ABSTRACT Although the cardiac ablation procedure for atrial fibrillation has a wide adoption rate, it also has a high recurrence of arrhythmia primarily due to the creation of suboptimal lesions. The procedure is also associated with complications including cardiac perforation, tamponade, atrio-esophageal fistulas and thrombus. Repeated and prolonged X-ray exposure for the clinician can also lead to enhanced risk of cancer. A fluoroless approach using intracardiac echocardiography (ICE) is becoming a more widely adopted imaging option due to the absence of ionizing radiation and the possibility of real-time monitoring of the created lesions. However, the ICE- guided approach suffers from significant shortcomings, which include poor dexterity of the ICE catheter, difficulty in simultaneously manipulating the ICE and ablation catheters, unintuitive image orientation and noisy image quality. There is therefore an unmet need to overcome these shortcomings of the ICE-guided approach to enable better lesion creation and reduced complications associated with the cardiac ablation procedure. The long-term goal of this research is to develop robotic technologies, control and machine learning algorithms to enable ICE- guided cardiac ablation procedures. The objective is to develop a novel robotic manipulator, a steerable ICE catheter, and machine learning and control algorithms to manipulate the ICE catheter and monitor the created lesions in real-time. The rationale that underlies the proposed research is that the robot-assisted steerable ICE catheter with the catheter tracking algorithms will enable simultaneous manipulation of the ICE and ablation catheters. Further, the machine learning algorithms to monitor therapy will reduce the risk of complications, while ensuring the creation of necrotic lesions, thereby reducing the recurrence of AF. In this proposal, we plan to pursue the following specific aims: 1) Design, develop, and model a steerable 3D ICE catheter with enhanced dexterity. 2) Design and develop a robotic manipulator and associated control algorithms to allow for precise manipulation of the ICE catheter. 3) Develop machine-learning and vision-based algorithms integrated with a navigation system for tracking the ablation catheter, and monitoring therapy. 4) Validate the robotic ICE system in heart phantom and porcine models. The proposed research is significant since it will allow for better therapeutic outcomes by reducing recurrence rates and complications associated with cardiac ablation, and avoiding exposure of the patient and clinical care team to X-ray radiation. The proposed research is innovative in that it builds on state-of-the-art robotics technology, machine learning and vision algorithms to enable fluoroless ICE- guided cardiac ablation procedures.