System and method for interactively and iteratively developing algorithms for detection of biological structures in biological samples

1. A method for categorizing biological structure of interest (BSOI) in digitized images of biological tissues comprising:

detecting, by a generic detector, one or more BSOIs in at least one pre-obtained digitized image of biological tissue from a set of images;

extracting image patches that contain, each, a BSOI;

annotating the image patches according to the detected BSOIs;

generating a BSOI categories classifier based on said annotation;

applying the categories classifier to at least some of the digitized images;

presenting an image that comprises at least one BSOI to a user, based on (i) a level of entropy of the BSOI, and (ii) a level of imbalance of a classified category of the image in the set of images;

receiving from the user input indicative of a category to be associated with the at least one BSOI; and

continuously annotating the image patches, to train the cell categories classifier according to the category of the BSOI provided by the user, as additional annotations become available,

wherein presenting the image comprises:

calculating a weighing component, configured to balance between a level of entropy of classified BSOIs and a level of imbalance of the classified category in the set of images;

ranking BSOIs based on the weighing component; and

displaying images containing BSOIs to the user ordered by the ranking.

2. The method of claim 1 , wherein each of the plurality of images comprises, at least one BSOI with high level of entropy.

3. The method according to claim 1 , wherein the weighing component is calculated by the function:

Weight= E*A+B *( N−E )* P minority

wherein

E=Entropy (class proportion)

A=Acquisition function as defined in Active Learning

B=parameter

N=number of categories

Pminority=output of a neural network that detects cell categories that give the probability for the minority category.

4. The method of claim 1 , wherein presenting an image further comprises:

automatically annotating BSOIs to be displayed to the user; and

allowing the user to intervene, and change the automatically produced annotation.

5. A method for categorizing biological structure of interest (BSOI) in digitized images of biological tissues comprising:

detecting, by a generic detector, one or more BSOIs in at least one pre-obtained digitized image of biological tissue from a set of images;

extracting image patches that contain, each, a BSOI;

annotating the image patches according to the detected BSOIs;

generating a BSOI categories classifier based on said annotation;

applying the categories classifier to at least some of the digitized images;

presenting an image that comprises at least one BSOI to a user, based on (i) a level of entropy of the BSOI, and (ii) a level of imbalance of a classified category of the image in the set of images;

receiving from the user input indicative of a category to be associated with the at least one BSOI; and

continuously annotating the image patches, to train the cell categories classifier according to the category of the BSOI provided by the user, as additional annotations become available,

wherein an order of presenting the images which comprise, each, at least one BSOI with high level of entropy, is responsive to the received user input indicative of a category of a B SOI, so that priority of presenting of images which await presenting to the user and comprise BSOI of the category that was indicated by the user, is made higher in response to the user's input.

6. A system for categorizing biological structure of interest (BSOI) in digitized images of biological tissues, said system comprising:

a non-transitory storage device comprising program code stored thereon; and

a processor configured to load the program code and execute the program code to perform:

detecting, by a generic detector, one or more BSOIs in at least one pre-obtained digitized image of biological tissue from a set of images;

extracting, image patches that contain, each, a BSOI;

annotating the image patches according to the detected BSOIs;

generating a BSOI categories classifier based on said annotation;

applying the categories classifier to at least some of the digitized images;

presenting an image that comprises at least one BSOI to a user, based on (i) a level of entropy of the BSOI, and (ii) a level of imbalance of a classified category of the image in the set of images;

receiving from the user input indicative of a category to be associated with the at least one BSOI; and

continuously annotating the image patches, to train the cell categories classifier according to the category of the BSOI provided by the user, as more annotations become available,

wherein the processor is further configured to present the image by:

calculating a weighing component, configured to balance between a level of entropy of classified BSOIs and a level of imbalance of the classified category in the set of images;

ranking BSOIs based on the weighing component and displaying images containing BSOIs to the user, ordered by the ranking.

7. The system according to claim 6 , wherein the weighing component is calculated by the function:

Weight= E*A−B *( N−E )* P minority

wherein

E=Entropy (class proportion)

A=Acquisition function as defined in Active Learning

B=parameter

N=number of categories

Pminority=output of a neural network that detects cell categories that give the probability for the minority category.