11 January 2024 in IEEE Sensors Journal
Robot-assisted membrane peeling provides several benefits including improvement of manipulation precision and safety. A forceps which is capable of sensing peeling force, scleral force, and insertion depth is the basis of developing robot control method. But such forceps still lacks. To relieve these challenges, a multi-function sensing forceps based on fiber bragg gratings (FBG) is proposed. Six FBG sensors are divided into three groups and attached along the grooves of a 23G forceps. The first FBG group aims to measure the peeling force, while the other two FBG groups measure the scleral force and insertion depth. The calculation algorithm is built. Both linear calibration method and radial basis function (RBF) neural network calibration method are proposed to improve the sensing accuracy. Experimental results show that the precision of the RBF neural network is approximately 76.28% higher than that of linear calibration method. Verification experiments results show that the average root mean square error of the proposed forceps are 0.28 mN (peeling force), 0.37 mN (scleral force), and 0.29 mm (insertion depth), respectively. The tested sensing range are 30 mN (peeling force), 100 mN (scleral force), 10 mm (insertion depth), respectively.
15 April 2025 in IEEE/ASME Transactions on Mechatronics
In retinal surgery, membrane peeling is a challenging procedure that requires surgeons to limit peeling force and manage disturbance force at the millinewton level. This article presents a novel handheld compliant robot designed to assist in membrane peeling, featuring a five-chain parallel compliant module and a force-sensing tube. The compliant module is used to adjust the tip's transverse position actively, while the tube is designed to adsorb the membrane and sense the peeling force. The Jacobian matrix of the proposed robot is also derived. Then, a data process method based on wavelet decomposition and Kalman filtering is introduced to identify the disturbance force. A control method based on the Jacobian matrix is also proposed to compensate the disturbance force. Finally, the proposed robot is evaluated through bench test and handheld test, in which chicken chorioallantoic membrane is utilized as the ex-vitro model. The handheld test results show that the root mean square and maximum of disturbance force are reduced by 44.1% and 52.7%, respectively. The membrane detaches from the tube when the peeling force approaches 15 mN, which provides a passive safety guarantee of membrane peeling.