System and method for detecting landmarks in a three-dimensional image volume

An apparatus for detecting target landmarks in an image volume, comprising:

• a processor;
• and
• a memory to store computer program instructions, the computer program instructions when executed on the processor cause the processor to perform operations comprising:
• detecting a set of landmark candidates for each of a plurality of target landmarks in an image volume;
• generating a node potential value for each landmark candidate based at least in part on an error value determined using spatial histogram-based error regression;
• generating edge potential values for pairs of landmark candidates based on a bifurcation analysis of the image volume using vessel tracing;
• and
• determining an optimal landmark candidate for each target landmark, using a Markov random field model based on the node potential values and the edge potential values.

A non-transitory computer readable medium encoded with computer executable instructions for detecting target landmarks in an image volume, the computer executable instructions defining steps comprising:

• detecting a set of landmark candidates for each of a plurality of target landmarks in an image volume;
• generating a node potential value for each landmark candidate based at least in part on an error value determined using spatial histogram-based error regression;
• generating edge potential values for pairs of landmark candidates based on a bifurcation analysis of the image volume using vessel tracing;
• and
• determining an optimal landmark candidate for each target landmark, using a Markov random field model based on the node potential values and the edge potential values.