Treatment methods for spinal disorders involving spinal fusion surgery often entail lengthy and expensive procedures, with a risk of damage to surrounding tissues. Breach detection tools aid in precise pedicle screw placement, thereby reducing the risk of complications or misalignment, and minimizing the need for follow-up surgeries. Accurate placement of pedicles in the vertebrae is crucial in spinal fusion surgery, as improper positioning can result in breaches that potentially harm neural tissues. This study proposes a handheld bone sensing device capable of measuring real-time reflected light from tissue using carefully selected optical elements. The selection of these optical elements was guided by analyzing porcine absorption spectra and conducting Monte Carlo light reflectance simulations. The device is designed to provide consistent real-time feedback by incorporating an electronic circuit optimized to minimize environmental signal interference. A prototype, named the optics-based bone quality sensing device, was developed using electrical discharge machining parts, 3D printed parts, an Arduino Uno, a custom-made printed circuit board (PCB), and off-the-shelf optical components such as laser and photodiodes coupled into optical fibers. Experimental validation with ex vivo porcine vertebrae revealed that the device could not detect sufficient reflected light to differentiate between cortical and cancellous bone. Despite this, the study marks significant progress in creating a handheld bone sensing device that utilizes real-time reflected light sensing minimizing interference. This device represents an important step towards developing tools to assist in precise pedicle screw placement for the treatment of spinal disorders.

Accurate placement of screws in spinal surgery is vital to mitigate risks like instability, nerve damage, or spinal cord injury, which can lead to various complications such as pain, sensory loss, muscle weakness, or paralysis below the injury site [33]. However, success rates in screw placement vary widely, ranging from 27.6% to 100% [56].

Advanced techniques, like pedicle screw placement (PSP) with enhanced guidance systems, aim to provide surgeons with better spatial awareness to prevent cortical breaches. Various guidance methods,
such as fluoroscopy, computer-assisted techniques, neurophysiological monitoring, and Electrical Impedance Spectroscopy (EIS) enabled by the PediGuard, have been developed and utilized in spinal fusion surgeries [18]. Despite their efficacy in reducing complications, these systems have limitations such as increased radiation exposure, prolonged surgery time, or lack of real-time feedback.

This study compares two sensing techniques, Electrical Impedance Spectroscopy (EIS)) and Diffuse Reflection Spectroscopy (DRS), for spinal fusion surgery. Both methods are employed to detect cortical
breaches during PSP by distinguishing between cancellous and cortical bone. Before experimentation, certain gaps in understanding must be addressed, including exploring the working principle of EIS for
vertebrae tissue and developing classifiers for both single and multiple frequencies. Additionally, efforts are made to create a vertebrae phantom with realistic electric and optical properties. EIS demonstrates an increase in impedance for cortical bone compared to cancellous bone due to the latter’s weblike structure with highly vascularized yellow bone marrow. Single-frequency analysis outperforms multiple-frequency analysis in distinguishing between vertebrae tissues. However, the development of a better model or the use of multiple single-frequency classifiers may enhance predictive
accuracy. Despite attempts to create an accurate vertebrae phantom, limitations in adjusting dielectric properties result in the utilization of the best-fit recipe.

Comparison between EIS and DRS reveals both are effective in detecting cortical breaches and tissues, with DRS exhibiting a longer look-ahead distance. EIS operates more like a step function, making it susceptible to distortions, suggesting DRS outperforms EIS in spinal fusion surgeries.

