OBOT-ASSISTED minimally invasive surgery (RMIS) has emerged as an approach that allows surgeons to perform complicated surgical procedures with improved dexterity, visualization, and precision and accuracy that can collectively enhance treatment outcomes [1,2]. Advanced robotic surgical systems, such as the da Vinci ® system (Intuitive Surgical, Inc.) and Senhance Surgical System (TransEntrix Inc.), offer multiple advantages, including increased degrees of freedom, high-definition visualization of the surgical site with accurate depth perception, and enhanced scalability [3]. RMIS performed using these surgical systems provides unique benefits to patients, including reduced pain and discomfort, smaller incisions, minimal blood loss, and faster recovery time [4]. Accordingly, RMIS is becoming increasingly used for a wide range of specialties, including thoracoscopic, hepatobiliary, gynecologic, urologic and gastrointestinal surgery [5].
Despite the numerous advantages and benefits, one of the widely recognized limitations of RMIS is the absence of tactile sensations (i.e., touch-and force-related sensations) [6]. During traditional open surgery, surgeons often use tactile feedback through manual palpation to examine the pathologic conditions of the tissues [7]. In particular, because pathologic tissues, such as tumors [8] and fibrosis [9], are stiffer than normal tissues, intra-operative manual palpation enables surgeons to identify diseased tissues that must be surgically treated. However, during RMIS, surgeons rely on visual information to assess the tissues because the use of robot arms for surgical operation limits their ability to receive tactile feedback [10].
One potential solution to overcoming the loss of tactile perceptions during RMIS is to remove tissue from the organ and evaluate the biological, mechanical, or physiological characteristics of the removed tissue through conventional assays, such as immunostaining, tensile testing, and gene profiling [11,12]. While the ex vivo measurements can provide detailed insight into the molecular and cellular characteristics of the tissues, they are slow and destructive as tissue removal is required for the downstream analyses. Further, the accuracy of measurements may be compromised as the native microstructure and biological properties of the tissues can change when removed from the body. Pre-operative imagingbased analysis modalities, such as computed tomography (CT), magnetic resonance imaging (MRI), and elastography, can allow rapid and non-destructive determination of tissue structure and properties. However, these imaging modalities remain limited by their special resolution and only provide a single historical snapshot which is often difficult for the surgeon to utilize in real-time during an operation [13,14].
Since mechanical properties of tissue are often associated with tissue state and pathology [15], several studies have investigated quantitative tissue stiffness assessment using mechanical palpation approaches, including rolling indentation [16][17][18], vertical indentation [19][20][21][22], vibration [23], grasping [24,25], and sweeping [26]. Typically, the palpation devices incorporate tactile sensors [27], such as electrical-based and optical-based sensors, to obtain force-and touch-related information. By applying deformation on the tissue surface and measuring the resulting forces, tissue stiffness can be quantified through mathematical modeling of the force-deformation relationship. Such devices have been useful in generating highresolution stiffness distribution maps of tissues, which have effectiveness in localizing tumors in ex vivo [16,28] and in vivo [19,21] cancer models. Further, few studies have explored a combinative approach using quantitative analyses and experimental palpation to determine the viscoelastic properties of healthy and pathological tissues [22,29].
The size, sensitivity, and sterilizability of the measurement probe are key factors considered in the development of palpation devices for RMIS. To minimize the required space for intraoperative application of a palpation device, few studies have developed wireless and Bluetooth-based palpation probes [21,30], while several studies have explored the incorporation of the palpation mechanism into existing surgical tools, such as endoscopes [29,31], robotic arms [19,32], and surgical forceps [24,25,33]. Notably, both approaches allow for the deployment of the probe through the RMIS trocar port, facilitating intraoperative stiffness measurements.
Additionally, integrating the palpation probes into surgical tools can reduce the time required for surgeons' training [33]. Moreover, the sensitivity of the palpation probe is a crucial design factor, as a highly sensitive probe allows precise identification of tumor margins that must be surgically resected and the detection of anatomical structures, such as nerves and blood vessels, that must be avoided during surgical procedures [21]. Furthermore, a number of palpation devices are designed to be either sterilizable [31] or disposable [34,35] to align with the clinical requirements.
