Image segmentation for sets of objects

A computing system comprising:

• memory storing a 3-dimensional (3D) image of a set of bones in a lower extremity of a patient;
• and
• processing circuitry configured to:
• partition the 3D image of the set of bones of a patient into a 3D image of an ankle region of the lower extremity of the patient, a 3D image of a forefoot region of the lower extremity of the patient, and a 3D image of a transition region of the lower extremity of the patient;
• perform a first segmentation process that generates a first segmentation mask based on the 3D image of the ankle region, wherein to perform the first segmentation process, the processing circuitry is configured to apply a first neural network trained to recognize bones in the ankle region, and applying the first neural network comprises loading a first set of weights into a memory at a first time;
• perform a second segmentation process that generates a second segmentation mask based on the 3D image of the forefoot region, wherein to perform the second segmentation process, the processing circuitry is configured to apply a second neural network trained to recognize bones in the forefoot region, and applying the second neural network comprises loading a second set of weights into the memory at a second time different from the first time;
• perform a third segmentation process that generates a third segmentation mask based on the 3D image of the transition region, wherein to perform the third segmentation process, the processing circuitry is configured to apply a third neural network trained to recognize bones in the transition region, and applying the third neural network comprises loading a third set of weights into the memory at a third time different from the first time and the second time;
• composite the first segmentation mask, the second segmentation mask, and the third segmentation mask to generate a fourth segmentation mask;
• and
• output the fourth segmentation mask for display.

A non-transitory computer-readable storage medium having instructions stored thereon that, when executed by processing circuitry, cause the processing circuitry to:

• partition a 3-dimensional (3D) image of a set of bones in a lower extremity of a patient into a 3D image of an ankle region of the lower extremity of the patient, a 3D image of a forefoot region of the lower extremity of the patient, and a 3D image of a transition region of the lower extremity of the patient;
• perform a first segmentation process that generates a first segmentation mask based on the 3D image of the ankle region, wherein performing the first segmentation process comprises applying a first neural network trained to recognize bones in the ankle region, and applying the first neural network comprises loading a first set of weights into a memory at a first time;
• perform a second segmentation process that generates a second segmentation mask based on the 3D image of the forefoot region, wherein performing the second segmentation process comprises applying a second neural network trained to recognize bones in the forefoot region, and applying the second neural network comprises loading a second set of weights into the memory at a second time different from the first time;
• perform a third segmentation process that generates a third segmentation mask based on the 3D image of the transition region, wherein performing the third segmentation process comprises applying a third neural network trained to recognize bones in the transition region, and applying the third neural network comprises loading a third set of weights into the memory at a third time different from the first time and the second time;
• composite the first segmentation mask, the second segmentation mask, and the third segmentation mask to generate a fourth segmentation mask;
• and
• output the fourth segmentation mask for display.