Research on autonomous surgery has largely focused on simple task automation in controlled environments. However, real-world surgical applications demand dexterous manipulation over extended durations and robust generalization to the inherent variability of human tissue. These challenges remain difficult to address using existing logic-based or conventional end-to-end learning strategies. To address this gap, we propose a hierarchical framework for performing dexterous, long-horizon surgical steps. Our approach uses a high-level policy for task planning and a low-level policy for generating low-level trajectories. The high-level planner plans in language space, generating task-level or corrective instructions that guide the robot through the long-horizon steps and help recover from errors made by the low-level policy. We validated our framework through ex vivo experiments on cholecystectomy, a commonly practiced minimally invasive procedure, and conducted ablation studies to evaluate key components of the system. Our method achieves a 100% success rate across eight different ex vivo gallbladders, operating fully autonomously without human intervention. The hierarchical approach improved the policy’s ability to recover from suboptimal states that are inevitable in the highly dynamic environment of realistic surgical applications. This work demonstrates step-level autonomy in a surgical procedure, marking a milestone toward clinical deployment of autonomous surgical systems.

Laser speckle contrast imaging (LSCI) utilizes the speckle pattern of a laser to determine the blood flow in tissues. The current approaches for its use in a clinical setting require a camera system with a laser source on a separate optical axis making it unsuitable for minimally invasive surgery (MIS). With blood flow visualization, bowel viability, for example, can be determined. Thus, LSCI can be a valuable tool in gastrointestinal surgery. In this work, we develop the first-of-its-kind dual-display laparoscopic vision system integrating LSCI with a commercially available 10mm rigid laparoscope where the laser has the same optical axis as the laparoscope. Designed for MIS, our system permits standard color RGB, label-free vasculature imaging, and fused display modes. A graphics processing unit accelerated algorithm enables the real-time display of three different modes at the surgical site. We demonstrate the capability of our system for imaging relative flow rates in a microfluidic phantom with channels as small as 200 μm at a working distance of 1–5 cm from the laparoscope tip to the phantom surface. Using our system, we reveal early changes in bowel perfusion, which are invisible to standard color vision using a rat bowel occlusion model. Furthermore, we apply our system for the first time for imaging intestinal vasculature during MIS in a swine.

The authors describe the development of an ultrafast three-dimensional (3D) optical coherence tomography (OCT) imaging system that provides real-time intraoperative video images of the surgical site to assist surgeons during microsurgical procedures. This system is based on a full-range complex conjugate free Fourier-domain OCT (FD-OCT). The system was built in a CPU-GPU heterogeneous computing architecture capable of video OCT image processing. The system displays at a maximum speed of $10volume/s$ for an image volume size of $160×80×1024$$(X×Y×Z)$ pixels. We have used this system to visualize and guide two prototypical microsurgical maneuvers: microvascular anastomosis of the rat femoral artery and ultramicrovascular isolation of the retinal arterioles of the bovine retina. Our preliminary experiments using 3D-OCT-guided microvascular anastomosis showed optimal visualization of the rat femoral artery ($diameter<0.8mm$), instruments, and suture material. Real-time intraoperative guidance helped facilitate precise suture placement due to optimized views of the vessel wall during anastomosis. Using the bovine retina as a model system, we have performed “ultra microvascular” feasibility studies by guiding handheld surgical micro-instruments to isolate retinal arterioles ($diameter∼0.1mm$). Isolation of the microvessels was confirmed by successfully passing a suture beneath the vessel in the 3D imaging environment.

Accurate, real-time identification and display of critical anatomic structures, such as the nerve and vasculature structures, are critical for reducing complications and improving surgical outcomes. Human vision is frequently limited in clearly distinguishing and contrasting these structures. We present a novel imaging system, which enables noninvasive visualization of critical anatomic structures during surgical dissection. Peripheral nerves are visualized by a snapshot polarimetry that calculates the anisotropic optical properties. Vascular structures, both venous and arterial, are identified and monitored in real-time using a near-infrared laser-speckle-contrast imaging. We evaluate the system by performing in vivo animal studies with qualitative comparison by contrast-agent-aided fluorescence imaging.

Intestinal anastomosis is a surgical procedure that restores bowel continuity after surgical resection to treat intestinal malignancy, inflammation, or obstruction. Despite the routine nature of intestinal anastomosis procedures, the rate of complications is high. Standard visual inspection cannot distinguish the tissue subsurface and small changes in spectral characteristics of the tissue, so existing tissue anastomosis techniques that rely on human vision to guide suturing could lead to problems such as bleeding and leakage from suturing sites. We present a proof-of-concept study using a portable multispectral imaging (MSI) platform for tissue characterization and preoperative surgical planning in intestinal anastomosis. The platform is composed of a fiber ring light-guided MSI system coupled with polarizers and image analysis software. The system is tested on ex vivo porcine intestine tissue, and we demonstrate the feasibility of identifying optimal regions for suture placement.