In this study, we present a robotic tissue palpation device that can accurately, rapidly, and minimally invasively quantify tissue stiffness during RMIS by determining elastic modulus (E) in situ (Figs. 1A and1B). Notably, the measurement probe in our device incorporates a novel electrical-based sensing method to monitor the predefined tissue deformation. Existing palpation probes utilize optical [16,20], mechanical [21,36], or computer-assisted [17,19,22,28] techniques to determine tissue deformation. Within our probe's design, two contact electrodes are integrated adjacent to the sensor head (i.e., hemispheric indenter) to provide real-time feedback on the maximum predefined deformation (Figs, 1C and1D). To the best of our knowledge, the use of contact electrodes to indicate resulting predefined deformation has not been previously demonstrated. We constructed the device and tested its functionality of the device prototype by measuring the stiffness of tissue mimics and biological tissues obtained from animal models. The measurement accuracy and reproducibility were confirmed by comparing the readouts with the values reported in the literature. Our device was able to accurately locate tumor mimetics in cancer models. Collectively, we confirmed that our device has the great potential to provide an accurate and objective tissue stiffness assessment that can ultimately allow surgeons to detect and treat pathologic tissues, such as small tumors, during robotic surgery.
The robotic tissue palpation device has been designed to intra-operatively and non-destructively quantify the stiffness of soft tissues (e.g., lung, heart, and liver) by measuring the elastic modulus of the tissue (Fig. 1A). The measurement sensor probe incorporated into the distal end of the catheter consisted of a thin force sensor, a rigid hemispheric indenter head, two contact electrodes, and an imaging probe (Figs. 1B and1C). To determine the tissue stiffness, the probe was directed downward perpendicular to the tissue surface (Fig. 1D,i). A force was then continuously applied against the tissue surface at a rate of 5 mm/min via the hemispheric indenter to compress the tissue. Predefined magnitude of tissue deformation, which was determined by the height of the indenter (i.e., 2 mm), was confirmed when the electrical signal was detected across the two contact electrodes (Fig. 1D, ii). Subsequently, the elastic modulus of the tissue was calculated by correlating the magnitude of the compression force (FC) and the deformation length of the tissue (LC) using Hertz's equation developed for spherical indentation:
where E is the elastic modulus of the tissue, FC is the force measured by force sensor at maximum deformation, LC is the maximum deformation, is Poisson's ratio, and r is the radius of the hemispheric indenter (Supplementary Methods). The Poisson's ratio was assumed to be 0.5 for the tissue phantoms and the biological tissues [37]. A single indentation was performed against the tissue surface for each measurement.
The device consisted of three key components: i) a deployable sensing probe mounted on a steerable catheter that can compress local tissue for stiffness measurement, ii) a motion control module that enables multi-directional device movements, such as linear displacement, rotation, and deflection of the device, and iii) a micro-optical imaging module that provides the visual information during the probe navigation and tissue stiffness measurement.
The stiffness measurement sensing probe consisted of a thin film-based force sensor (diameter: 3 mm; GD03-10N, UNEO), two spiral-headed pogo pins as contact electrodes (diameter: 1 mm, MilMax), and a rigid hemispheric compression head made of acrylic plastic (height: 2 mm). To utilize the force sensor in the stiffness measurement application, a force-to-voltage circuit was created using a digital data acquisition device (Arduino UNO, Rev 3), a reference resistor (R, Microchip), and customwritten MATLAB code (MATLAB R2021). In addition, an electrical circuit was formed between the contact electrodes (i.e., pogo pins) to quantify tissue deformation through acquisition of voltage as the readouts.
The steerable catheter arm was designed as a wire-driven continuum robot and constructed with driving disks, driving wires, and flexible tubing. The robotic catheter was integrated into the tissue palpation device via a 3D printed adapters made of poly(lactic acid) (PLA; MakerBot), which was mounted on the device using dovetail rail carriers and a dovetail optical rail. A complete description of the measurement probe, the deflectable catheter, and the electrical/optical circuits is provided in the Supplementary Methods.
