Deep learning for digital pathology image analysis

Background: 

Computational pathology as well as the medical discipline pathology is a wide and diverse field which encompasses scientific research and day-to-day work in clinical settings. This novel field can be define as follows:
Computational pathology investigates a complete probabilistic treatment of scientific and clinical workflows in general pathology, i.e. it combines experimental design, statistical pattern recognition and survival analysis within a unified framework to answer scientific and clinical questions in pathology [1].
The analysis of large datasets plays increasingly important roles in many areas of science and engineering. Computational pathology builds on this base by demonstrating the use of methods, including machine learning as well as statistical and computational modeling, to understand complex disease states. To support the development of novel computer-aided diagnosis and predictive algorithms for selecting appropriate therapies, quantitative characterization of combined molecular, structural, and spatial 3D information from multiple biomedical inputs will be necessary [2].

References:
[1] Fuchs, Thomas J., and Joachim M. Buhmann. "Computational pathology: Challenges and promises for tissue analysis." Computerized Medical Imaging and Graphics 35.7 (2011): 515-530.
[2] Louis, David N., et al. "Computational pathology: a path ahead." Archives of pathology & laboratory medicine, 140.1 (2015): 41-50.

Description: 

Data analysis of novel multi-modal high-dimensional digital pathology signals is a challenging task. This data is redundant and needs a study by both the engineering and medical communities. In this project, the aim is to process multispectral images with different machine learning techniques for detecting cancer in tissues. For doing so, the following research questions should be investigated:

  • Developing an appropriate preprocessing stage for efficient processing of artificial neural network models
  • Designing a machine learning method for cancer detection in MSI data based on neural networks.

Candidates need to have a good basic knowledge of machine learning and in particular artificial neural networks. The programming language is preferably Python or C++.

Location: 
This project will be executed internally at the TU/e.
Keywords: 
Deep learning, Image analysis, signal analysis, machine learning, healthcare/oncology
Contact person: 
Partners: