Clinical studies have shown that speech pauses can reflect the cognitive function differences between Alzheimer’s Disease (AD) and non-AD patients, while the value of pause information in AD detection has not been fully explored. Herein, we propose a speech pause feature extraction and encoding strategy for only acoustic-signal-based AD detection. First, a voice activity detection (VAD) method was constructed to detect pause/non-pause feature and encode it to binary pause sequences that are easier to calculate. Then, an ensemble machine-learning-based approach was proposed for the classification of AD from the participants’ spontaneous speech, based on the VAD Pause feature sequence and common acoustic feature sets (ComParE and eGeMAPS). The proposed pause feature sequence was verified in five machine-learning models. The validation data included two public challenge datasets (ADReSS and ADReSSo, English voice) and a local dataset (10 audio recordings containing five patients and five controls, Chinese voice). Results showed that the VAD Pause feature was more effective than common feature sets (ComParE: 6373 features and eGeMAPS: 88 features) for AD classification, and that the ensemble method improved the accuracy by more than 5% compared to several baseline methods (8% on the ADReSS dataset; 5.9% on the ADReSSo dataset). Moreover, the pause-sequence-based AD detection method could achieve 80% accuracy on the local dataset. Our study further demonstrated the potential of pause information in speech-based AD detection, and also contributed to a more accessible and general pause feature extraction and encoding method for AD detection.

Background: Transforaminal percutaneous endoscopic lumbar surgeries (PELS) for lumbar disc herniation and spinal stenosis are growing in popularity. However, there are some problems in the establishment of the working channel and foraminoplasty such as nerve and blood vessel injuries, more radiation exposure, and steeper learning curve. Rapid technological advancements have allowed robotic technology to assist surgeons in improving the accuracy and safety of surgeries. Therefore, the purpose of this study is to develop a robot-assisted system for transforaminal PELS, which can provide navigation and foraminoplasty. Methods: The robot-assisted system consists of three systems: preoperative planning system, navigation system, and foraminoplasty system. In the preoperative planning system, 3D visualization of the surgical segment and surrounding tissues are realized using the multimodal image fusion technique of computed tomography and magnetic resonance imaging, and the working channel planning is carried out to reduce the risk for injury to vital blood vessels and nerves. In the navigation system, the robot can obtain visual perception ability from a visual receptor and automatically adjust the robotic platform and robot arm to the appropriate positions according to the patient’s position and preoperative plan. In addition, the robot can automatically register the surgical levels through intraoperative fluoroscopy. After that, the robot will provide navigation using the 6 degree-of-freedom (DOF) robot arm according to the preoperative planning system and guide the surgeon to complete the establishment of the working channel. In the foraminoplasty system, according to the foraminoplasty planning in the preoperative planning system, the robot performs foraminoplasty automatically using the high speed burr at the end of the robot arm. The system can provide real-time feedback on the working status of the bur through multi-mode sensors such as multidimensional force, position, and acceleration. Finally, a prototype of the system is constructed and performance tests are conducted. Discussion: Our study will develop a robot-assisted system to perform transforaminal PELS, and this robot-assisted system can also be used for other percutaneous endoscopic spinal surgeries such as interlaminar PELS and percutaneous endoscopic cervical and thoracic surgeries through further research. The development of this robot-assisted system can be of great significance. First, the robot can improve the accuracy and efficiency of endoscopic spinal surgeries. In addition, it can avoid multiple intraoperative fluoroscopies, minimize exposure to both patients and the surgical staff, shorten the operative time, and improve the learning curve of beginners, which is beneficial to the popularization of percutaneous endoscopic spinal surgeries.

Objective : Current surgical robots face challenges to understand preoperative images like human surgeons, which hindering robots from making full use of preoperative information to operate stably and efficiently. We offer a method to estimate drilling force information based on preoperative images to provide a priori force information for surgical robots performing bone drilling tasks. Methods: A visual sensing computing framework is proposed to help robots calculate drill-tissue contact area 3D image information in a one-dimensional signal format. Under this computing framework, a computed tomography (CT) image-weighted bone drilling mechanical model is built, which further considers both targets bone shape and material properties to predict the thrust force, torque, and radial force of a drilling process based on preoperative CT images. Results: The built model can respond to multiple bone drilling process factors, such as personalized surgery plans, varying tissue densities, uneven drilling surfaces, different drilling speeds, feed rates, and drill bit geometries. The best trust force prediction error on bovine bones is 1.13 ± 0.95 N, and the best normalized average prediction error on porcine bones is 0.07 ± 0.08. Experiments in spinal pedicle screw placement surgery also show potential application abilities. Conclusion: Our method predicts the bone drilling force well based on preoperative images, providing robots with more efficient preoperative information. Significance: This work offers a new perspective to study the interaction relationship between robot surgical instruments and tissues with the assistance of preoperative images.