Deflection, tilting, and linear displacement movements were enabled in the device to guide the measurement probe to the target tissue surface (Fig. 2, Fig. S1). The movement control module consisted of a linear motor (HDLS-8-50-12V, Robotzone) that controlled the tilting movement, a servo motor (HS-785HB, HITEC RCD) that modulates the translational movement, and two high-precision servo motors (HS-311, HITEC RCD) that control the deflection movements of the catheter's distal end. During the device operation, all the motors were controlled simultaneously using two motor controllers (605105, Acrobatics).
The deflection movement of the catheter arm was enabled by a wire-driven conformation that allowed multi-angular movements of the device tip in three-dimension (3D) space (Fig. 2A, Fig. S2, Movie S1). The desirable probe deflection was achieved by pushing and pulling the driving wires via pulleys, mounted on servo motors. Tilting movement was enabled by extending or retracting a linear servo integrated into the device (Fig. 2B). Specific tilting motions of the device could be achieved by adjusting the length of the linear servo arm (Movie S2). Furthermore, a servo motor was installed on the back of the rail to control the linear motion of the measurement probe (Fig. 2C,Movie S3).
A custom-built imaging module was integrated into the tissue palpation device that allowed vision-assisted navigation during device operation and tissue stiffness measurements (Fig. 3A). The imaging module consisted of an LED illumination light source (M565L3, Thorlabs), a flexible optical-fiber imaging bundle with embedded SELFOC micro-lens (FIGH-30-650S, Fujikura), a monochrome CMOS camera (Mako, Allied Vision), an achromatic doublet (tube lens; AC254-150-A-ML, Thorlabs), a 10 objective lens (Olympus), a filter holder (STK01, Thorlabs), a fiber bundle adapter (S120-SMA, Thorlabs), a translating lens mount (CXY1, Thorlabs), and extension tubes (Thorlabs).
For imaging an object, such as an organ, the distal end of the imaging fiber bundle (i.e., imaging tip) was introduced into the imaging channel of the catheter, and the image formed on the proximal end passed through the objective lens and collected by the camera's imaging sensor (Fig. 3B and3C). The resolution of the imaging module was approximately 80 m/line pair (one light line and one dark line) and 74 m/line pair for USAF 1951 and NBS 1951 targets, respectively, at 1 mm working distance (Fig. 3D and Supplementary Methods). Using this imaging module, we obtained bright-field images of entire structure as well as smaller tissue regions of the ex vivo rat lungs (Fig. 3E, Fig. S3) and rat liver (Fig. S3).
Prior to all stiffness measurements, the force sensor circuit was calibrated by measuring the output voltages against known forces applied to the sensor using a force gage (M5-2, Mark-10) (Fig. 4A). The circuit consisted of a digital data acquisition device (Arduino UNO, Rev 3), a reference resistor (R, Microchip), and custom-written MATLAB code (MATLAB R2021) The output voltage (VF) versus force load (F) curves were generated for all resistors (force range: 0-5N) (Fig. 4B). Different reference resistors (R; 1, 10, and 100 k) integrated into the sensor circuit to investigate the role of the resistance on the measurement outcomes (Figs. 4B and4C). The measurements showed that increasing the value of reference resistance increased the sensitivity of the force measurement at lower force range (below 0.5 N). Since lung and liver tissues are soft, the 100 k-resistor was used for the stiffness measurements (Fig. 4C).
Gelatin-based tissue phantoms were prepared by dissolving gelatin powder (G2500, Sigma-Aldrich) in 1 phosphatebuffered saline (1 PBS; Gibco) at 70C (concentration: 5, 10, and 15%) and physically crosslinking the solutions at 4C for 30 min. All tissue phantoms were left at room temperature for 30 minutes prior to stiffness measurements. To create tissue phantom with spatially heterogenous stiffness, a cylindrical Indocyanine Green (ICG)-labeled poly(dimethylsiloxane) (PDMS) (ratio of prepolymer to crosslinker: 10:1; diameter: 12 mm, thickness: 2.5 mm) was embedded at the center of a gelatin block (concentration: 10% w/v; length: 35 mm, width: 35 mm, thickness: 10 mm). The details of the preparation method are provided in the Supplementary Methods.