As people’s life expectancy increases, more individuals will develop degenerative diseases over their lifetime. Surgeries like spinal fusion can help manage these diseases to improve mobility and alleviate pain in the aging population. During spinal fusion, two or more adjacent vertebrae are connected to stabilize part of the spine. Since it takes several months for the bones to grow together, connecting rods and pedicle screws provide support in the meantime. Inserting these screws into the vertebrae is a delicate procedure that, if not done correctly, could harm nearby nerves and blood vessels. To find the correct screw trajectory, surgeons rely on tactile feedback and their experience, as they cannot see inside the body. To support surgeons and enhance safety, the procedure can be guided by technology. Although surgical navigation systems exist, their high cost is a barrier to widespread use. A simple and affordable alternative is tissue sensing technology, which provides insights into the structure and composition of the tissue being examined. This information helps detect the border of the vertebra during spinal fusion to prevent breaches into surrounding tissue. Even where surgical navigation is available, tissue sensing offers an extra layer of safety, for instance, if the spine has moved during surgery....@en

Purpose: In percutaneous renal ablation (PRA), reaching the target accurately is challenging due to target motion and limited real-time visualization of the target, requiring a suitable navigation strategy. In the context of this thesis research, skin markers are used in combination with an the Azurion C-arm system (Philips Healthcare, Best, The Netherlands), an image-guided therapy system, equipped with ClarifEye (Philips Healthcare, Best, The Netherlands), an Augmented Reality (AR) surgical navigation system. The skin markers are tracked by ClarifEye, virtually representing the position of the patient. This thesis aims to explore if skin markers can facilitate sufficiently accurate navigation in PRA by investigating the skin marker motion (SMM) and kidney tumor marker motion (TMM) during needle insertion, considering the insertion speed (IS) and level of perirenal fat (LPF). Methods: An ex-vivo porcine kidney model, with simulated skin, tumors and perirenal fat was developed and validated. A skin marker model was constructed on the model. Consequently, 60 robotic-assisted needle insertions were planned with ClarifEye (Philips Healthcare, Best, The Netherlands) navigation and performed on 8 kidney models (4 with low LPF (LLPF), 4 with high LPF (HLPF)), targeting 20 tumors at varying ISs: continuous insertions of 2.5 mm/s, 75 mm/s and consecutive 5 mm increments of 10 mm/s. Cone beam computed tomography (CBCT)-scans were made before and after insertion to analyze SMM and TMM. Results: The SMM (overall mean SMM = 0.16 mm, standard deviation (SD) = 0.14 mm) was smaller than the voxel-size (1D) of dViewX (dicom-viewer software, Philips Healthcare, Best, The Netherlands) (voxel-size 1D = 0.2425 mm), regardless of the direction of the SMM and regardless of the IS and LPF. The TMM values in the z-direction were predominantly negative. The IS and LPF did not impact the amount and direction of TMM.
Conclusion: Any local skin deformation caused by needle insertion resulted in negligible SMM. Although the skin markers do not account for the TMM caused by needle insertions, the median TMM values (1.2-1.9 mm) are still smaller than the minimally required clinical insertion accuracy of 3-5 mm. The acceptable TMM values, in combination with the negligible SMM values, imply that skin markers can be used to facilitate navigation in PRA.

Summary
Introduction
The increasing use of cardiac catheterisation procedures has raised concerns about occupational radiation exposure and its associated health risks for clinicians. Radiation exposure can lead to deterministic and stochastic effects, including cataract and various cancers. To minimise these risks, the ALARA (As Low As Reasonably Achievable) principle must be followed in the clinical setting by applying the available measures. Radiation protection consists of passive and active measures. Passive protection includes architectural shielding, stationary shielding, and personal protective equipment, while active protection focuses on minimising radiation during interventions, employing adjustable lead screens, modifying imaging techniques and provide feedback to improve the usage of these measures. However, current methods for measuring and providing feedback on radiation exposure are insufficient.
The primary goal of this research is to find a comprehensive and improved approach to guidance and implementation of protective measures, to ensure the safety and well-being of healthcare workers who face radiation exposure in their daily practice. An integral component of achieving this goal is to create a model that calculates radiation exposure for clinicians in the cathlab with the use of a lead screen. This research aims to determine the feasibility of using dose rate data from measurements without a lead screen to estimate the dose rate in an environment where a lead screen is employed. Moreover, potential methods for communicating dose rate data to clinicians during a procedure will be explored, aiming to increase their awareness of radiation exposure and encourage the adoption of appropriate protective measures without distracting them from the procedure itself.

