This study presents a tissue-independent neural network approach for predicting needle deflection in minimally invasive procedure. Precision in percutaneous interventions critically depends on accurate needle deflection estimation. A Radial Basis Function Network (RBFN) with Xavier initialization is implemented to ensure optimal weight scaling, preventing vanishing/exploding gradients and enhancing network stability. Additionally, a triangular learning rate scheduler is used to adaptively adjust learning rates, accelerating convergence and improving generalization across various insertion depths. RBFN’s localized activation mechanism ensures that deflection predictions are primarily influenced by data points near cluster centroids, effectively capturing region-specific needle-tissue interaction dynamics. The model represents needle-tissue interaction as a combination of a distributed force and a concentrated force at the bevel tip, the primary cause of deflection. Experimental validation across ten trials (90 insertions) shows a mean absolute error of 2.1 ± 1.9 mm. Upon comparing the performances of the proposed neural network model and conventional beam theory-based approach with the ground truth through statistical analysis (two-sample t-test, ANOVA) confirmed that the deflection predicted by proposed neural network-based approach is closer to the ground truth, likely to be suitable for clinical aspects.

Modular software design is a technique that increase reusability and portability of software composed from separate parts, called modules. We have designed and developed a reusable integrated software solution for robotic prostate brachytherapy procedure. The application is capable of concurrent handling of all aspects of the image-guided brachytherapy procedure: ultrasound image acquisition, anatomic delineation, target modeling, dosimetry planning and analysis, seed delivery, and visualization of all surgerical steps involved in the procedure. Based on force feedback and visual feedback, the control module of the application is capable of controlling the robotic system (i.e. motions of the ultrasound probe and the needles), supervising the flow of the procedure via built-in strategies for emergency handling and recovery, collision avoidance, manual takeover (if necessary), needle tracking and real-time dose updates. The implementation of the developed software solution to the two brachytherapy robotic systems has been presented.

Respiratory and cardiac motions induce displacement and deformation of the tumor-volume in various internal organs. To accommodate this undesired movement and other errors, physicians incorporate a large margin around the tumor to delineate Planning Target Volume (PTV), so that the Clinical Target Volume (CTV) receives the prescribed radiation dose under any scenario. Consequently, a large volume of healthy tissue is irradiated and sometimes it is difficult to spare critical organs adjacent to the tumor. In this study we have proposed a novel approach to 4D Active Tracking and Dynamic Delivery (ATDD) together with tumor motion prediction. Proposed algorithm can predict tumor position and the robotic system can continuously track the tumor during radiation dose delivery, so that a precise dose is given to a moving target while reducing dose to nearby critical organs for improved patient treatment outcome. The efficacy of the proposed method has been investigated by extensive computer simulation. The results have been presented in this article.

Prediction of tumor motion is one of the important steps in active tracking of tumor and dynamic delivery of radiation dose to tumor. In this paper, we have presented a novel adaptive acceleration-enhanced normalized least mean squares (AE-nLMS) prediction filter based on the adaptive normalized least mean squares (nLMS) filter with predicted acceleration and ratio between the real and predicted acceleration taken into account. We have compared the performances of nLMS, artificial neural network (ANN), and AE-nLMS filter for predicting the respiration motion during normal and irregular respiration. The results revealed that the ANN filter has the best performance in the prediction of normal respiration motion, whereas the AE-nLMS filter outperformed other filters in the prediction of irregular respiration motion.

In low-dose-rate (LDR) brachytherapy, precise placement of needle and accurate delivery of radioactive seeds are very important and challenging. Therefore, development of robotic systems for performing brachytherapy is gaining significant momentum. In this paper, we present a multichannel robotic system developed for image-guided brachytherapy (IGBT), especially radioactive seed implantation in prostate gland. The developed multichannel robotic system is capable of inserting a large number of needles concurrently and depositing seeds automatically. Numerous techniques perfected by a variety of experiments have been implemented in the system design and development. Thus, this system possesses potential several advantages such as reduction of target displacement, edema, and operating time as compared to single needle insertion technique. Feasibility and efficacy of this robot has been demonstrated with experimental results. Seeds can be delivered with about within 0.2 mm accuracy.