Lungs and livers were obtained from Sprague-Dawley rats (SAS SD rats; total: 6 rats; weight: 250-270 g; Charles River Laboratories). All animal care, handling, and experimental work were conducted under the animal protocol approved by the Stevens Institute of Technology Institutional Animal Care and Use Committee (IACUC). A complete description of the animal experiment procedure is provided in the Supplementary Methods. Briefly, animals were euthanized, and lungs and livers were isolated and rinsed with 1PBS prior to the stiffness measurements. Lungs were ventilated for 10 min with a tidal volume of 2.2 mL to eliminate the variations that could be caused by inflation and deflation history.
Further, freshly isolated swine organs, including lungs, heart, liver, and abdominal skin and muscle (ventral region), were purchased from a local slaughterhouse (Dealaman Enterprise, Warren, NJ). The stiffness measurements of all organs were conducted on the day of harvesting. To minimize the variability between the lungs resulting from different inflation and inhalation histories, the swine lungs were ventilated with a manual resuscitator (Ambu bag) connected to a pressure sensor for 10 min.
To produce an ex vivo lung cancer (tumor) model, an incision was made in the distal region of the swine lung using a blade, and an ICG-labelled PDMS (diameter: 2 cm; thickness: 5 mm) that mimics a nodule was inserted into the subpleural region, approximately 5 mm below the pleura surface. The incised area was then sutured (4-0 silk) to prevent air leak and lung collapse.
All experimental data were presented as mean ± standard deviation (SD). All experiments were performed at least five times (n > 5; n: number of samples) to ensure reproducibility and to evaluate statistical significance between experimental groups. One-way analysis of variance (One-way ANOVA) was performed, followed by Tukey's honest significant post hoc test to compare different groups. A difference was considered statistically significant when p < 0.05.
To evaluate the functionality and measurement accuracy of the palpation device, we evaluated the stiffness of gelatin-based tissue phantoms (Fig. 5). The phantoms were compressed using the measurement probe while the deformation was continuously monitored using a camera (Fig. 5A). To determine the elastic modulus, the sensor head was gently pushed against the tissue phantom while the applied force was being recorded continuously (Fig. 5B, region b). The tissue deformation continued until the tip of both contact electrodes touched the tissue surface due to maximum tissue deformation (i.e., 2 mm). The contact between tissue and electrodes was confirmed by detecting an electrical signal as the electrical circuit between the contact electrodes was closed (Fig. 5C, Fig. S4). When the maximum deformation was achieved, the voltage (VE) recorded during the measurement showed a step response of 3.2 volts (Fig. 5C,region c). The E values were then calculated based on the maximum deformation length (LC: 2 mm) and the corresponding force measured (FC) using an equation developed for spherical indenters (Fig. 5D and Supplementary Methods).
The rigid substrate made of an acrylic plastic (8473K112, McMaster, E: ~3.1 GPa), onto which tissue mimics or tissue samples were placed could influence the stiffness readouts in indentation-based measurements [38]. Accordingly, we evaluated the stiffness of gelatin tissue phantoms (concentration: 10% w/v) of different thicknesses (2.5, 5, 7.5, 10, 15, and 20 mm) to find the critical tissue thickness at which the effect of the substrate is negligible (Fig. 5E). The measured E values were 49.5 2.6, 42.1 2.8, 37.4 0.7, 25.7 2.3, 25.7 2.2, and 24.5 2.3 kPa for phantoms with the thickness of 2.5, 5, 7.5, 10, 15, and 20 mm, respectively. These results show that the effect of substrate on the measured stiffness decreased as the phantom thickness increased, without significant difference in stiffness for tissue phantoms thicker than 10 mm (Fig. 5E).
