Method for examining distributed objects by segmenting an overview image

A method for examining a plurality of distributed objects using an overview image of an area in which the distributed objects are distributed, comprising:

• converting the overview image into a binary image by classifying intensity values of pixels of the overview image as to whether they are below or above a predetermined threshold;
• cleaning the binary image of structures which are smaller than the distributed objects to generate a cleaned-up image;
• morphologically closing the cleaned-up image to form a binary object mask indicating which locations of the area belong to the distributed objects and which locations of the area belong to no object;
• irradiating at least one location in the area at which, according to the binary object mask, at least one of the distributed objects is located, the at least one location being irradiated with a light beam;
• evaluating one or more responses of the at least one of the distributed objects to the irradiation;
• and
• scanning the light beam relative to the at least one of the distributed objects based on the binary object mask, wherein a timing program of the scanning taking into account a spatial intensity distribution of the light beam is selected such that the spatial intensity distribution of a total received light dose is homogenized over the at least one of the distributed objects.

A method for examining a plurality of distributed objects using an overview image of an area in which the distributed objects are distributed, comprising:

• converting the overview image into a binary image by classifying intensity values of pixels of the overview image as to whether they are below or above a predetermined threshold;
• cleaning the binary image of structures which are smaller than the distributed objects to generate a cleaned-up image;
• morphologically closing the cleaned-up image to form a binary object mask indicating which locations of the area belong to the distributed objects and which locations of the area belong to no object;
• irradiating at least one location in the area at which, according to the binary object mask, at least one of the distributed objects is located, the at least one location being irradiated with a light beam;
• and
• evaluating one or more responses of the at least one of the distributed objects to the irradiation,
• wherein biological cells, which are considered as the distributed objects and provided on a plurality of substrates, are irradiated with the light beam, and wherein the responses of the individual biological cells to the irradiation are merged across the substrates,
• wherein, based a totality of the one or more responses, a multivariate analysis is used to evaluate which properties of the biological cells change from one substrate to another substrate in a statistically significant manner, and
• wherein at least one of (i) the multivariate analysis includes the Principal Component Analysis (PCA), (ii) a classifier is evaluated from the multivariate analysis, the classifier mapping responding to the irradiation obtained on the biological cells on a further substrate to at least one value of a property applicable to this substrate, or (iii) spectra of pure components are fitted to measured spectra using least squares fitting.

A non-transitory computer-accessible medium which includes a machine readable program with instructions, wherein, when the program is executed on a computer, the computer is configured to execute procedures comprising:

• converting the overview image into a binary image by classifying intensity values of pixels of the overview image as to whether they are below or above a predetermined threshold;
• cleaning the binary image of structures which are smaller than the distributed objects to generate a cleaned-up image;
• morphologically closing the cleaned-up image to form a binary object mask indicating which locations of the area belong to the distributed objects and which locations of the area belong to no object;
• causing an irradiation of at least one location in the area at which, according to the binary object mask, at least one of the distributed objects is located, the at least one location being irradiated with a light beam;
• and
• evaluating one or more responses of the at least one of the distributed objects to the irradiation,
• wherein biological cells, which are considered as the distributed objects and provided on a plurality of substrates, are irradiated with the light beam, and wherein the responses of the individual biological cells to the irradiation are merged across the substrates,
• wherein, based a totality of the one or more responses, a multivariate analysis is used to evaluate which properties of the biological cells change from one substrate to another substrate in a statistically significant manner, and
• wherein at least one of (i) the multivariate analysis includes the Principal Component Analysis (PCA), (ii) a classifier is evaluated from the multivariate analysis, the classifier mapping responding to the irradiation obtained on the biological cells on a further substrate to at least one value of a property applicable to this substrate, or (iii) spectra of pure components are fitted to measured spectra using least squares fitting.