Kimberley M Timmins, Maarten J Kamphuis, Iris N Vos, Birgitta K Velthuis, Irene C van der Schaaf and Hugo J Kuijf. Future Unruptured Intracranial Aneurysm Growth Prediction using Mesh Convolutional Neural Networks
Abstract. The growth of unruptured intracranial aneurysms (UIAs) is a predictor of rupture. Therefore, for further imaging surveillance and treatment planning, it is important to be able to predict if an UIA is likely to grow based on an initial baseline Time-of-Flight MRA (TOFMRA). It is known that the size and shape of UIAs are predictors of aneurysm growth and/or rupture. We perform a feasibility study of using a mesh convolutional neural network for future UIA growth prediction from baseline TOF-MRAs. We include 151 TOF-MRAs, with 169 UIAs where 49 UIAs were classified as growing and 120 as stable, based on the clinical definition of growth (>1 mm increase in size in follow-up scan). UIAs were segmented from TOF-MRAs and meshes were automatically generated. We investigate the input of both UIA mesh only and region-of-interest (ROI) meshes including UIA and surrounding parent vessels. We develop a classification model to predict UIAs that will grow or remain stable. The model consisted of a mesh convolutional neural network including additional novel input edge features of shape index and curvedness which describe the surface topology. It was investigated if input edge mid-point co-ordinates influenced the model performance. The model with highest AUC (63.8%) for growth prediction was using UIA meshes with input edge mid-point co-ordinate features (average F1 score = 62.3%, accuracy = 66.9%, sensitivity = 57.3%, specificity = 70.8%). We present a future UIA growth prediction model based on a mesh convolutional neural network with promising results.
Clara Brémond-Martin, Camille Simon-Chane, Cédric Clouchoux and Aymeric Histace. TDA-Clustering Strategies for the Characterization of Brain Organoids
Abstract. We propose to use Topological Data Analysis (TDA) to characterize the morphological development of brain organoids. We combine TDA with clustering strategies to characterize the morphology of three developmental stages of segmented brain organoid images. We calculate a linear regression of the H1 feature diagrams as well as entropy, dispersion, and average persistence of H0 and H1 features separately for each developmental stage. We also compute pairwise average Wasserstein distances between features, and within and inter clusters for each developmental stage. To explore all feature vectors from each group, we calculate K-L divergence on the t-SNE reduction of the TDA results. Early stages are characterized by a persistence diagram with a low slope and intercept, a high entropy within features, and a high Wasserstein distance between clusters near and far from the origin. The opposite is true for later stages. K-L divergences are particularly high between midstages and early or late stages, and punctually between some clusters. Results highlight specific morphological patterns at 14 days, corresponding to neuroepithelial formations.
Liu Li, Qiang Ma, Zeju Li, Cheng Ouyang, Weitong Zhang, Anthony Price, Vanessa Kyriakopoulou, Lucilio Cordero-Grande, Antonis Makropoulos, Joseph Hajnal, Daniel Rueckert, Bernhard Kainz and Amir Alansary. Fetal Cortex Segmentation with Topology and Thickness Loss Constraints
Abstract. The segmentation of the fetal cerebral cortex from magnetic resonance imaging (MRI) is an important tool for neurobiological research about the developing human brain. Manual segmentation is difficult and time-consuming. Limited image resolution and partial volume effects introduce errors and labeling noise when attempting to automate the process through machine learning. The significant morphological changes observed during brain growth pose additional challenges for learning-based image segmentation methods, which may drastically increase the amount of necessary training data. In this paper, we propose a framework to learn from noisy labels by using additional regularization via shape priors for the accurate segmentation of the cortical gray matter (CGM) in 3D. Firstly, we introduce a novel structure consistency loss based on persistent homology analysis of the cortical topology. Secondly, a regularization loss term is proposed by integrating assumptions about the cortical thickness within each sample. Our experiments on the developing human connectome project (dHCP) dataset show that our method can predict accurate CGM segmentation learned from noisy labels