Co-localization of multiple internal organs based on images obtained during surgery

A method, comprising:

• in a surgical assistive device that comprises a surgical image capture device and co-localization circuitry:
• capturing, by the surgical image capture device, capture a sequence of video frames, based on insertion of the surgical image capture device in a body of a subject;
• generating, by the co-localization circuitry, first localization likelihood values for an internal organ of interest and second localization likelihood values for a neighboring organ having an anatomical adjacency with the internal organ of interest in a test video frame selected from the captured sequence of video frames;
• determining, by the co-localization circuitry, a extent of an anatomical adjacency boundary between the internal organ of interest and the neighboring organ in the test video frame, based on the generated first localization likelihood values and the generated second localization likelihood values;
• segmenting, by the co-localization circuitry, the test video frame into a plurality of regions;
• generating, by the co-localization circuitry, a first region for the internal organ of interest by integration of a first set of regions from the plurality of regions in the segmented test video frame, based on a conditional probabilistic distribution for a localization parameter for the internal organ of interest in the test video frame with respect to the anatomical adjacency boundary and a localization parameter for the neighboring organ;
• and
• localizing, by the co-localization circuitry, the internal organ of interest within the body of the subject based on the generated first region in the test video frame, in accordance with extent of the anatomical adjacency boundary.