Premium
A new automatic method for microglial‐cell quantification in whole‐mount mouse retinas
Author(s) -
DE HOZ R,
GALLEGO BI,
ROJAS B,
RAMIREZ AI,
SALAZAR JJ,
TRIVIÑO A,
DE GRACIA PACHECO P,
RAMIREZ JM
Publication year - 2013
Publication title -
acta ophthalmologica
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.534
H-Index - 87
eISSN - 1755-3768
pISSN - 1755-375X
DOI - 10.1111/j.1755-3768.2013.t010.x
Subject(s) - retinal , retina , outer plexiform layer , automated method , inner plexiform layer , segmentation , matlab , computer science , artificial intelligence , ophthalmology , pathology , biomedical engineering , biology , medicine , neuroscience , operating system
Abstract Purpose To study the proliferative microglial behavior in a laser‐induced ocular hypertension (OHT) model with an automatic method that allows the assessment of the number of microglial cells and comparisons of the results with those found by direct human observation. Methods Albino Swiss mice were divided into: age‐matched control, n=6; and lasered, n= 6. Retinal whole‐mounts were immunostained with anti‐Iba1. In Matlab, new algorithms of segmentation and control of distances were developed to determine the number of Iba‐1+ cells. The automatic results were compared with those from direct human observation of the images. Results The algorithm method automatically detected the number of Iba‐1+ retinal cells in the inner and outer plexiform layers, both in naïve and OHT retinas. The number of cells present in the image samples was not an obstacle for the program to run properly. The time required for counting Iba 1+ cells decreased from the human guide to the program‐based counting method from days to one hour. the results showed a strong correlation between automatic and manual methods (Pearson correlation test, R = 0.979; P=0.000 and R= 0.942; P=0.000 for outer and inner plexiform layer, respectively) indicating the reliability of the automatic counting. Conclusion A new consistent and fast algorithm method was developed with Matlab to quantify Iba‐1+ microglial cells as well as cellular density maps through retinal whole‐mounts, in both naïve and OHT. Through this new automatic method, a larger set of images or samples could be included in future studies to analyze the behavior of microglial cells under proliferative conditions.