2019-01486 - PhD Position F/M [CORDIS2019-Morpheme]: Statistical and Geometrical Analysis of Filamentous Networks in Biology from Microscopy Images.

Contract type : Public service fixed-term contract

Level of qualifications required : Graduate degree or equivalent

Fonction : PhD Position

About the research centre or Inria department

The Inria Sophia Antipolis - Méditerranée center counts 37 research teams and 9 support departments. The center's staff (about 600 people including 400 Inria employees) is composed of scientists of different nationalities (250 foreigners of 50 nationalities), engineers, technicians and administrators. 1/3 of the staff are civil servants, the others are contractual. The majority of the research teams at the center are located in Sophia Antipolis and Nice in the Alpes-Maritimes. Six teams are based in Montpellier and a team is hosted by the computer science department of the University of Bologna in Italy. The Center is a member of the University and Institution Community (ComUE) "Université Côte d'Azur (UCA)".

Context

 

Assignment

Filamentous (Or Fiber) networks play an essential role in biology. We can exemplify the mitochondrial network within cells, neuronal networks, actin filaments or even the extra-cellular matrix. Studying the geometry and the topology of these networks is essential to exhibit biomarkers of some given pathologies. To extract such information, confocal microscopy images are adapted due to the resolution of the data and to the numerous technics of fluorescence. Currently, such studies are conducted for example to evaluate the toxicity of some endocrine disrupters from the mitochondrial network, to evaluate some cancers from the extra cellular matrix or to understand some brain pathologies from the neuron growth process. In these projects, the topological and geometrical properties of the networks are extracted from the images. Some data analysis tools, such as classification or PCA allow populations characterization or even phenotype explanation or prediction. To achieve this goal it is necessary to develop automatic algorithms to compare and classify networks. This implies to define relevant metrics that can be biologically interpretable as well as a rigorous statistical framework to evaluate mean shapes, variability between and within populations.

Some software have been developed to extract linear structures from images, for example to detect road or vascular networks. We can distinguish adapted filters such as Gabor functions, local operators based on Hessian matrix analysis or more global stochastic approaches. However, a versatile algorithm is still an open issue due to the huge variability of situations of the objects of interest(mitochondria, neurons, actin filaments) but also of the microscopes. Concerning the metrics between graphs the proposed approaches are often based on the topological properties, as for instance in the "Tree Edit Distance " (TED). In our case, the geometrical information is also very important (size and shape of branches). In a previous work, we have combined these two types of information to classify axonal trees [Mottini]. This new approach still has to be generalized to any graph.

 

The main goal of this PHD project is to build a methodological framework for analyzing filamentous networks in biology. In the first part, we will derive a versatile and ergonomic software for detecting linear network in microscopy images. This approach could be built in three main steps : linear structures enhancement, binarization and reconnection [Risser] . This approach could be compared to a more global one based on marked point processes [Lacoste]. The second part will consist in developing a rigorous statistical framework to analyze biological graphs. We will define a metric in a well chosen space by generalizing the works described in [Mottini] that combine a geometrical metric, proposed in [Srivastava], with the topological information. In this setup, some algorithms will be proposed to compute geodesic path on the considered manifold, mean graphs or covariace matrices computed in the tangent space. This will provide a framework to derive classification and modeling algorithms for biological graphs obtained from real data.

This theoretical framework will be applied to different concrete problems in collaboration with biological institutes, such as the C3M and IPMC for studying mitochondrial networks respectively in the case of prostate cancer and Alzheimer disease. Neurons and extra cellular matrix applications will be addressed in collaboration with iBV.

 

Bibliography :

[Gillette] : T. Gillette, K. Brown, G. Ascoli (2011). The diadem metric: Comparing multiple reconstructions of the same neuron. Neuroinformatics, 9, 233–245.

[Mottini] : A. Mottini, X. Descombes, F. Besse (2015). From Curves to Trees: A Tree-like Shapes Distance Using the Elastic Shape Analysis Framework. Neuroinformatics, Humana Press, 13 (2), pp.175-191.

[Risser] : L. Risser and F. Plouraboue and X. Descombes (2008). Gap Filling of 3-D Microvascular Networs by Tensor Voting. IEEE Trans. Medical Imaging, 27(5): pages 674-687.

[Lacoste] : C. Lacoste and X. Descombes and J. Zerubia (2005). Point Processes for Unsupervised Line Network Extraction in Remote Sensing. IEEE Trans. Pattern Analysis and Machine Intelligence, 27(10): pages 1568-1579.

[Srivastava] : A. Srivastava, S. Joshi, W. Mio, and C. Liu (2005). Statistical Shape Anlaysis: Clustering, Learning and Testing. IEEE Transactions on Pattern Analysis and Machine Intelligence 27(4), pages 590-602,

 

 

Main activities

Main activities (5 maximum) :

Research

 

Skills

Technical skills and level required :

master in applied mathematic, computer science or a related field

Languages :

Englsih

Relational skills :

work in a multidisciplinary context

 

Benefits package

  • Subsidized meals
  • Partial reimbursement of public transport costs
  • Leave: 7 weeks of annual leave + 10 extra days off due to RTT (statutory reduction in working hours) + possibility of exceptional leave (sick children, moving home, etc.)
  • Possibility of teleworking (after 6 months of employment) and flexible organization of working hours
  • Professional equipment available (videoconferencing, loan of computer equipment, etc.)
  • Social, cultural and sports events and activities
  • Access to vocational training
  • Social security coverage

Remuneration

Duration: 36 months
Location: Sophia Antipolis, France
Gross Salary per month: 1982€brut per month (year 1 & 2) and 2085€ brut/month (year 3)