We further measured the stiffness of phantoms with different gelatin concentrations (5, 10, and 15% w/v; phantom thickness: 10 mm; Fig. 6, Fig. S5). The slope of the acquired forcedisplacement curves increased with the gelatin concentration (Fig. 6A). For 5, 10, and 15% w/v tissue phantoms, the measured E was 18.3 1.5, 25.7 2.2, and 42.5 1.6 kPa, respectively (Fig. 6B). The results were in the range of elastic moduli of physically-crosslinked gelatin phantoms measured via our published vacuum-based method [13].
Further, we used tissue phantoms to investigate whether the palpation device can profile tissues with spatially heterogenous stiffness, such as a tumor formed in tissue (Fig. 7). To this end, an ICG-labeled cylindrical-shaped PDMS, which recapitulated a stiff nodule, was embedded at the central region of a soft 10% gelatin block (Fig. 7A and Fig. S6). The location of the ICGlabeled PDMS nodule was confirmed visually via near-infrared (NIR) imaging (Fig. 7A and Supplementary Methods). Then, the stiffness of the tissue phantom was mapped using our palpation device (Figs. 7B and7C). Starting from the upper left corner of the phantom, the stiffness was measured by gently compressing the surface of the tissue mimic and recording the force and electrical signals. The stiffness map showing the elastic modulus distribution with a spatial resolution of 5 mm was generated (Fig. 7C). Notably, while the E values at peripheral areas of the phantom block varied between 23 and 27 kPa, there was a substantial increase in the E values (range: 38-50 kPa) at the center of the tissue phantom where the PDMSbased nodule was located, highlighting the ability of our device in accurate and quantitative localization of a stiff tissue mimic.
We investigated whether our palpation device can accurately measure the elastic modulus of human-sized swine lungs that were subjected to various internal air pressures (Fig. 8). The tissue stiffness was without ventilating the lungs to minimize motion-induced measurement errors. Specifically, the intra-peak inspiratory pressure (PIP) of the lungs were adjusted to a constant level (e.g., 2, 25, and 45 cmH2O) during each measurement using an Ambu bag and a pressure sensor integrated into the tracheal tubing (Fig. 8A). The measured E values were measured to be 9.1 2.3, 16.8 1.8, and 26.0 3.6 kPa when the internal pressure of the lungs (PIP) was subjected to 2, 25, and 45 cmH2O, respectively (Fig. 8B). Notably, the lung tissue stiffness increased with PIP due to changes in the tension (T) of the lung parenchymal tissue within the pleural network and alveolar septa with respect to the pressure inside the lungs [39]. The E values measured using our device were in the range reported in the literature [40,41].
We further created a lung tumor model that mimicked the presence of small nodules in the distal regions of the lungs to determine whether our palpation device can accurately discriminate the nodules from healthy tissue (Fig. 8C). The lung cancer model was generated using an explanted swine lung and an ICG-labelled PDMS block as a nodule mimic (size: 2 cm; thickness: 5 mm) (Fig. 8C). An incision was created in the right lobe, and the nodule was placed in the subpleural region, approximately 5 mm below the pleura surface (Fig. 8C, Fig. S7). We used the imaging module of the palpation device to visualize the location of the nodule in the subpleural region (Fig. 8C). The measurement probe was then moved to different regions of the lobe (distance between measured regions: 1 cm) to evaluate the stiffness across the lung tissue near the region where the nodule mimic was implanted. Following the scanning, a stiffness map showing the distribution of the elastic modulus across the entire lobe was generated (Fig. 8D). Notably, the E values ranged from 9.3 to 28.3 kPa, where the stiffness of 28.3 kPa corresponded to the precise location of the nodules in the lung lobe, confirming the ability of the device to locate the small nodule in the lung.
