DocumentCode :
3540378
Title :
The value of multispectral observations in photon-limited quantitative tissue analysis
Author :
Harmany, Zachary T. ; Jiang, Xin ; Willett, Rebecca
Author_Institution :
Dept. of Electr. & Comput. Eng., Duke Univ., Durham, NC, USA
fYear :
2012
fDate :
5-8 Aug. 2012
Firstpage :
237
Lastpage :
240
Abstract :
Multispectral fluorescence data can be used within sparse decomposition methods to separate key cellular structures, even when the number of photons per spectral band is very small. However, there are two key costs associated with multispectral data acquisition: (a) for a fixed data acquisition time, increasing the number of spectral bands means decreasing the number of photons (and hence SNR) per band, and (b) the optical system becomes more complex and expensive. These costs lead to important tradeoffs between the information content and the noise of the observations. This paper describes a mathematical framework for assessing this tradeoff and supporting experimental results.
Keywords :
cellular biophysics; data acquisition; endoscopes; fluorescence; medical image processing; tissue engineering; tumours; cellular structures; data acquisition time; information content; multispectral data acquisition; multispectral fluorescence data; photon limited imaging; photon limited quantitative tissue analysis; sparse decomposition methods; Coherence; Dictionaries; Image reconstruction; Muscles; Noise; Photonics; Tumors; Photon-limited imaging; Poisson noise; convex optimization; fluorescence imaging; microscopy; sparse decomposition; spectral imaging;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Statistical Signal Processing Workshop (SSP), 2012 IEEE
Conference_Location :
Ann Arbor, MI
ISSN :
pending
Print_ISBN :
978-1-4673-0182-4
Electronic_ISBN :
pending
Type :
conf
DOI :
10.1109/SSP.2012.6319670
Filename :
6319670
Link To Document :
بازگشت