| In the medical field,ultrasound imaging has been widely used due to its advantages of non-invasiveness and convenient diagnosis.However,in medically backward areas(such as small islands,fishing villages,poor mountainous areas,and remote military bases),due to the lack of professional ultrasound imaging doctors,patients cannot receive professional and timely treatment,thus missing the golden diagnosis time.In addition,the long-term monotonous and repetitive actions greatly increase the risk of ultrasound imaging doctors suffering from frozen shoulder,eye diseases and other occupational diseases,which greatly affects the normal work and life of doctors.Based on the above problems,researching a new ultrasonic scanning method has become an urgent problem to be solved.In recent years,robotic arm-assisted ultrasound scanning technology has gradually entered people’s field of vision because of its simplicity and efficiency,and has been widely explored and applied in abdomen,ablation,and sacral scanning.During the application process,many problems and deficiencies are exposed,such as limited auxiliary scanning parts,large time delay,and low position accuracy.In view of the common problems in this research direction,this paper realizes a remote robotic arm-assisted ultrasound diagnosis and treatment system by combining robotic arm,vision and attitude information,aiming to provide an intelligent platform for clinical remote ultrasound scanning and diagnosis.The main contributions of this paper are as follows:(1)Design and implement a point cloud-based method for locating the desired scanning area.This method mainly solves the defect that the robotic arm can only scan along the midline of the exploration site or a specific straight line in the existing research.By extracting the pose information of the desired point selected by the sonographer under the point cloud,and then performing coordinate transformation,Finally,the robot arm is controlled to move to the desired point through workspace planning.The experimental results show that the response time and path execution time of the point cloud-based path planning method are 0.6572s and 0.9577s,respectively,and the average position error of the three axes is about 2.1mm.In contrast,the response time and position error of this system are better than other methods,and have better applicability.(2)Design and implement a real-time following method of manipulator based on pose information.The method solves the problem that the existing robotic arm-assisted ultrasonic scanning method is limited to detect parts.The position and orientation sensor is used to capture the doctor’s operation action in real time,and the robot inverse kinematics algorithm is used to realize the remote accurate real-time follow-up of the robotic arm.After simulation experiments and real clinical tests,the position tracking trajectory of the robot arm end in this method is consistent with the position trajectory of the virtual probe,and the robot arm can completely reproduce the motion posture of the virtual probe.The average position errors of the X-axis,Yaxis and Z-axis are about 0.74mm,0.88mm and 0.80mm.The required time delay and followup accuracy are better than those of the existing research,and it has good universality and application prospects.(3)Design and implement a remote robotic arm-assisted ultrasound diagnosis and treatment system.In the system,by querying the real-time streaming addresses of the robotic arm end,the expert end and the ultrasound probe end,the real-time information of the three is uploaded to the system for display,which is convenient for doctors to observe the operation of the remote robotic arm in real time.At the same time,multiple experts can analyze and diagnose the patient’s ultrasound images online,and use the intelligent diagnosis module of the system to identify and process the ultrasound images collected by the robotic arm in real time,which is helpful to promote the development of clinical intelligent medicine. |
PDF Full Text Request