We further determined the tissue stiffness of rat liver and lung, and swine heart, liver, skin, and muscle (Fig. 9 and Fig. S8). The measured E values were 2.6 0.3 and 9.2 0.5 kPa for rat lung (peak inspiratory pressure: 2 cmH2O) and liver, respectively (Fig. 9A). The measured E values were 33.0 5.4, 19.2 2.2, 33.5 8.2, and 22.6 6.0 kPa, for swine heart, liver, abdominal skin, and muscle, respectively (Fig. 9B), and were similar to the values reported in the literature (Table S2) [42].
The adoption and utilization of robot-assisted surgery has continued to rise, and it is anticipated to persist as technology evolves and patient outcomes improve. Given this trend, there is a growing need to develop technologies to improve surgeons' ability to access tissue health and identify pathologic lesions, such as tumors, in real-time. Current technologies, such as preoperative localization with wires, markers, and dye are limited in their effectiveness and applicability [43,44]. Other intraoperative imaging modalities such as ultrasound, fluorescent imaging, photoacoustic, and Raman spectroscopy have failed to have a significant clinical impact on the accuracy of localizing small tumors, due to their limited spatial resolution [45]. Inaccurate localization and delineation of tumor margins may result in either excessive or sufficient removal of tissue during tumor resection, which may necessitate re-operation [46]. During robot-assisted procedures, the surgeon may evaluate the tissue stiffness by grasping it with surgical tools and visually inspecting its response to stretch and compression. This assessment, however, is subjective and may result in tissue damage and reduced blood flow due to the lack of tactile perceptions during tissue-surgical tools contact [47]. In contrast, quantitative profiling of tissue stiffness can allow surgeons to make objective and data-driven decisions during tissue resection procedures.
To improve the precision and accuracy in detection of pathologic tissues during RMIS, we created a robotic tissue palpation device that can quantify the stiffness of tissues (Fig. 1). Our palpation device was integrated with a sensing probe, steerable catheter, a motion control module, and an optical fiber-based imaging module. The tissue stiffness measurement is achieved by gently compressing soft tissue with a rigid hemispherical indenter to deform the tissue to a specified depth (LC). The magnitude of the force (FC) required to deform the local tissue to correlated with LC to estimate the tissue stiffness or elastic modulus (E). Significantly, this compression-based approach can allow intraoperative tissue stiffness assessment in real time without damaging the local tissue. Multi-directional movements, including translational, tilting, and deflection motion of the device, were enabled by simultaneously controlling servo motors (Fig. 2). The conformable and controllable device motions facilitate the intra-operative navigation and tissue compression within a tight space, such as chest cavity. In particular, the wire-driven design of the catheter can provide dexterity, allowing the probe to apply the normal force to tissue with irregular surface topology. The majority of palpation devices reported in the literature have a limited range of motion and flexibility [20,28,34,48,49], which makes them difficult to use during robotic surgery. In contrast, our device is capable of maneuvering in confined spaces, allowing surgeons to access to the surgical targets and survey questionable tissue rapidly [50,51]. We envision our device will be able to integrate into a standard robotic or laparoscopic device arm and be controlled by the surgeon with existing interfaces. Each scenario requires thorough considerations, such as the adaptability to clinical requirements, interoperability with other robotic systems and surgical instruments, and optimization of the device's size. Additionally, In future designs, the size of the current catheter (diameter: 20 mm) can be minimized to fit into the surgical ports used in minimally invasive and robotic surgery (diameter: 5 to 12 mm) [52].
The optical fiber imaging probe incorporated into the device allows visual monitoring of the local tissue during stiffness measurement. Notably, the imaging module can be customized to enable visualization at the cellular level, as demonstrated in our previous study [53]. By implementing a real-time image processing scheme, such as Gaussian filtering, the quality of images and videos can be substantially improved (Fig. 3, Fig. S9). Moreover, the flexibility of the imaging fiber facilitates integration into the steerable catheter. Further, the bifurcated geometry of the imaging fiber enables simultaneous illumination and imaging, with the light passing through the "transmitting bundle" to the fiber tip and the fluorescent signal passing through the "receiving bundle" into the camera. This imaging capability can be useful during intra-operative tumor resection, where surgeons can administer fluorescent molecules that can specifically label tumors to improve the accuracy of tumor identification and resection [54].