Method
Data collection is necessary to construct a model that can predict the radiation scattering. Measurements in the catheterisation laboratory were performed by placing Philips DoseAware Detectors (PDDs) at various locations and heights. Combinations of fluoroscopy or acquisition mode, anteroposterior (AP) view or a 40° tilted view of the C- arm, and the presence or absence of a lead screen, resulted in eight different setups, with each 168 measurements.
In the programming process careful consideration has been given to the ease of refining and improving the model afterwards. Three potential source options for the radiation were considered: the centre of the phantom, a point where 50% of the scattering has occurred, and a 3D representation of the entire phantom. The modelled lead screen together with the source was used to create an attenuation grid, which will be used for the estimation of the effect of the lead screen.
The adequacy of the estimated dose rates with the lead screen was evaluated using boxplots and counting the number of correctly estimated points. Then, the data was interpolated using Kriging. The interpolated data was visualised, and the radiation doses for the assistant and cardiologist were calculated at the chest and head levels.
The exposure calculation was performed for three lead screen positions and a scenario without the lead screen. The placements considered were close to the clinician, close to the phantom, and at the edge of the phantom.
To explore radiation feedback preferences, a literature review and a questionnaire for interventional cardiologists were conducted.

Results
Measurements were taken for eight different setups, each with 168 measurements, including the combinations of fluoroscopy or acquisition mode, anteroposterior (AP) view or a 40° tilted view of the C-arm, and the presence or absence of a lead screen. The interpolated data revealed that the dose rate is higher when the acquisition mode is used, and radiation levels are higher when the C-arm is rotated. Three different source types were incorporated into the measurements, and the accuracy of the estimations was evaluated based on whether they were within a deviation of 0.1 mSv/h from the actual measurements. The 3D source was found to be the most accurate representation of the source in the model. Visualisation of inaccurately estimated values showed that the model overestimated the amount of radiation blocked in certain cases. The radiation exposure for clinicians increased as the lead shield was positioned closer to the radiation source.
A questionnaire was conducted with five interventional cardiologists to investigate their radiation safety practices. The results indicated that radiation safety is deemed important, with an average score of 8/10. The lead screen is an essential shielding component, but its placement varies, and there is uncertainty regarding its optimal positioning. Monthly reports on total radiation dose exposure were provided. The clinicians hardly paid attention to this information, as they believed that any potential harm had already occurred.
The highest-rated techniques were advice on optimal lead screen position (displayed on the lead screen) and a visualisation using augmented reality that displays the scatter pattern on the monitor.

Discussion and Conclusion
This study addresses radiation safety in interventional cardiology by exploring protective measures. A preliminary model was developed to predict radiation exposure based on the location of the lead screen enabling the computation of the radiation dose received by clinicians during a procedure. This dose rate model helps to visualise the impact of lead screen positioning on radiation exposure, which can aid in real-time decision-making during procedures. The model also revealed that placing the lead screen closer to the clinician, rather than closer to the source, could potentially enhance radiation attenuation by a factor of 6.2. This results in a possible dose reduction up to 32.5 times.
The most preferred option by the interventional cardiologists was to display the suggestion of the optimal position of the lead screen directly on the shield. A potential addition to the feedback system could be an augmented reality (AR) visualisation of the radiation effect on the monitor, clearly showing safe areas for personnel. The precise feedback mechanism should be piloted and refined to ensure seamless integration into clinical practice without distracting the clinician from the procedure itself.
Understanding the limitations of the measurements and the model is crucial for optimising its application in practice and the improve the reliability. The accuracy of measurements obtained with the DoseAware badges, while sufficient for this project, could be improved by using more accurate instruments. The phantom used in the study does not accurately represent real patients, and future research should involve patient- specific inputs and more realistic phantoms. One of the key future objectives include collecting more data in different settings and positions of the C-arm, as well as investigating different positions and orientations of the ceiling-mounted lead screen.
The model itself has limitations in its representation of the scattering source and interpolation methods. To improve the model, researchers should account for other sources of scatter, and assess alternative interpolation techniques using more powerful computing resources. Moreover, to calculate the received radiation dose, the model should consider the entire body, accounting for the protection provided by a lead apron and the concept of effective radiation dose.
Accomplishing these objectives will allow the model to play an important role in enhancing cathlab radiation safety, benefiting the health and safety of clinicians.