We have developed a multimodal imaging system that combines three modalities, optical spectroscopy, ultrasonography and acoustic radiation force (ARF) for improving diagnosis of breast cancer based on noninvasive interrogation of vasculature. This paper presents a detailed system design. The safety issues regarding the use of laser and ultrasound have also been addressed in this paper. The maximum exposure to skin for laser was controlled within 0.2 W·cm-2(ANSI Z136.1); exposure from ARF fields were maintained below the FDA diagnostic limit (0.72 W·cm-2). This multimodal system has the potential to improve tumor detection by deploying ARF to produce a measurable difference in the dynamic behavior of the tissue blood supply environment as interrogated by optical spectroscopy, which was demonstrated to be highly diagnostic in a murine tumor model. Pilot clinical study is being carried out.

Breast biopsy guided by imaging techniques such as ultrasound is widely used to evaluate suspicious masses within the breast. The current procedure allows the clinician to determine the location and extent of a tumor in the patient breast before inserting the needle. However, there are several problems with this procedure: the complex interaction dynamics between the needle force and the breast tissue will likely displace the tumor from its original position, necessitating multiple insertions, causing clinicianspsila fatigue, patient's discomfort, and compromising the integrity of the tissue specimen. In this paper, we present a new concept for real-time manipulation of a tumor using a robotic controller that monitors the image of the tumor to generate appropriate external force to position the tumor at a desired location. The idea here is to demonstrate that it is possible to manipulate a tumor in real time by applying controlled external force in an automated way such that the tumor does not deviate from the path of the needle. Experiments on breast phantoms are presented to demonstrate the essence of this concept. The success of this approach has the potential to reduce the number of attempts a clinician makes to capture the desired tissue specimen, minimize tissue damage, improve speed of biopsy, reduce patient discomfort, and eliminate false negative results.

Dynamic contrast enhanced magnetic resonance imaging (DCE MRI) in combination with ultrasound and/or X-ray imaging offers an excellent roadmap for intraoperative surgical guidance for complete excision of the true extent of disease in early stage breast cancer. However, incorporation of image-guidance in the operating room is challenging due to patient positioning, precise targeting under tissue deformation, instrument incompatibility and limited maneuverability, and complex human-technology interface. In this paper, we outline a novel strategy and preliminary simulation data on a proposed robot-assisted lumpectomy technique in the prone position using DCE MRI for real-time guidance. Extension to CT guidance for X-ray enhancing lesions would require minimal additional specifications. For ultrasound enhancing lesions, a method to ensure co-registration of the target with MRI is presented.

Breast biopsy guided by imaging techniques such as ultrasound is widely used to evaluate suspicious masses within the breast. The current procedure allows the physician to determine the location and extent of a tumor in the patient breast before inserting the needle. However, there are several problems with this procedure: the complex interaction dynamics between the needle force and the breast tissue will likely displace the tumor from its original position necessitating multiple insertions, causing surgeons' fatigue, patient's discomfort, and compromising the integrity of the tissue specimen. In this work, we present a new concept for real-time manipulation of a tumor using a robotic controller that monitors the image of the tumor to generate appropriate external force to position the tumor at a desired location. The idea here is to demonstrate that it is possible to manipulate a tumor in real-time by applying controlled external force in an automated way such that the tumor does not deviate from the path of the needle. A laboratory experiment has been presented on a phantom that demonstrates the essence of this concept. The success of this approach has the potential to reduce the number of attempts a surgeon makes to capture the desired tissue specimen, minimize tissue damage, improve speed of biopsy, and reduce patient discomfort.

