Technologies for intra-operative ligament balancing using machine learning

A system for generating intra-operative input data for a trained machine learning model (MLM) used in a ligament balancing procedure, the system comprises:

• a plurality of heterogenous sensors configured to generate distinct types of data, the plurality of heterogenous sensors comprising one or more load sensors and one or more optical sensors;
• at least one processor coupled to the plurality of heterogenous sensors;
• at least memory coupled to the processor, the memory comprising the trained MLM and a plurality of instructions stored thereon that, in response to execution by the at least one processor, cause the system to:
• receive, by the at least one processor, real-time force data from the one or more load sensors, the real-time force data comprising measurements of a plurality of forces applied to the one or more load sensors at a plurality of points between flexion and extension of the joint, the one or more load sensors being attached to one or more bone sections of a joint;
• receive, by the at least one processor, real-time displacement data from the one or more optical sensors, the real-time displacement data comprising measurements of a displacement between the one or more bone sections of the joint;
• compute, by the at least one processor, one or more normalized relationships between real-time force data from the one or more load sensors and real-time displacement data from the one or more optical sensors, across a range of motions between flexion and extension of the joint, to provide a derived biomechanical representation of the joint state;
• and
• provide, by the at least one processor, the derived biomechanical representation of joint performance as one or more input parameters into the trained MLM.

A method for generating intra-operative input data for a trained machine learning model (MLM) used in a ligament balancing procedure, the method comprising:

• receiving, by at least one processor and from one or more load sensors, real-time measurements of a plurality of forces applied to the one or more load sensors at a plurality of points between flexion and extension of the joint, the one or more load sensors being attached to one or more bone sections of a joint;
• receiving, by the at least one processor and from one or more optical sensors, real-time measurements of a plurality of positions associated with one or more contact points of the joint, measured across a range of motions between flexion and extension of the joint, the real-time measurements of the plurality of positions being operative to provide a plurality of real-time displacement measurements between the one or more bone sections of the joint;
• computing, by the at least one processor, one or more normalized relationships between the plurality of forces and the plurality of real-time displacement measurements, to provide a derived biomechanical representation of the joint state;
• and
• providing, by the at least one processor, the derived biomechanical representation of joint performance as one or more input parameters into the trained MLM.