Introduction Sudden cardiac death is one of the major deaths worldwide. Several treatments are available to lower risks of sudden cardiac deaths. Atrial fibrillation is one of the leading causes for sudden cardiac deaths. Atrial fibrillation can be treated with drugs or with procedures or surgeries. Catheter ablation is one of these procedures. With the use of radiofrequency, the area in the heart causing the disrupting of the signal pathway and therefore causing the atrial fibrillation can be restored. When performing the ablation, the force on the heart tissue and insert angle of the catheter tip can influence the results of the catheter ablation. In this paper we investigate whether the impedance readout can be used to determine the force and the angle the multielectrode catheter tip is exerted on a phantom tissue. Methods To research the impedance readout with respect to the insert angle and force an experimental setup was designed. The electrodes of a catheter were connected to a HP impedance analyzer (HP 4192A). This catheter was guided through a 3D print with hollow tubes (90, 60, 45 and 30 degrees) to achieve the desired insert angle on a phantom. The 3D print is mounted to a linear stage to control the movement of the catheter with millimeter precision movement. The catheter was pressed onto a phantom. Cardiac phantoms were made with the base material Gellan Gum, Agar and PVA. The phantom was put in a container with a skinpatch counter electrode attached to it to be able to measure the impedance change. Below the container a scale is placed to readout the force that is being exerted on the phantom. Results Samples with base materials as Gellan Gum, Agar and PVA were made and the catheter was being exerted on it with various insert angles. Impedance change with increasing force and varying insert angles were recorded and graphs were taken out of the data that is being recorded with the use of Excel. The data has been taken from fresh made samples, 1 week old samples and 2 weeks old samples. From the graphs the relative impedance increase decreases when the insert angle is reduced from 90 to 30 degrees. Conclusion Impedance can be used to determine how much pressure is being exerted on a heart tissue. However, the catheter that approaches the heart tissue will have different insert angles. The impedance change is dependent on the various insert angles. Therefore it is not sure if this impedance change method can be used to determine the force that is being exerted on the heart tissue in a reliable way when the catheter tip angle is unknown.

Spinal fusion is a surgical treatment involving stiffening parts of the spinal column to eliminate any relative motion and improve spine stability. A critical part of spinal fusion surgery is the placement of pedicle screws because it is highly complex and can damage the vascular structures and nerves near the vertebrae. To prevent complications, a new anchoring method is developed that strengthens screw contact by following the curved trajectory of the dense cortical shell while avoiding a cortical breach. Diffuse reflectance spectroscopy (DRS) is a technique that could be used in a bone drill to obtain real-time information about the location of the drill within the vertebra. In order to drill the correct trajectory, an optical probe using DRS should determine the parallel closeness of cortical bone to the drill tip. To show the potential of this novel technology for improving spinal fusion surgery, this master thesis introduces and validates a flexible sideways-looking probe using DRS to identify the proximity of cortical bone when measuring along the cortical shell of a vertebra.

In this study, the ideal optical design to detect the proximity of cortical bone is investigated by looking into the effect of fiber angulation on the photon propagation using Monte Carlo simulations and phantom experiments. A conceptual design was created based on the ideal optical design combined with the mechanical requirements to make the probe insertable into a vertebra. As a proof of concept, a prototype was developed and tested on a bone-mimicking phantom and porcine vertebrae.

