System and method for predicting endometrium receptivity

1 . A computer-implemented system for predicting endometrium receptivity, comprising:

a processor; and

a memory coupled to the processor and storing processor-executable instructions that, when executed, configure the processor to:

maintain a data set representing a neural network having a plurality of weights;

obtain a first image of an endometrium with a first timestamp;

extract a first set of target endometrium physical features from the first image, including a thickness of the endometrium, wherein extracting the thickness of the endometrium comprises identifying morphologically-relevant points of the endometrium in the first image and determining the thickness of the endometrium based on at least some of the morphologically-relevant points; and

generate, using the neural network and based on the first set of target endometrium physical features including the thickness of the endometrium, a predicted value indicating an endometrium receptivity for embryo implantation in the endometrium in the first image, the predicted value for determining a window of the embryo implantation.

2 . The system of claim 1 , wherein the processor-executable instructions, when executed, further configure the processor to:

generate a value representative of a likelihood of a successful embryo implantation.

3 . The system of claim 1 , wherein the first set of target endometrium physical features further comprises at least one of: a length of the endometrium, a surface area of the endometrium, and a pattern of the endometrium.

4 . The system of claim 3 , wherein the pattern of the endometrium comprises a trilaminar pattern.

5 . The system of claim 1 , wherein the processor-executable instructions, when executed, configure the processor to:

receive a second image of the endometrium with a second timestamp;

extract a second set of target endometrium physical features from the second image; and

generate, using the neural network and based on the first and second sets of target endometrium physical features, the predicted value indicating the endometrium receptivity for embryo implantation in the endometrium.

6 . The system of claim 5 , wherein the second set of target endometrium physical features comprises at least one of: a thickness of the endometrium, a length of the endometrium, a surface area of the endometrium, and a trilaminar pattern of the endometrium.

7 . The system of claim 5 , wherein the processor-executable instructions, when executed, configure the processor to:

determine a difference between the first image and the second image; and

analyze the difference to generate the predicted value indicating the endometrium receptivity for embryo implantation in the endometrium.

8 . The system of claim 1 , wherein the predicted value indicating the endometrium receptivity comprises a probability value.

9 . The system of claim 1 , wherein the neural network is trained based on a set of training data comprising:

a plurality of ultrasound images of one or more endometria, each of the plurality of ultrasound images showing a respective endometrium; and

for each of the plurality of ultrasound images, a respective label indicating an outcome of a respective embryo implantation in the respective endometrium in the respective ultrasound image.

10 . The system of claim 9 , wherein each of the plurality of ultrasound images is associated with training data comprising a blastocyst quality of an embryo transferred into a respective endometrial cavity in the respective ultrasound image.

11 . A computer-implemented method for predicting endometrium receptivity, the method comprising:

maintaining a data set representing a neural network having a plurality of weights;

obtaining a first image of an endometrium with a first timestamp;

extracting a first set of target endometrium physical features from the first image including a thickness of the endometrium, wherein extracting the thickness of the endometrium comprises identifying morphologically-relevant points of the endometrium in the first image and determining the thickness of the endometrium based on at least some of the morphologically-relevant points; and

generating, using the neural network and based on the first set of target endometrium physical features including the thickness of the endometrium, a predicted value indicating an endometrium receptivity for embryo implantation in the endometrium in the first image, the predicted value for determining a window of the embryo implantation.

12 . The method of claim 11 , further comprising:

generating a value representative of a likelihood of a successful embryo implantation.

13 . The method of claim 11 , wherein the first set of target endometrium physical features further comprises at least one of: a length of the endometrium, a surface area of the endometrium, and a pattern of the endometrium.

14 . The method of claim 13 , wherein the pattern of the endometrium comprises a trilaminar pattern.

15 . The method of claim 11 , further comprising:

receiving a second image of the endometrium with a second timestamp;

extracting a second set of target endometrium physical features from the second image; and

generating, using the neural network and based on the first and second sets of target endometrium physical features, the predicted value indicating the endometrium receptivity of the endometrium.

16 . The method of claim 15 , wherein the second set of target endometrium features comprises at least one of: a thickness of the endometrium, a length of the endometrium, a surface area of the endometrium, and a trilaminar pattern of the endometrium.

17 . The method of claim 15 , further comprising:

determining a difference between the first image and the second image; and

analyzing the difference to generate the predicted value indicating the endometrium receptivity of the endometrium.

18 . The method of claim 11 , wherein the neural network is trained based on a set of training data comprising:

a plurality of ultrasound images of one or more endometria, each of the plurality of ultrasound images showing a respective endometrium; and

for each of the plurality of ultrasound images, a respective label indicating an outcome of a respective embryo implantation in the respective endometrium in the respective ultrasound image.

19 . The method of claim 18 , wherein each of the plurality of ultrasound images is associated with training data comprising a blastocyst quality of an embryo transferred into a respective endometrial cavity in the respective ultrasound image.

20 . A non-transitory computer-readable medium having stored thereon machine interpretable instructions which, when executed by a processor, cause the processor to perform:

maintaining a data set representing a neural network having a plurality of weights;

obtaining a first image of an endometrium with a first timestamp;

extracting a first set of target endometrium physical features from the first image including a thickness of the endometrium, wherein extracting the thickness of the endometrium comprises identifying morphologically-relevant points of the endometrium in the first image and determining the thickness of the endometrium based on at least some of the morphologically-relevant points; and

generating, using the neural network and based on the first set of target endometrium physical features including the thickness of the endometrium, a predicted value indicating a endometrium receptivity for embryo implantation in the endometrium in the first image, the predicted value for determining a window for the embryo implantation.