A generalized Potts model for confocal microscopy images
Much as being among the least invasive mainstream imaging technologies in life sciences, the resolution of confocal microscopy is limited. Imaged structures, e.g., chromatin-fiber loops, have diameters around or beyond the diffraction limit, and microscopy images show seemingly random spatial densit...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article (Journal) |
| Language: | English |
| Published: |
26 January 2015
|
| In: |
International journal of modern physics. B, Condensed matter physics etc.
Year: 2015, Volume: 29, Issue: 08, Pages: 1550048 |
| ISSN: | 1793-6578 |
| DOI: | 10.1142/S0217979215500484 |
| Online Access: | Verlag, Volltext: http://dx.doi.org/10.1142/S0217979215500484 Verlag, Volltext: http://www.worldscientific.com/doi/abs/10.1142/S0217979215500484 
|
| Author Notes: | Gabriell Máté, Dieter W. Heermann |
| Summary: | Much as being among the least invasive mainstream imaging technologies in life sciences, the resolution of confocal microscopy is limited. Imaged structures, e.g., chromatin-fiber loops, have diameters around or beyond the diffraction limit, and microscopy images show seemingly random spatial density distributions only. While such images are important because the organization of the chromosomes influences different cell mechanisms, many interesting questions can also be related to the observed patterns. These concern their spatial aspects, the role of randomness, the possibility of modeling these images with a random generative process, the interaction between the densities of adjacent loci, the length-scales of these influences, etc. We answer these questions by implementing a generalization of the Potts model. We show how to estimate the model parameters, test the performance of the estimation process and numerically prove that the obtained values converge to the ground truth. Finally, we generate images with a trained model and show that they compare well to real cell images. |
|---|---|
| Item Description: | Gesehen am 16.08.2017 |
| Physical Description: | Online Resource |
| ISSN: | 1793-6578 |
| DOI: | 10.1142/S0217979215500484 |