DS6, Deformation-Aware Semi-Supervised Learning: Application to Small Vessel Segmentation with Noisy Training Data

Abstract

Blood vessels of the brain provide the human brain with the required nutrients and oxygen. As a vulnerable part of the cerebral blood supply, pathology of small vessels can cause serious problems such as Cerebral Small Vessel Diseases (CSVD). It has also been shown that CSVD is related to neurodegeneration, such as Alzheimer’s disease. With the advancement of 7 Tesla MRI systems, higher spatial image resolution can be achieved, enabling the depiction of very small vessels in the brain. Non-Deep Learning-based approaches for vessel segmentation, e.g., Frangi’s vessel enhancement with subsequent thresholding, are capable of segmenting medium to large vessels but often fail to segment small vessels. The sensitivity of these methods to small vessels can be increased by extensive parameter tuning or by manual corrections, albeit making them time-consuming, laborious, and not feasible for larger datasets. This paper proposes a deep learning architecture to automatically segment small vessels in 7 Tesla 3D Time-of-Flight (ToF) Magnetic Resonance Angiography (MRA) data. The algorithm was trained and evaluated on a small imperfect semi-automatically segmented dataset of only 11 subjects; using six for training, two for validation, and three for testing. The deep learning model based on U-Net Multi-Scale Supervision was trained using the training subset and was made equivariant to elastic deformations in a self-supervised manner using deformation-aware learning to improve the generalisation performance. The proposed technique was evaluated quantitatively and qualitatively against the test set and achieved a Dice score of 80.44 ± 0.83. Furthermore, the result of the proposed method was compared against a selected manually segmented region (62.07 resultant Dice) and has shown a considerable improvement (18.98%) with deformation-aware learning.

DDoS-UNet: Incorporating temporal information using Dynamic Dual-channel UNet for enhancing super-resolution of dynamic MRI

“…The UNet architecture (Ronneberger et al, 2015), including its 3D version (C ¸ic ¸ek et al, 2016), is a versatile neural network consisting of two paths: contraction and expansion. Originally proposed for image segmentation, different flavours of UNet have been developed and deployed in plenty of applications such as image segmentation (Milletari et al, 2016;Zhou et al, 2018;Oktay et al, 2018;Chatterjee et al, 2020a), audio source separation (Jansson et al, 2017;Stoller et al, 2018;Choi et al, 2019) and image reconstruction (Hyun et al, 2018;Iqbal et al, 2019). 3D UNet and its variants have been used for MR super-resolution as well (Pham et al, 2019;Sarasaen et al, 2021;Chatterjee et al, 2021b).…”

“…Following the hypothesis of using batch size one to be able to learn an exact mapping function between the specific pair of low-high-resolution images (Chatterjee et al, 2021a), batch size during training and inference in this research was also set to one. The loss during training was calculated using perceptual loss (Johnson et al, 2016), with the help of a perceptual loss network Chatterjee et al (2020a), and was minimised using the Adam optimiser with a learning rate of 10 −4 for 100 epochs. The code of the implementation is available on GitHub: https://github.com/soumickmj/DDoS.…”

UNet Deep Learning Architecture for Segmentation of Vascular and Non-Vascular Images: A Microscopic Look at UNet Components Buffered With Pruning, Explainable Artificial Intelligence, and Bias

“…The proposed architecture provided posthoc interpretability and explainability methods and incorporates all libraries related to interpretability and explainability like LIME, SHAP and TorchRay and extended to apply on 2D and 3D deep learning models for images. Authors used the segmentation model from DS6 [210] paper and the models were UNet,UNet-MSS(multi-scale supervision) and UNet-MSS with deformation. In order to evaluate proposed architecture for segmentation model, vessel segmentation was chosen.…”

Liver Segmentation in Time-resolved C-arm CT Volumes Reconstructed from Dynamic Perfusion Scans using Time Separation Technique

“…Unlike typical deep learning models, this model receives the input image in the original scale and in three downsampled scales, which are supplied in the inner encoding blocks. Similarly, the output of the model is compared at different scales as well -known as deep supervision [15] or multi-scale supervision [16]. The final output of the model (output from the final decoding block), along with three more outputs from the inner decoding blocks, are compared against the original ground-truth (used in [7]), as well as three downscaled versions of the ground-truth, respectively.…”