Tracking operating room phase from captured video of the operating room

A system comprising:

• a plurality of image capture devices positioned within an operating room and each capturing video of the operating room, each image captured device having a different location within the operating room;
• and
• a surgical tracking server coupled to the plurality of image capture devices, the surgical tracking server including a non-transitory computer readable storage medium having instructions encoded thereon that, when executed by a processor, cause the processor to:
• identify objects within frames of the video obtained from one or more of the plurality of image capture devices by application of one or more computer vision models by the surgical tracking server;
• determine a state of each identified object by the surgical tracking server applying one or more models to characteristics of the video including the identified objects;
• determine a phase of the operating room from a set of predetermined phases by application of one or more phase classification models to the determined states for each identified object, wherein each phase classification model is trained to:
• access a set of rules, each rule corresponding to a phase of the set of predetermined phases and criteria identifying locations of objects within frames of video and a specific state of each object;
• and
• for a frame of the obtained video, determine the phase of the predetermined set corresponding to a rule for which the frame satisfies a maximum number of criteria;
• and
• store the determined phase and a time when the determined phase was determined in association with the operating room.