Purpose: The problem of reliable automatic estimation of density of corneal endothelial cells in images from optical microscopy was addressed. The reliability of estimated densities should be comparable to that of manual cell count. Methods: The spatial frequencies contained in endothelium images are extracted with suitable mathematical techniques (2-dimension Discrete Fourier Transform, DFT). A circular band in the DFT of the images is shown to contain the frequency information related to the cell density. An algorithm for reliably identifying this spatial frequency information and for extracting from it an estimate of the cell density has been developed. A prototype of the algorithm was implemented in the Matlab® language and run on a personal computer. A preliminary evaluation was performed on a data set containing 50 corneas, with two 100X images each; manual count was performed by an expert ophthalmologist on two 200X images for each cornea. Results: Mean percent difference of automatic densities vs. manual ones was 2.6% (std dev 5.5%, max 15.7%), with a Pearson correlation coefficient of 0.90. Mean percent absolute difference was 4.7% (std dev 3.8%). Running times of the prototype were in the order of 40 seconds per image. Conclusions: A new algorithm was developed for the reliable automatic estimation of endothelial cell density. A preliminary clinical evaluation of the proposed technique confirmed its capability of reliably estimating corneal endothelium cell density. Implementation of the algorithm with a more efficient computer language, e.g. C++, will allow execution times in the order of 5-10 seconds.
A software for the reliable automatic estimation of endothelial cell density
RUGGERI, ALFREDO;GRISAN, ENRICO;
2004
Abstract
Purpose: The problem of reliable automatic estimation of density of corneal endothelial cells in images from optical microscopy was addressed. The reliability of estimated densities should be comparable to that of manual cell count. Methods: The spatial frequencies contained in endothelium images are extracted with suitable mathematical techniques (2-dimension Discrete Fourier Transform, DFT). A circular band in the DFT of the images is shown to contain the frequency information related to the cell density. An algorithm for reliably identifying this spatial frequency information and for extracting from it an estimate of the cell density has been developed. A prototype of the algorithm was implemented in the Matlab® language and run on a personal computer. A preliminary evaluation was performed on a data set containing 50 corneas, with two 100X images each; manual count was performed by an expert ophthalmologist on two 200X images for each cornea. Results: Mean percent difference of automatic densities vs. manual ones was 2.6% (std dev 5.5%, max 15.7%), with a Pearson correlation coefficient of 0.90. Mean percent absolute difference was 4.7% (std dev 3.8%). Running times of the prototype were in the order of 40 seconds per image. Conclusions: A new algorithm was developed for the reliable automatic estimation of endothelial cell density. A preliminary clinical evaluation of the proposed technique confirmed its capability of reliably estimating corneal endothelium cell density. Implementation of the algorithm with a more efficient computer language, e.g. C++, will allow execution times in the order of 5-10 seconds.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.