Doctoral Thesis

Doctoral thesis

Data-driven Representation Learning from Histopathology Image Databases to Support Digital Pathology Analysis

[ thesis | slides ]

Abstract

Cancer research is a major public health priority in the world due to its high incidence, diversity and mortality. Despite great advances in this area during recent decades, the high incidence and lack of specialists have proven that one of the major challenges is to achieve early diagnosis. Improved early diagnosis, especially in developing countries, plays a crucial role in timely treatment and patient survival. Recent advances in scanner technology for the digitization of pathology slides and the growth of global initiatives to build databases for cancer research have enabled the emergence of digital pathology as a new approach to support pathology workflows. This has led to the development of many computational methods for automatic histopathology image analysis, which in turn has raised new computational challenges due to the high visual variability of histopathology slides, the difficulty in assessing the effectiveness of methods (considering the lack of annotated data from different pathologists and institutions), and the need of interpretable, efficient and feasible methods for practical use. On the other hand, machine learning techniques have focused on exploiting large databases to automatically extract and induce information and knowledge, in the form of patterns and rules, that allow to connect low-level content with its high-level meaning. Several approaches have emerged as opposed to traditional schemes based on handcrafted features for data representation, which nowadays are known as representation learning. The objective of this thesis is the exploration, development and validation of precise, interpretable and efficient computational machine learning methods for automatic representation learning from histopathology image databases to support diagnosis tasks of different types of cancer. The validation of the proposed methods during the the- sis development allowed to corroborate their capability in several histopathology image analysis tasks of different types of cancer. These methods achieve good results in terms of accuracy, robustness, reproducibility, interpretability and feasibility suggesting their potential practical application towards translational and personalized medicine.

Thesis proposal

An Automatic Knowledge Discovery Strategy In Biomedical Images

[ proposal | slides ]

Abstract

The growing on variety, volume and velocity of public biomedical databases in the last years have generate an explosion of big data in biology and medicine. Most of these databases comprise structural, molecular and genetic information from different kind of images acquisition modalities and associated metadata having a great potential, not yet exploited, as a source of information and knowledge which could impact biomedical research in different application fields. In fact, new research areas are emerging in this direction, known as bioimage informatics and computational pathology, which are areas basically attempting to apply different methods of image processing, pattern recognition, machine learning and data mining, in multimodal biomedical databases. However, the proposed tools and methods for image collection analysis have some research challenges coming with deluge of big data in biomedicine such as: visual appearance variability, semantic gap between image content and high-level meaning, structural and interpretable representation of image content, semantic inclusion of multimodal information sources, and scalability support with the increasing volume of databases. In this way, the research proposal is addressing the problem of automatic extraction of knowledge from biomedical image collections. Specifically, the goal is to devise methods to automatically find: visual patterns that compactly explain the visual richness of biomedical images, relationships between visual patterns, and relationships between visual patterns and their meaning in a particular biomedical context. In order to solve it, the proposed methodology has three main stages: part-based bioimage representation, semantic bioimage representation and biomedical knowledge discovery. Each stage of methodology state-of-the-art methods from computer vision, image processing, machine learning and data mining will be explored to provide interpretable learning methods supported by high-performance computing.

Figure 1. Schematic representation of the proposed strategy.

International Conferences

  • Others

National Conferences

International Journals

      • Microscopy Research and Technique [Impact Factor: 1.850][avg time: ]

      • Journal of Biomedical Informatics [Impact Factor: 2.432][avg time: ]

      • International Journal of Medical Informatics [Impact Factor: 3.126][avg time: ]

      • Journal of Medical Systems. ISSN:0148-5598 [Impact Factor: 1.064][avg time: 2.75 months][Publindex: A2][url]

      • Journal in Computer Methods and Programs in Biomedicine [Impact Factor: 1.144][avg time: 6.86 months]

      • Computers in biology and medicine [Impact Factor: 1.269] [avg time: 11.24 months]

      • IEEE Transactions on Information Technology in Biomedicine [Impact Factor: 1.694] [avg time 11.04 months]

      • Computerized medical imaging and graphics [Impact factor. 1.041] [avg time: 9.74]

      • Journal in Artificial Intelligence in Medicine [impact factor: 1.645][5-year impact factor: 2.447][issues per year: 9][Publindex: A2][url]

      • International Journal of Medical Informatics [Impact Factor: 3.126][5-Year Impact Factor: 3.061][Issues per year: 12][Publindex: ][url]

      • Journal in Computer Methods and Programs in Biomedicine [Impact Factor: 1.144][5-Year Impact Factor: 1.362][Issues per year: 12][Publindex: ][url]

      • Journal of Pathology Informatics http://www.jpathinformatics.org/

National Journals

Research Groups

People