Electrical impedance tomography: a potential tool for intraoperative imaging of the tongue base

The presence of a tumor in the tongue is a pathology that requires surgical intervention from a certain stage. This type of surgery is difficult to perform because of the limited space available around the base of the tongue for the insertion of surgical tools. During the procedure, the surgeon has to stretch and then fix the tongue firmly in order to optimize the available space and prevent tissue movement. As a result, the preoperative images of the inside of the tongue no longer give a reliable indication of the position and shape of the cancerous tissue due to the deformation of the overall tissue in the area. Thus, new images are needed during the operation, but are very difficult to obtain using conventional techniques due to the presence of surgical tools. Electrical Impedance Tomography (EIT) is an imaging technique that maps the resistivity or difference of resistivity of biological tissues from electrical signals. The small size of the electrodes makes it a potentially interesting tool to obtain intraoperative images of the inside of the tongue. In this paper, the possibility of using EIT for this purpose is investigated. A detection method is proposed, including an original configuration of the electrodes, consistent with the anatomical specificities of the tongue. The proposed method is studied in simulation and then a proof of concept is obtained experimentally on a 3D printed test tank filled with saline solution and plant fibres.

Tactile Sensing for Minimally Invasive Surgery: Conventional Methods and Potential Emerging Tactile Technologies

As opposed to open surgery procedures, minimally invasive surgery (MIS) utilizes small skin incisions to insert a camera and surgical instruments. MIS has numerous advantages such as reduced postoperative pain, shorter hospital stay, faster recovery time, and reduced learning curve for surgical trainees. MIS comprises surgical approaches, including laparoscopic surgery, endoscopic surgery, and robotic-assisted surgery. Despite the advantages that MIS provides to patients and surgeons, it remains limited by the lost sense of touch due to the indirect contact with tissues under operation, especially in robotic-assisted surgery. Surgeons, without haptic feedback, could unintentionally apply excessive forces that may cause tissue damage. Therefore, incorporating tactile sensation into MIS tools has become an interesting research topic. Designing, fabricating, and integrating force sensors onto different locations on the surgical tools are currently under development by several companies and research groups. In this context, electrical force sensing modality, including piezoelectric, resistive, and capacitive sensors, is the most conventionally considered approach to measure the grasping force, manipulation force, torque, and tissue compliance. For instance, piezoelectric sensors exhibit high sensitivity and accuracy, but the drawbacks of thermal sensitivity and the inability to detect static loads constrain their adoption in MIS tools. Optical-based tactile sensing is another conventional approach that facilitates electrically passive force sensing compatible with magnetic resonance imaging. Estimations of applied loadings are calculated from the induced changes in the intensity, wavelength, or phase of light transmitted through optical fibers. Nonetheless, new emerging technologies are also evoking a high potential of contributions to the field of smart surgical tools. The recent development of flexible, highly sensitive tactile microfluidic-based sensors has become an emerging field in tactile sensing, which contributed to wearable electronics and smart-skin applications. Another emerging technology is imaging-based tactile sensing that achieved superior multi-axial force measurements by implementing image sensors with high pixel densities and frame rates to track visual changes on a sensing surface. This article aims to review the literature on MIS tactile sensing technologies in terms of working principles, design requirements, and specifications. Moreover, this work highlights and discusses the promising potential of a few emerging technologies towards establishing low-cost, high-performance MIS force sensing.

Design and implementation of a medical device test stand for micro-catheters and guide-wires

The use of catheters and guide-wires in minimally invasive therapeutic approaches is an important part of clinical practice. In the neurovascular field, the unique nature of cerebral blood vessels necessitates very thin-walled and flexible catheters. The blood vessels in question are highly branched and at the same time can be less than one millimetre in diameter. This results in high demands on micro-catheters and guide-wires for successful endovascular therapy. The interaction of these surgical tools and the vessel wall is of especial interest. Depending on the catheter stiffness, this interaction can be friction, punctual collision or straightening. The work aims to design and implement a test setup for evaluation of these interactions with the vessel wall. For this purpose, a standardized vessel course with representative characteristics is necessary. Furthermore, by implementing suitable measuring equipment, an endovascular intervention can be simulated.

Assessing Generalisation Capabilities of CNN Models for Surgical Tool Classification

Accurate recognition of surgical tools is a crucial component in the development of robust, context-aware systems. Recently, deep learning methods have been increasingly adopted to analyse laparoscopic videos. Existing work mainly leverages the ability of convolutional neural networks (CNNs) to model visual information of laparoscopic images. However, the performance was evaluated only on data belonging to the same dataset used for training. A more comprehensive evaluation of CNN performance on data from other datasets can provide a more rigorous assessment of the approaches. In this work, we investigate the generalisation capability of different CNN architectures to classify surgical tools in laparoscopic images recorded at different institutions. This research highlights the need to determine the effect of using data from different surgical sites on CNN generalisability. Experimental results imply that training a CNN model using data from multiple sites improves generalisability to new surgical locations.

Adaptive kernel selection network with attention constraint for surgical instrument classification

Computer vision (CV) technologies are assisting the health care industry in many respects, i.e., disease diagnosis. However, as a pivotal procedure before and after surgery, the inventory work of surgical instruments has not been researched with the CV-powered technologies. To reduce the risk and hazard of surgical tools’ loss, we propose a study of systematic surgical instrument classification and introduce a novel attention-based deep neural network called SKA-ResNet which is mainly composed of: (a) A feature extractor with selective kernel attention module to automatically adjust the receptive fields of neurons and enhance the learnt expression and (b) A multi-scale regularizer with KL-divergence as the constraint to exploit the relationships between feature maps. Our method is easily trained end-to-end in only one stage with few additional calculation burdens. Moreover, to facilitate our study, we create a new surgical instrument dataset called SID19 (with 19 kinds of surgical tools consisting of 3800 images) for the first time. Experimental results show the superiority of SKA-ResNet for the classification of surgical tools on SID19 when compared with state-of-the-art models. The classification accuracy of our method reaches up to 97.703%, which is well supportive for the inventory and recognition study of surgical tools. Also, our method can achieve state-of-the-art performance on four challenging fine-grained visual classification datasets.