Dr. Nicholas Devaney

B.Sc, PhD

Contact Details

Lecturer Above The Bar
School of Physics
NUIG

T: Ext. 5188
E: nicholas.devaney@nuigalway.ie
 
researcher

Biography



Ph.D. (NUIG then UCG)1989
Royal Greenwich Observatory:Research Associate1989-1991
Observatoire de Meudon, Paris: Research Associate1991-1992
Instituto de Astrofisica de Canarias: Support Astronomer1993-1995
GRANTECAN, Instituto de Astrofisica de Canarias: Adaptive Optics Manager1995-2005
National University of Ireland, Galway: Lecturer2005-



My background is in Astronomical Imaging with high spatial resolution, mostly involving Adaptive Optics and Image Processing. My current work in this area concerns the detection of extrasolar planets in data obtained using adaptive optics. I am also interested in the application of adaptive optics and other techniques to the enhancement of microscopy. In addition, I am supervising projects on Image processing of medical images; specifically on the study of retinal disease using adaptive-optics retinal images, the optimal detection of lesions in mammography and computation imaging. The latter is concerned with the combination of optics and image processing to provide new imaging modalities such as 3D imaging, depth estimation, face and gesture detection etc.

Research Interests

Computational Imaging

Computational imaging involves the combination of optics and image processing to enhance some aspect of imaging or allow the extraction of new information. For example, aperture masks may be designed which extend the depth of focus of a camera, or allow depth to be extracted from images. Plenoptic cameras and camera arrays provide information on the 'light field' i.e. the direction of the rays from the objects in the scene. This information allows digital re-focusing, the generation of synthetic views and recovery of 3D object shape. The gain in angular information is traded against spatial resolution, and we are investigating techniques to provide higher spatial resolution. This will be particularly important for applications in microscopy as well as macroscopic applications, such as face and gesture recognition.
 
High-resolution retinal image processing

Adaptive optics is now used to provide images of the retina with unprecedented detail. Image processing can further enhance these images by selecting those images with the highest contrast, and precisely registering the images. We use a multi-resolution approach based on wavelets to carry out these tasks. We are also developing ways to automatically detect photoreceptors (cones and rods) and nerve fivers in the retinal images. Quantification of these structures at high resolution will allow early detection of retinal disease, and tracking the effects of treatment. We are carrying out this work in close collaboration with ophthalmologists.
 
Enhancing microscopy using Adaptive Optics and deconvolution

Images obtained with microscopes may be blurred by aberrations introduced by the sample. Adaptive optics has been used recently to correct for these aberrations and provide clearer images. Alternatively, the aberrations can be measured and the correction applied in post-processing. We are comparing these approaches in order to understand the relative benefits as applied to different modes of microscopy.

Thermal Imaging
 
We are investigating the combination of thermal and visual imaging for novel applications. The thermal imager has much lower spatial resolution than the visual camera, and we are developing techniques to 'fuse' the data from the two systems in order to enhance enhance the function. An example of application is the measurement of core body temperature from the nasal corner of the eye.
Exoplanet Detection

A large number of planets have been detected outside our solar system using indirect techniques (e.g. see here). Imaging of these exoplanets is extremely challenging given how close they are to their parent stars, and how much fainter. The brightness ratio with the parent star is of the order 10^-6 to 10^-9. Obtaining such images will require using large (or extremely large) telescopes, with practically perfect adaptive optics. Even so, residual speckle effects severely hamper our ability to unambiguously detect exoplanets. We have developed optimal techniques based on using knowledge of the data statistics. We have also developed a powerful inverse-problem approach applicable to multi-wavelength data. This exploits the fact that the position of residual speckles scales with wavelength, and this fact may be used to discriminate speckles from planets. We want to extend these approaches to future Extremely Large Telescopes.

Lesion detection in mammography

The sort of optimal detection techniqes which we have developed for exoplanet images can be applied to many tasks; one example is the detection of lesions in mammographs. This is complicated by the fact that there are many different kinds of lesion, and the background tissue varies greatly between subjects.

Research Projects

  Project Start Date End Date
EI CS 2008 0124 ITN PROPOSAL N DEVANEY 01-AUG-08 31-MAR-09

Peer Reviewed Journals

  Year Publication
(2015) 'Performance analysis of cone detection algorithms'
L. Mariotti and N. Devaney (2015) 'Performance analysis of cone detection algorithms'. Journal Of The Optical Society Of America A-Optics Image Sci, 32 [Details]
(2014) 'Registration of adaptive optics corrected retinal nerve fiber layer (RNFL) images'
Ramaswamy, G,Lombardo, M,Devaney, N (2014) 'Registration of adaptive optics corrected retinal nerve fiber layer (RNFL) images'. Biomedical Optics Express, 5 :1941-1951 [DOI] [Details]
(2013) 'Pre-processing, registration and selection of adaptive optics corrected retinal images'
Ramaswamy, G,Devaney, N (2013) 'Pre-processing, registration and selection of adaptive optics corrected retinal images'. Ophthalmic And Physiological Optics, 33 :527-539 [DOI] [Details]
(2013) 'Adaptive optics technology for high-resolution retinal imaging'
M. Lombardo, S. Serrao, N.Devaney, M. Parravano, G. Lombardo (2013) 'Adaptive optics technology for high-resolution retinal imaging'. Sensors, [Details]
(2010) 'Enhanced faint companion photometry and astrometry using wavelength diversity'
Burke, D,Devaney, N (2010) 'Enhanced faint companion photometry and astrometry u