Project

Task1: Registration 

4)Q

Don't use the same dataset for model training and model evaluation. If you want to build a reliable machine learning model, you need to split your dataset into the training, validation, and test sets. If you don't, your results will be biased, and you'll end up with a false impression of better model accuracy.

5)Q

Image registration is a fundamental task in multiple medical image analysis applications. With the advent of deep learning, there have been significant advances in algorithmic performance for various computer vision tasks in recent years, including medical image registration. The last couple of years have seen a dramatic increase in the development of deep learning-based medical image registration algorithms. Consequently, a comprehensive review of the current state-of-the-art algorithms in the field is timely, and necessary. This review is aimed at understanding the clinical applications and challenges that drove this innovation, analysing the functionality and limitations of existing approaches, and at providing insights to open challenges and as yet unmet clinical needs that could shape future research directions. To this end, the main contributions of this paper are: (a) discussion of all deep learning-based medical image registration papers published since 2013 with significant methodological and/or functional contributions to the field; (b) analysis of the development and evolution of deep learning-based image registration methods, summarising the current trends and challenges in the domain; and (c) overview of unmet clinical needs and potential directions for future research in deep learning-based medical image registration.

As discussed previously, medical image registration differs greatly from other medical image analysis tasks. Future research should introduce more registration priors to DL networks, making DL networks more specific to image registration, and more application specific. To improve registration performance DLIR networks could be imbued with prior information related to the expected type of deformation, spatial relationship between anatomical structures, and the topology and morphology of anatomical structures. For example, although ground-truth spatial transformations are seldom available, other labels could serve as the ground-truth to guide the training process. Several methods on weakly-supervised image registration have been proposed, which generally achieve better performance than its corresponding unsupervised variant (at no additional cost in terms of execution speed). More informative priors combined with training data that is synthetically modified, such as, blackening pixels in the moving image, or generating adversarial examples, could enhance the ability of networks to generalise to unseen data, while remaining robust to variable image quality. Consequently, combining different types of spatial and temporal priors with DL networks is a promising direction for future research in the field.