Object detection in medical images using deep learning is a challenging task, due to the imbalance often present in the data. Deep learning algorithms require large amount of balanced data to achieve optimal performance, as well as close monitoring and ne-tuning of hyper parameters. For most applications, such performance monitoring is done by simply feeding unseen data trough the network, and then using the loss function for evaluation. In the case of small or sparse objects, the loss function might not able to describe the features needed, but such features can be hard to capture in a loss function. In this paper we introduce a lesion-wise whole volume validation tool, which allows more a more accurate performance monitoring of segmentation of small and sparse objects. We showcase the efficacy of our tool by applying it to the task of microbleed segmentation, and compare the behaviour of lesionwise-whole volume validation compared to well known segmentation loss functions. Microbleeds are visible as small (less than 10 mm), ovoid hypo-intensities on T2*-weighted and susceptibility weighted magnetic resonance images. Detection of microbleeds is clinically relevant, as microbleeds can indicate the risk of recurrent stroke, and are used as imaging biomarker for various neurodegenerative diseases. Manual detection or segmentation is time consuming and error prone, and suffers from high inter- and intraobserver variability. Due to the sparsity and small size of the lesions, the data is severely imbalanced.
Prion diseases are a group of progressive neurodegenerative conditions which cause cognitive impairment and neurological deficits. To date, there is no accurate measure that can be used to diagnose this illness, or to quantify the evolution of symptoms over time. Prion disease, due to its rarity, is in fact commonly mistaken for other types of dementia. A robust tool to diagnose and quantify the progression of the disease is key as it would lead to more appropriately timed clinical trials, and thereby improve patients’ quality of life. The approaches used to study other types of neurodegenerative diseases are not satisfactory to capture the progression of human form of Prion disease. This is due to the large heterogeneity of phenotypes of Prion disease and to the lack of consistent geometrical pattern of disease progression. In this paper, we aim to identify and select imaging biomarkers that are relevant for the diagnostic on Prion disease. We extract features from magnetic resonance imaging data and use genetic and demographic information from a cohort affected by genetic forms of the disease. The proposed framework consists of a multi-modal subjectspecific feature extraction step, followed by a Gaussian Process classifier used to calculate the probability of a subject to be diagnosed with Prion disease. We show that the proposed method improves the characterisation of Prion disease.
Dealing with pathological tissues is a very challenging task in medical brain segmentation. The presence of pathology can indeed bias the ultimate results when the model chosen is not appropriate and lead to missegmentations and errors in the model parameters. Model fit and segmentation accuracy are impaired by the lack of flexibility of the model used to represent the data. In this work, based on a finite Gaussian mixture model, we dynamically introduce extra degrees of freedom so that each anatomical tissue considered is modelled as a mixture of Gaussian components. The choice of the appropriate number of components per tissue class relies on a model selection criterion. Its purpose is to balance the complexity of the model with the quality of the model fit in order to avoid overfitting while allowing flexibility. The parameters optimisation, constrained with the additional knowledge brought by probabilistic anatomical atlases, follows the expectation maximisation (EM) framework. Split-and-merge operations bring the new flexibility to the model along with a data-driven adaptation. The proposed methodology appears to improve the segmentation when pathological tissue are present as well as the model fit when compared to an atlas-based expectation maximisation algorithm with a unique component per tissue class. These improvements in the modelling might bring new insight in the characterisation of pathological tissues as well as in the modelling of partial volume effect.