The localization phase begins by identifying each fluorophore within the data set. To be detected, a fluorophore must emit more photons than a certain threshold. The threshold is determined by measuring the local background and the noise level for each image pixel. A statistical test is used to determine if the photons emitted from a fluorophore are statistically significant when compared to the background distribution for that region of the image. The probability that a peak is outside the background distribution is determined, and a range of probability values is used to determine if a fluorophore should be accepted. The user-defined Local maximum factor is used to adjust the range of accepted probability values with larger values leading to more detected fluorophores.
Note: Accepting a larger range of probabilities will lead to more detected fluorophore events. However, it may also introduce false positives. Due to their low signal‐to‐noise ratio, these false positives will not be fit well by a 2‐D Gaussian model. |
After the fluorophores are detected, their sub-pixel positions are determined using a multiple-Gaussian-fitting routine. The localization precision for each fluorophore is determined by the number of photons emitted, with higher photon counts leading to more precise determinations of fluorophore position (as shown by smaller localization precision values). The x-and y-coordinates, photon measurements, localization precision, frame number, and time stamp are saved for each fluorophore as a *_LOC.txt output file.