Path planning for minimally-invasive knee surgery using a hybrid optimization procedure

A Planar Model of an Ankle Joint with Optimized Material Parameters and Hertzian Contact Pairs

The ankle is one of the most complicated joints in the human body. Its features a plethora of elements with complex behavior. Their functions could be better understood using a planar model of the joint with low parameter count and low numerical complexity. In this study, an accurate planar model of the ankle with optimized material parameters was presented. In order to obtain the model, we proposed an optimizational approach, which fine-tuned the material parameters of two-dimensional links substituting three-dimensional ligaments of the ankle. Furthermore, the cartilage in the model was replaced with Hertzian contact pairs. The model was solved in statics under moment loads up to 5 Nm. The obtained results showed that the structure exhibited angular displacements in the range of the ankle joint and that their range was higher in dorsiflexion than plantarflexion. The structure also displayed a characteristic ramp up of the angular stiffness. The results obtained from the optimized model were in accordance with the experimental results for the ankle. Therefore, the proposed method for fine-tuning the material parameters of its links could be considered viable.

Structural and Material Optimization for Automatic Synthesis of Spine-Segment Mechanisms for Humanoid Robots with Custom Stiffness Profiles

Typical artificial joints for humanoid robots use actual human body joints only as an inspiration. The load responses of these structures rarely match those of the corresponding joints, which is important when applying the robots in environments tailored to humans. In this study, we proposed a novel, automated method for designing substitutes for a human intervertebral joint. The substitutes were considered as two platforms, connected by a set of flexible links. Their structural and material parameters were obtained through optimization with a structured Genetic Algorithm, based on the reference angular stiffnesses. The proposed approach was tested in three numerical scenarios. In the first test, a mechanism with angular stiffnesses corresponded to the actual L4–L5 intervertebral joint. Scenarios 2 and 3 featured mechanisms with geometry and structure comparable to the joint, but with custom stiffness profiles. The obtained results proved the effectiveness of the proposed method. It could be employed in the design of artificial joints for humanoid robots and orthotic structures for the human spine. As the approach is general, it could also be extended to different body joints.