The conceptual design comprises a flexible needle with two optical fibers attached to a rigid tip with a diameter of 2.9 mm. Two mirrors are positioned in the rigid tip, emitting light 45\degree off-axis and collecting light 90\degree off-axis with a source-detector separation of 2.1 mm. These light directions were chosen because they optimize the reflectance spectrum difference between cancellous and cortical bone due to a longer path length, making detecting cortical bone closeness easier. A prototype was made that demonstrates that the probe can detect the parallel proximity of cortical bone at a depth up to 1.5 mm in an ideal phantom setting. The prototype can distinguish between pure cancellous and cortical bone in porcine vertebrae. Additionally, the prototype can also be inserted into the pedicles and bent along the cortical wall of a vertebra. These positive findings show that the flexible sideways-looking probe created in this study can potentially improve spinal fusion surgery by enabling an increase in cortical bone contact without perforating it. Although proof of concept has been provided, it is essential to improve the design for safety and usability and improve tissue type identification accuracy before continuing to clinical trials.

In the Netherlands, 362.000 people suffered from atrial fibrillation in 2019. During a catheter intervention, the tissue that causes atrial fibrillation is destroyed by heat generated by an alternating current exerted from an ablation catheter. Two problems of this treatment are over-ablation of the heart tissue and the non-contingent surrounding of scar tissue around the pulmonary veins. Better surgical outcomes can be obtained if an extra imaging technique is added to prevent these problems from occurring. Impedance spectroscopy has potential, it is already widely used for measurements and food characterization and it enables the measurement of the distinct dielectric properties and impedance of different tissue types. Based on the dielectric properties and impedance, different tissue types can be distinguished from each other and the surgeon can adapt his procedure according to this information.

The goal of this Master Thesis is to distinguish different tissue types in the radiofrequency range with impedance spectroscopy during a cardiac ablation procedure. Impedance data for this Master Thesis are gathered via impedance measurements on three porcine hearts with two-needle electrodes over a frequency range from 100 Hz to 1 MHz. Ablated lesions are created with an electrosurgical knife in fulguration and pure cut mode. To distinguish between different tissue types with impedance spectroscopy there is a need for a parametric model that can parameterize the impedance spectra. This enables the comparison of different tissue types based on a couple of parameters instead of a complete spectrum. A parametric model is fitted to the measured impedance spectra and the resulting best-fit parameters are evaluated. The parametric model that is used is a series combination of a CPE model and a two-pole Cole impedance model. The CPE model models the electrode polarization contribution to the impedance spectrum. This way no a-priory knowledge regarding the electrode polarization is needed. The two-pole Cole impedance model models the tissue contribution to the impedance spectrum.

In the measurements that are performed in this Master Thesis, an increase in impedance is seen after ablation. The differences between the healthy and ablated measurements are mostly described by (significant) differences in the parameters of the CPE model (K and m) and the second Cole term R2, t2 and a2). The mean values of R2 and a2 increased after ablation, while the values of t2, K and m decreased after ablation for most measurements. Classification of healthy and ablated tissue is performed based on a formula consisting of a combination of these model parameters. The sensitivity and specificity are both around the 80%. This classification serves as a first impression, with the results indicating that discrimination of healthy from ablated tissue is possible with parametric modeling.

