Contact Compliance Based Visual Feedback for Tool Alignment in Robot Assisted Bone Drilling.

In recent decades, robot-assisted surgery has been proven superior at providing more accurate outcomes than the conventional one, particularly in minimally invasive procedures. However, there are still limitations to these kinds of surgical robots. Accurate bone drilling on the steep and hard surface of cortical bone is still challenging. The issues of slipping away from the target entry point on the bone surface and subsequently deviating from the desired path are still not completely solved. Therefore, in this paper, a force control is proposed to accompany the resolved motion rate controller in a handheld orthopedic robot system. The force control makes it possible to adjust the contact compliance of the drill to the bone surface. With the proper contact compliance, the drill can be prevented from deflecting in contact with the bone surface, and will eventually be directed to the target entry point. The experiments on test jig and vertebra phantom also show that the robot under the proposed contact compliance visual feedback control structure could produce better usability positioning accuracy under various contact disturbances than its counterpart.

A robotized handheld smart tool for orthopedic surgery.

BACKGROUND: Handheld surgical robots offer functionalities, such as active guidance, tremor suppression and force reflection, for surgeons to enhance their skill in manipulating surgical tools during medical intervention. In orthopedic surgery, the robot additionally has to offer sufficient rigidity and power for bone machining. The size and weight of the mechanical design, together with the control behaviour associated with involuntary hand motion, navigation and reflected force to the human, all influence the overall performance of an orthopedic handheld robot. METHODS: The paper proposes a miniature and compact design for an embedded robot, which is a similar weight as a handpiece. Then, a shared controller is proposed to address the coupling among involuntary and voluntary hand motions, robot navigation, tool feedback forces and force artefacts from actuation. RESULTS: The handheld robot is able to stabilize the drill positioning by removing involuntary tremors as well as reduce force artefacts from motor actuation in experiments involving pedicle tunnelling on a porcine spine. CONCLUSION: The paper has successfully realized a compact handheld orthopedic robot which provides high performance of usability, tremor suppression and force reflection for bone drilling.

Coordinated control of bone cutting for a CT-free navigation robotic system in total knee arthroplasty.

BACKGROUND: Robots are gradually becoming intelligent tools for surgeons in computer assisted orthopedic surgery. A hands-on robot combining CT-free navigation software and coordinated control has been designed for total knee arthroplasty. METHODS: The hands-on robot is under bilateral force control so that the robot not only follows the force commands from the operator but also shapes the reflected force back to the operator. The proposed coordinated control also defines three operating modes to fulfill intelligent bone cutting by tuning appropriate scaling to the admittance gains. RESULTS: Experimental results demonstrated assistive, resistant and emergent actions to meet various bone cutting conditions by the coordinated controller. CONCLUSIONS: The proposed coordinated controller enables the robot to perform bone cutting more safely, rapidly and accurately compared with being performed by manual bone saw. The intelligent bone cutting has been successfully verified by a cadaver test.

Vascular morphologic information of three-dimensional power Doppler ultrasound is valuable in the classification of breast lesions.

Doppler ultrasound imaging provides vascular information that could characterize benign and malignant breast masses in many previous publications. In this study, we applied vascular quantification and morphology features derived from three-dimensional power Doppler ultrasound as classifiers based on support vector machine. An Az value under the receiver operating characteristic (ROC) curve was used to measure the significance of each vascularization feature. Sixty solid breast tumors were assessed. According to the Az value for the ROC curve of the selected features, the classification performance of the proposed method was 0.8423, indicating that vascular morphologic information is valuable in the classification of breast lesions.

Development of a stiffness measurement accessory for ultrasound in breast cancer diagnosis.

When ultrasound imaging is used for breast cancer diagnosis, the lesion's morphology is usually used to determine if the lesion is benign or malignant. Sometimes, the information provided by the procedure may not be adequate to make an accurate judgment. Additional information is needed, such as the stiffness of the lesion relative to its surrounding tissue. This paper presents an ultrasound accessory device designed to achieve this purpose. The device is easy to operate and is similar in use to a normal clinical breast ultrasound examination. A sonologist must only attach an ultrasound probe to the device and then slide it across the lesion maintaining a constant compression depth. The built-in inverse biomechanical model of the device will then calculate the predicted stiffness ratio of the lesion relative to its surrounding tissue based on the measured palpation data. Modelling and experiments have been performed on phantoms with embedded inclusions. The experimental results show that the stiffness ratio of the inclusion to its surrounding material can be accurately predicted by the handheld device. A preliminary clinical test was also performed to demonstrate the use of this device in vivo, to offer additional information to aid classification of the tumor as benign or malignant when complemented with ultrasound images.

Lateral exploration strategy for differentiating the stiffness ratio of an inclusion in soft tissue.

The stiffness ratio between an inclusion and the surrounding tissue provides critical information for tumor classification. Malignant tumors are usually harder than benign ones. Accuracy and efficiency of computing tissue stiffness depends on how external excitations are applied to the tissue and what kind of biomechanical model is used. In this paper, a lateral exploration strategy combined with an inverse biomechanical model based on an artificial neural network has been proposed to identify inclusion properties. The experimental results showed that the proposed method was able to predict the inclusion properties with better accuracy and significantly improved computational efficiency as compared to the conventional indentation method.