Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

In several fields of application, multi-gigapixel images must be analysed to gather information and take decision. This analysis is often performed manually, which is a tedious task given the volume of data to process. For instance, in cytology, branch of medical sciences which focuses on study of cells, cytopathologists analyse cell samples microscope slides in order to diagnose diseases such as cancers. Typically, malignancy is assessed by the presence or absence of cells with given characteristics. In geology, climate variations can be analysed by studying the concentration 
of micro-organisms in core samples. The concentration is usually evaluated by smearing the samples onto microscope glass slides and counting those micro-organisms.

In those situations, computer sciences and, especially, machine learning and image processing provide a great alternative to a pure-human approach as they can be used to extract relevant information automatically. Especially, those kind of problems can be expressed as object detection and classification problems.

This thesis presents the elaboration and assessment of a generic framework, extit{SLDC}, for object detection and classification in multi-gigapixel images. Especially, this framework provides implementers with a concise way of formulating problem dependent-components of their algorithm (i.e. segmentation and classification) while it takes care of problem-independent concerns such as parallelization and large image handling. 

The performances of the framework are then assessed on a real-world problem, thyroid nodule malignancy. Especially, a workflow is built to detect malignant cells in thyroid cell samples whole-slides.

Results are promising: the effective processing time for an image containing 8 gigapixels is less than 10 minutes. In order, to further reduce this execution time, some improvements are proposed.

The framework implementation can be found on GitHub: https://github.com/waliens/sldc.

machine learning --- cytomine --- image processing --- cytology --- object detection --- Ingénierie, informatique & technologie > Sciences informatiques

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

In many biomedical applications, manual annotations of whole slide images take a tremendous amount of time. In the computer vision literature, semi-automatic tools using deep learning, known as deep interactive learning, have emerged to speed up the annotation process. These semi-automatic tools exploit the interactions of the annotators in various forms to produce the annotations more rapidly. In recent years, deep interactive learning seems to gain more attention for its performance. However, do the additional information provided by the annotators help to improve the results of automatic tools? An exploration in the literature was made, resulting in the finding of a promising architecture, named NuClick, which uses the scribbles of the annotators in combination with the images to produce decent annotations more quickly. In this thesis, results of the conducted experiments on various datasets show that the additional information provided by the scribbles improve drastically the performance of the segmentation for tissues, such as bronchi, glands, or infiltrations. However, this interactive approach fails to produce accurate segmentation for more complex tissues, such as tumours or inflammations. Also, results indicate that the quality of the scribbles highly influences the produced segmentation. Therefore, care should be taken when the annotators scribble the objects of interest. These results tend to support the benefit that can be gain from the interactions of the annotators, although this thesis shows that there is room for improvements with these semi-automatic tools.

Cytomine --- Pathology --- Deep Learning --- Computer Vision --- Image Segmentation --- Ingénierie, informatique & technologie > Sciences informatiques

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

In the biomedical research field, and more precisely the field of morphometric studies, the detection of anatomical landmarks is a crucial step in order to quantify the shape and size of an object under study. The annotations of theses landmarks is quite laborious and often requires dedicated human expertise. Thus the use of automatic landmark detection techniques using artificial intelligence began to gain importance.

This thesis explores different kinds of approach in order to tackle the problem on butterflies from the Morpho genus. Each butterfly possesses two types of landmarks: true landmarks and semi-landmarks, and has two views: ventral and dorsal, resulting in a total of 4 distinct datasets. Moreover, the Morpho genus contain about 30 species which induces a great variety between the images.

Several approaches have been experimented, namely computer vision, machine learning and deep learning. The results shows that the deep learning approach outperforms the others in most cases. Nevertheless the machine learning approach has proven its performance on smaller part of the data, extit{i.e.} datasets composed of one specie only. Unfortunately, the computer vision approach did not lead to any convincing results.

Finally, this work presents the Cytomine application that has been built along with the most consistent model from the experimentations. This application provides a user-friendly interface for both training and predicting with a new model.

Machine Learning --- Deep Learning --- Automatic Landmark Detection --- Landmark --- Morpho --- Cytomine --- Ingénierie, informatique & technologie > Sciences informatiques

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Multimodal imaging analyses are large scale work, combining experience from many professionals in different disciplines, providing different modalities (i.e. data produced by an experiment) linked together.
The growth in multimodal analyses induces a demand for software to make some workflows possible, or help automate some other workflows, at lest partially.

Mass Spectrometry Imaging (MSI), although more than 50 years old, continues to see some development in the data processing domain, particularly with machine learning and deep learning applications, where some approaches tackle the preprocessing and the analysis of MSI datasets.
The analysis performed on MSI data greatly contributes from multimodal studies, providing a spatial distribution for the molecular content of the sample, thus adding valuable information to the study.

Multimodal analyses currently lack an open, collaborative web platform : such tools would allow for a greater share of experience thanks to the collaborative aspect, enable reproducibility because the analyses would run in the cloud, always on the same hardware, and the results would be available to all.
Such tools are being developed : Cytomine aims to add more effective multimodal tools to improve its capabilities, but integrating MSI data is not trivial.

The analysis of MSI data is not an easy task : file formats for this kind of data are abundant, but often vendor specific. imzML is an open effort to unify all these formats, which is supported by many pieces of software already. However, imzML is not the most appropriate format as its structure is very different from most imaging data format, making it ill-suited for visualization applications such as in Cytomine.

This master's thesis introduce a new, versatile and open format based on OME-Zarr, which is suitable for many modalities, including MSI. This file format is benchmarked against imzML to show its potential in server applications, such as Cytomine.

In addition to the new file format, the developed pieces of software includes a convertor from imzML and some preprocessing tools designed for the file format. Using the developed file format, a machine learning workflow classifies spectra from a multimodal dataset with label coming from other modalities, and provide a list of important features as a mean of interpretation.

While these pieces of software are currently developed to be run on a local machine, they lay the ground for cloud based application that can be integrated with Cytomine.

multimodal --- mass spectrometry imaging --- msi --- cytomine --- template matching --- machine learning --- image registration --- bioimaging --- Ingénierie, informatique & technologie > Sciences informatiques