Tumors in thorax region incur significant amount of motion and deformation due to respiratory and cardiac cycles. To accommodate this undesired movement, physicians incorporate a large standard margin around the tumor to delineate a Planning Target Volume (PTV), so that the Clinical Target Volume (CTV) receives the prescribed dose under any scenario. Consequently, a large volume of healthy tissue might be irradiated and sometimes it is difficult to spare critical organs adjacent to the tumor. For compensating this tumor motion, techniques such as breath-hold, gating, and Active Tracking and Dynamic Delivery (ATDD) are used. Although, ATDD is the most effective technique, it is the most challenging one. The ATDD can be accomplished in three different ways: adjusting the Multi-Leaf Collimator (MLC), adjusting the couch, and adjusting the MLC and couch simultaneously. The first two techniques have been explored and/or implemented in practice. However, the third approach has not been investigated extensively. In this study, we have proposed a novel approach for ATDD that exploits the advantages of both the MLC (or MLC-bank) and the HexaPOD™ robotic couch. In our proposed new approach, we have decomposed and allocated the tumor motion trajectory to the subsystems (MLC or MLC-bank and HexaPOD™ robotic couch) based on their natural frequency domains using wavelet technique. The efficacy of the proposed method has been investigated by extensive computer simulation and the results are presented in this paper.

Breast biopsy guided by imaging techniques such as ultrasound is widely used to evaluate suspicious masses within the breast. Currently, ultrasound imaging does not allow live guidance during breast biopsy. The current procedure allows the physician to determine the location and extent of a tumor in the patient breast before inserting the needle. However, there are several problems with this procedure: the complex interaction dynamics between the needle force and the breast tissue will likely displace the tumor from its original position necessitating multiple insertions, causing surgeons' fatigue, and compromising the structural integrity of the tissue specimen. In this work, we present a new concept for real-time manipulation of a tumor using a robotic force controller that monitors the image of the tumor to generate appropriate force to position the tumor at a desired location. The idea here is to demonstrate that it is possible to manipulate a tumor in real-time by applying controlled external force in an automated way such that the tumor does not deviate from the path of the needle. A laboratory experiment has been presented on a phantom that demonstrates the essence of this concept. The success of this approach has the potential to reduce the number of attempts a surgeon makes to capture the desired tissue specimen, minimize tissue damage, improve speed of biopsy, and reduce patient discomfort.

As a technique for compensating for tumor motion induced by respiratory and cardiac motions, various dynamic approaches are being investigated and implemented in radiation therapy. This paper presents a novel system that can continuously track the tumor during radiation dose delivery so that a precise dose is given to a moving target while reducing dose to nearby critical organs for improved patient treatment outcome. The efficacy of the proposed methodology has been investigated by extensive computer simulation using a Proportional, Integral and Derivative (PID) controller. System design and simulation results are presented in this paper.

In this paper we present a semi-automated device designed and developed to deliver radio-active seeds for treating prostate cancer. In the brachytherapy procedure a slander needle is inserted through the perineum and passed through different types of tissues. Thus, the needle experiences significant amount of force which may cause it to buckle and bend. In our design, we have considered the buckling force and insertion force on needle by collecting invivo data from real patient and performing in-vitro experiments. Techniques to reduce force and organ/tissue deformation have been implemented into this new design. To track the axial force on the needle for detecting pubic arch interference and to improve robotic control, we have incorporated three force sensors. Rigidity and factor of safety of the device has been analyzed using finite element method which was very useful for iterative design process.

In this paper, we present a technological development of an autonomous underwater vehicle (AUV) docking system motivated by science requirements. Twenty-seven case studies were drafted after having elaborate discussions with twelve senior marine scientists from a wide range of oceanographic fields including physical oceanography, ocean chemistry, midwater ecology, biological oceanography, molecular biology, marine microbiology, geology, evolutionary biology, and benthic ecology. These science cases spread over coastal ocean, deep-ocean, mid-ocean ridge, and polar regions. All of these science missions can be significantly benefited from new and improved sets of data and samples that will be collected by using a docked AUV. The case studies are organized to address the science issue, significance of the problem, required data and measurements to address the scientific importance, a non-docked method for acquiring data, and the benefit of using a docked AUV. More than twenty science instruments (in addition to navigational sensors) are identified and evaluated for these science missions. Science pay loads for the AUV and the docking system are discussed in detail. Functional requirements for a general purpose docking system are identified and pros and cons of various docking subsystems are explained. A fuzzy logic based solution approach for an autonomous docking system design and development is also proposed in this paper.

