Prone-to-supine breast tumor registration using landmarks

During the preoperative planning for breast-conserving surgery, the surgeon makes use of an MRI scan of the breast cancer patient in the prone position to accurately locate the tumour. However, surgery is performed with the patient in the supine position. The surgeon needs to mentally translate the location of the tumour from the prone position...

Non-destructive Infield Quality Estimation of Strawberries using Deep Architectures

Strawberries have a short shelf-life time and thus need to be harvested at the right time to reduce waste. To this end, information about quality attributes is useful. Recently, many computer vision methods have been proposed. Most literature analyzes postharvest, which means that strawberries can only be analyzed after harvesting. As a result,...

The effect of recurrence on adversarial robustness

Traditionally, convolutional neural networks are feedforward networks with a deep and complex hierarchy. Conversely, the human brain has a relatively shallow hierarchy with recurrent connections. Replicating this recurrence may allow for shallower and easier to understand computer vision models that may possess characteristics usually attributed...

TRIDENT: Transductive Variational Inference of Decoupled Latent Variables for Few Shot Classification

The versatility to learn from a handful of samples is the hallmark of human intelligence. Few-shot learning is an endeavour to transcend this capability down to machines. Inspired by the promise and power of probabilistic deep learning, we propose a novel variational inference network for few-shot classification (coined as TRIDENT) to decouple...

Integrated learning of mutational signatures and prediction of DNA repair deficiencies

Motivation: Many tumors show deficiencies in DNA damage repair. These deficiencies can play a role in the disease, but also expose vulnerabilities with therapeutic potential. Targeted treatments exploit specific repair deficiencies, for instance based on synthetic lethality. To decide which patients could benefit from such therapies requires the...

Improving cell type matching across species in scRNA-seq data using protein embeddings and transfer learning

Knowing the relation between cell types is crucial for translating experimental results from mice to humans. Establishing cell type matches, however, is hindered by the biological differences between the species. A substantial amount of evolutionary information between genes that could be used to align the species, is discarded by most of the...

Predicting immune responses on multi-modal single-cell data with variational inference

Single-cell sequencing allows measuring individual cells' molecular features and their responses to perturbations. Understanding which cells respond to a particular perturbation and how these responses vary across populations can be used to, for example, improve vaccine immunogenicity. However, an exhaustive exploration of single-cell...

Explaining overthinking in Multi-Scale Dense networks: Why more computation does not always lead to better results

Traditional convolutional neural networks exhibit an inherent limitation, they can not adapt their computation to the input while some inputs require less computation to arrive at an accurate prediction than others. Early-exiting setups exploit this fact by only spending as much computation as is necessary and subsequently exiting the sample...

A multi-omic approach to discovering methylation-based markers of colon cancer in cfDNA

In recent years the advent of multi-omic techniques have shown great promise in the field of oncology. In light of these advancements, this thesis focuses on the use of multiple data types to find methylation markers around transcription start site regions for colorectal cancer in the cell-free DNA (cfDNA) domain. It combines several methods of...

Explainable Survival Analysis: for Urothelial Cancer

Survival analysis is a statistical method used to predict when an event will occur. Machine learning survival models have been used in many cancer studies. However, machine learning models may not always be interpretable. The current lack of research for explainable survival analysis for urothelial cancer prompted this study. This study offers...

Dynamics of the transcriptome and proteome in regions of the brain differ considerably

Human brain research is advancing, facilitated by improved high-throughput techniques and systematically preservation of brain tissue in brain banks. Due to the complex organization and regulation of the brain, most studies only focus on a single aspect of healthy and diseased samples. This research integrates analyses of the transcriptome and...

A Bayesian approach to haplotype-aware de novo genome assembly for polyploid organisms

Capturing all genetic variation within a polyploid organism is a challenge. Most current de novo assemblers have no notion of the concept “ploidy”. Consequently, when assembling the genome of diploid or higher ploidy organisms, the assembler mixes reads coming from either chromosome copy and builds a single DNA sequence representing an arbitrary...