Search for: All records

Tactile imaging sensor determines the tumor's mechanical properties such as size, depth, and Young's modulus based on the principle of total internal reflection of light. To improve the classifying accuracy of the Tactile imaging sensor, we introduce ultrasound signals and estimate the difference in the tumor tactile images. A developed vibro-acoustic tactile imaging sensor was used to classify benign and malignant tumors. We test the developed system on breast tumor phantoms. These vibrated tactile images are analyzed to improve the overall performance of tumor detection. 

Viscoelasticity of human tissue often carries important physiological information linking to many fatal diseases, such as heart failure, renal impairment, and liver failure. Fluid retention due to these diseases cause swelling of body parts (edema) and changes the viscoelastic characteristic of the tissue. We hypothesize that the viscoelastic behavior change of the tissue can be estimated by creating and quantifying the pit on the swelled body parts. Here, we present a smartphone tactile imaging probe with an indenter (STIP-I) system that measures the pitting parameters and characterizes the viscoelastic behavior. This system consists of tactile imaging sensing that utilizes a light diffusion in a polydimethylsiloxane (PDMS)-based optical waveguide and a Viscoelastic Pitting Recovery (VPR) computation model. The prototype STIP-I system is tested using edematous tissue phantoms, which show a moderate measurement error of 29.5% for the pitting parameters and excellent performance of 7.60 % error in elastic modulus estimation. The STIP-I system is expected to bring a new approach to understanding viscoelasticity changes due to various diseases. 

Here we present a hybrid hierarchical statistical control approach for the control of robotic manipulators. The bimodal dynamic imaging system considered in this paper utilizes two robotic manipulators to move the source and detector imaging modules. As this system contains both continuous and discrete dynamics, hybrid system control techniques are applied. The robotic arms used in this research are comprised of compliant joints, which have been shown to introduce process noise into the system. To address this, a full-state feedback statistical controller is developed to minimize joint angle variations for the system. The statistical controllers for the two robot arms are then coordinated using a hierarchical controller. Finally, the feasibility of the hybrid hierarchical statistical controller is demonstrated with numerical simulations.