A new approach to the fault-accommodating allocation of thruster forces of an autonomous underwater vehicle (AUV) is investigated in this paper. This paper presents a framework that exploits the excess number of thrusters to accommodate thruster faults during operation. First, a redundancy resolution scheme is presented that considers the presence of an excess number of thrusters along with any thruster faults and determines the reference thruster forces to produce the desired motion. This framework is then extended to incorporate a dynamic state feedback technique to generate reference thruster forces that are within the saturation limit of each thruster. Results from both computer simulations and experiments are provided to demonstrate the viability of the proposed scheme.

An approach to the allocation of thruster forces of an autonomous underwater vehicle (AUV) is investigated. Generally, the number of thrusters in an AUV is more than what is minimally required to produce the desired motion. We investigate how to exploit the excess number of thrusters to accommodate thruster faults during operation. First, a redundancy resolution scheme is presented that takes into account the presence of excess number of thrusters along with any thruster faults and determines the reference thruster forces to produce the desired motion. These reference thruster forces are utilized in the thruster controller to generate the required motion. This approach resolves the thruster redundancy in the Cartesian space and allows the AUV to track the task-space trajectories with asymptotic reduction of the task-space errors. Results from both computer simulations and experiments are provided to demonstrate the viability of the proposed scheme. The paper is built upon the preliminary concept proposed earlier by Podder and Sarkar (1999).

An approach to dynamics-based trajectory planning for systems composed of subsystems with different dynamic responses is presented. This approach requires that the task-space trajectory be represented in terms of Fourier series. Various frequency components from this series are then used to generate the reference joint-space trajectories based on the natural frequencies of the respective subsystems. This approach is further extended for trajectory planning of an autonomous underwater vehicle manipulator system, where the underwater vehicle has much slower response than the on-board manipulator. The advantage of this proposed dynamics-based planning methodology is its ability to generate both kinematically admissible and dynamically feasible trajectories. Results from computer simulations are presented to demonstrate the efficacy of the proposed scheme.

An approach to the decomposition of thruster forces of an autonomous underwater vehicle (AUV) is investigated. Generally, the number of thrusters is more than what is minimally required to produce the desired motion. We investigate how to exploit the excess number of thrusters to accommodate thruster faults during operation. First, a redundancy resolution scheme is presented that takes into account the presence of excess number of thrusters and distributes the thruster forces to produce the desired motion. Then we propose a general framework that accommodates thruster faults and allows the AUV to trade task-space trajectories with asymptotic reduction of task-space errors. Results from computer simulations are provided to demonstrate the viability of the proposed scheme.

A motion coordination algorithm for an autonomous underwater vehicle-manipulator system (UVMS) is proposed. This algorithm generates the desired trajectories for both the vehicle and the manipulator in such a way that the total drag effect on the system is minimized. Resolution of kinematic redundancy of the system is performed at the acceleration level so that this algorithm can be incorporated into the system dynamics. The dynamics of the UVMS is modeled using quasi-Lagrange approach. A state-space formulation of the system along with a model-based controller design for a trajectory following task is also presented. The results from computer simulations are used to demonstrate the effectiveness of this proposed method in reducing the drag on the system.

An impedance control scheme for an autonomous underwater robotic system is proposed. This robotic system is composed of a six degree-of-freedom autonomous underwater vehicle (AUV) and a robotic arm that is mounted on the AUV. First, a dynamic model for the whole underwater vehicle-manipulator system (UVMS) considering various hydrodynamic effects is derived using a quasi-Langrange method. This model is later used to implement the proposed impedance controller. The impedance controller is designed considering the whole UVMS as one dynamic system. Extensive computer simulations are performed to verify the efficacy of the proposed control scheme and the results are presented in the paper.

The control of underwater vehicle-manipulator system (UVMS) is a challenging task because the dynamics of such a system is highly nonlinear, coupled, time-varying, and subject to hydrodynamic uncertainties and external disturbances. The paper presents a non-regressor based adaptive control for UVMS where centralized and coordinated trajectory planning is followed by decentralized control-one for the slower subsystem (vehicle) and other for the faster subsystem (manipulator). To avoid joint limits, kinematic redundancy of the system is resolved using the weighted pseudoinverse of the Jacobian matrix. The results from computer simulations demonstrate the effectiveness of this proposed method.