Machine-learning-enabled imputation of spatial omics data based on histopathology image data

A method for generating synthetic spatial omics images, comprising:

• receiving a histopathology image depicting a diseased region of interest of an input tissue sample;
• and
• generating a synthetic spatial omics image depicting one or more stained structures of interest within the diseased region of interest by inputting the histopathology image into a trained machine learning model, wherein the machine learning model is trained by:
• receiving a training histopathology image of a training tissue sample, wherein the training histopathology image comprises a plurality of identified landmarks;
• receiving a training spatial omics image of the training tissue sample, wherein the training spatial omics image comprises a plurality of identified landmarks;
• registering the training histopathology image and the training spatial omics image based on the plurality of landmarks identified in the training histopathology image and the plurality of landmarks identified in the training spatial omics image;
• generating, based on the registration of the training histopathology image and the training spatial omics image, a training dataset comprising a histopathology image set and a corresponding spatial omics image set;
• and
• training the machine learning model based on the generated training dataset.

A non-transitory computer-readable storage medium storing one or more programs for generating synthetic spatial omics images, the one or more programs comprising instructions, which when executed by one or more processors of an electronic device, cause the electronic device to perform:

• receiving a histopathology image depicting a diseased region of interest of an input tissue sample;
• and
• generating a synthetic spatial omics image depicting one or more stained structures of interest within the diseased region of interest by inputting the histopathology image into a trained machine learning model, wherein the machine learning model is trained by:
• receiving a training histopathology image of a training tissue sample, wherein the training histopathology image comprises a plurality of identified landmarks;
• receiving a training spatial omics image of the training tissue sample, wherein the training spatial omics image comprises a plurality of identified landmarks;
• registering the training histopathology image and the training spatial omics image based on the plurality of landmarks identified in the training histopathology image and the plurality of landmarks identified in the training spatial omics image;
• generating, based on the registration of the training histopathology image and the training spatial omics image, a training dataset comprising a histopathology image set and a corresponding spatial omics image set;
• and
• training the machine learning model based on the generated training dataset.