An analysis of the influence of transfer learning when measuring the tortuosity of blood vessels

Background and Objective: Convolutional Neural Networks (CNNs) can provide excellent results regarding the segmentation of blood vessels. One important aspect of CNNs is that they can be trained on large amounts of data and then be made available, for instance, in image processing software. The pre-trained CNNs can then be easily applied in downstream blood vessel characterization tasks, such as the calcula-tion of the length, tortuosity, or caliber of the blood vessels. Yet, it is still unclear if pre-trained CNNs can provide robust, unbiased, results in downstream tasks involving the morphological analysis of blood ves-sels. Here, we focus on measuring the tortuosity of blood vessels and investigate to which extent CNNs may provide biased tortuosity values even after fine-tuning the network to a new dataset under study. Methods: We develop a procedure for quantifying the influence of CNN pre-training in downstream anal-yses involving the measurement of morphological properties of blood vessels. Using the methodology, we compare the performance of CNNs that were trained on images containing blood vessels having high tortuosity with CNNs that were trained on blood vessels with low tortuosity and fine-tuned on blood vessels with high tortuosity. The opposite situation is also investigated.Results: We show that the tortuosity values obtained by a CNN trained from scratch on a dataset may not agree with those obtained by a fine-tuned network that was pre-trained on a dataset having different tortuosity statistics. In addition, we show that improving the segmentation accuracy does not necessarily lead to better tortuosity estimation. To mitigate the aforementioned issues, we propose the application of data augmentation techniques even in situations where they do not improve segmentation performance. For instance, we found that the application of elastic transformations was enough to prevent an under-estimation of 8% of blood vessel tortuosity when applying CNNs to different datasets.Conclusions: The results highlight the importance of developing new methodologies for training CNNs with the specific goal of reducing the error of morphological measurements, as opposed to the traditional approach of using segmentation accuracy as a proxy metric for performance evaluation.(c) 2022 Elsevier B.V. All rights reserved.