To obtain precise tissue stiffness measurements with our palpation device, it is essential to ensure that the device's distal end is positioned perpendicular to the tissue surface, allowing the generation of a normal force during the measurement. Deviation from this perpendicular alignment can result in shear forces applied to the tissue due to an angled device position, which can introduce inaccuracies in the readings. The impact of these shear forces can be particularly pronounced when dealing with tissues that have slippery surfaces, especially if the slipperiness is not consistent across the tissue surface.
Orthogonal positioning of the device with respect to the tissue surface can be achieved by ensuring that the two contact electrodes simultaneously touch the tissue surface. Minor deviations (e.g., a few angles) from perfect orthogonality may lead to negligible impact on the accuracy of the measurements. In future design improvements, we can enhance accuracy by adding more contact electrodes, for example, four electrodes, to the device. This increased electrode count can help ensure perpendicular tissue compression, leading to greater precision in device positioning and accurate measurements [26].
Utilizing imaging guidance can greatly assist in achieving proper orthogonal positioning of the device during the procedure. In our experiments, we ensured the probe's correct alignment with the tissue surface by monitoring its position using a camera placed adjacent to the measurement site. Likewise, in robotic surgery, incorporating a surgical camera can aid the surgeon in accurately positioning the palpation device for stiffness measurements in real-time. Furthermore, advanced imaging methods, 3D topography in particular, can be employed to assess the tissue surface's structural features, facilitating precise alignment of the device to achieve the desired orientation [55][56][57].
The innovative feature of the palpation device developed in the current study is the integrated contact electrodes circuit, which informs the maximum tissue deformation (Fig. 5). In our design, the deformation of tissue to a specified degree can be accurately determined through non-invasive, real-time recording of voltage via the electrodes. The current deformation measurement methods, such as optically based and computerassisted approaches, are time-consuming and prone to errors [58]. On the other hand, a predetermined magnitude of tissue deformation can be achieved intraoperatively using our approach. Further, the deformation length can be easily customized by using a hemispheric indenter with different heights. The magnitudes of electrical voltage and current (voltage: 3.2 volts, current: 0.5 mA) used in this study, which could also be easily adjusted to different values, were within a safe and clinically relevant range (Figs. 456789) [11,59].
Despite the several advantages of the tissue palpation prototype presented in this work, some areas need to be improved. For example, the resolution of the stiffness map can be improved by applying finer discretization or using a smaller indenter. Moreover, a cylindrical nodule mimic with a constant stiffness value was used in the lung cancer model, and stiffness measurements were performed on ex vivo tissues. Given that tumors have different topologies and mechanical properties at various developmental stages, further studies can confirm the utility of our tissue palpation device in various in vivo scenarios [23]. Furthermore, the tissue phantoms and biological tissues in this study were assumed to be elastic within the applied deformation ranges. Nonlinear viscoelasticity of the tissues and its effect on stiffness measurements and localization of pathologic tissues can be studied more thoroughly to provide clear picture of the structural complexity of biological tissues [22].
Overall, the study highlights the potential of a novel robotic tissue palpation device as an intraoperative tool that can assess and localize pathologic tissues with altered mechanical properties, such as tumors. The quantitative data provided by the palpation device can be interpreted along with the results of other pre-and intra-operative techniques to give clearer insight of tumors location and margins to surgeons. Exciting prospects for the future of this device include even more sensitive and precise measurements through design improvements, validation of its functions through in vivo animal experiments, and the utilization of this innovative technology in real-world clinical settings. With these developments, robotic palpation technology may have a profound impact on tissue assessment and surgical procedures in the future.
This article has been accepted for publication in IEEE Transactions on Biomedical Engineering. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/TBME.2024.3357293 © 2024 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See https://www.ieee.org/publications/rights/index.html for more information. Authorized licensed use limited to: Stevens Institute of Technology. Downloaded on April 27,2024 at 00:52:44 UTC from IEEE Xplore. Restrictions apply.