadly applicable to other areas of medical imaging. Further work will investigate the integration of multiple different data modalities, particularly -omics data, and methods of extracting information from digital pathology slides at different magnifications.