The growing demand for cardiovascular treatments and the need to control the ever rising health care expenditures require efficient use of expensive resources, such as cardiac catheterization laboratories (cath labs). These in cardiac catheterization specialized operating rooms are labor and capital-intensive. Therefore, a high utilization is desired, although overtime should be prevented. However, this is complex in practice, because uncertainty in procedure duration and the number of patients makes it difficult to predict shift duration. At the moment, scheduling is mainly done by hand and the result strongly depends on the scheduler’s experience. Furthermore, little research is done on cath lab scheduling. Hence, this study aims to further define the cath lab scheduling process, identify the areas of improvement, and explore the usefulness of Monte Carlo simulation to support human schedulers. Schedulers of two Dutch hospitals, the Reinier de Graaf Gasthuis and the HagaZiekenhuis, are interviewed to map the scheduling process and identify the main difficulties. According to the schedulers, the main area of improvement is the estimation of procedure and shift duration, which currently is based on experience and rules of thumb. Based on the process in the HagaZiekenhuis, a Monte Carlo simulation is developed that computes the shift duration, utilization, and number of deferrals based on the blueprint schedule, the distributions of procedure duration and the emergency arrival rate. A first attempt to validate the model is done with input data from the VUmc as reported in another paper, as no historical data of the HagaZiekenhuis was available. The number of simulations required for convergence of the model was found to be at least 300. Next, the utilization, overtime, and undertime of the simulation were compared with these metrics from the VUmc. They were found to be close, even though the scheduling decisions are based on another hospital. To examine the sensitivity of the model, several scenarios are tested, namely, no emergency arrivals, a double amount of emergency arrivals, both halving and doubling the standard deviation of input distributions, adding extra patients to the inpatient waiting list to increase the demand, changing the threshold when patients are deferred and changing the type of input distribution from lognormal to normal. The effect of these scenarios is mainly as expected. Less variation in the output can be obtained by reducing the variance of the input distributions. Additionally, the type of input distribution seems to have limited effect on the output distribution. Furthermore, the effect of changing the deferral threshold, which is stochastic according to the interviews, seems limited. Moreover, no inexplicable behaviour was discovered by inspecting the realizations of the blue print schedule. Lastly, the data of the VUmc was applied to the blueprint of the HagaZiekenhuis, but inspection of the blueprint schedule revealed that the procedure duration in the VUmc data is longer than in the HagaZiekenhuis. This demonstrates that one cannot use the data and blueprint of different hospitals. An attempt was done to correct for this by scaling the distributions such that the mean corresponds to the scheduled duration. This led to more realistic results for some procedure types, although the standard procedure duration appears to not match the average duration for other types. The first results are promising, hence, a more in-depth validation of the model is recommended. The most promising application of this simulation seems the estimation of shift duration given a specific set of patients, integrated in the scheduling software.

Breast cancer is the most prevalent form of cancer among women, leading to 2.3 million new diagnosis and almost 685,000 deaths in 2020. Currently, one of the preferred treatments is breast conserving surgery (BCS), but its success depends on the correct identification of tumor margins during surgery. A promising optical intraoperative margin assessment technique is diffuse reflectance spectroscopy (DRS). However, the DRS instrumentation is complex, since it uses dedicated broadrange lights and sensors. In addition, the system is expensive and voluminous. Thus, a simpler alternative is desired. A possible approach is to probe the tissue’s optical response at only a few relevant wavelengths.

This thesis, developed within an ongoing collaboration between TU Delft and Philips, represents a first proof-of-concept study towards the development of a multi-wavelength hand-held probe for real-time tumor detection during BCS.

A multi-wavelength sensor was developed for testing based on an existing prototype, able to measure with an infrared photodetector the optical response from two independent light emitting diodes.

The working principle of the detection system was based on the known differences in fat/water-ratio between healthy and tumor tissue. It was investigated whether the optical response of tissue-mimicking phantoms with different fat and water content, measured with infrared LEDs in a transmittance-based configuration, could provide a means to quantify the composition of the phantoms. A measurement method was proposed based on the linear relationship observed between the percentage of fat in the phantoms and the ratio of transmitted light intensity measured at 970 nm or 1300 nm, and 1200 nm. The test was repeated with new phantoms, and it was found that the parameters of the fat-intensity relationship varied between the experiments. The scattering parameter was identified as the main cause of this variation.

To demonstrate the feasibility of integrating the system in a hand-held device, the sensor was tested with optical fibers in a reflectance-based configuration. Infrared lasers were found to be suitable light sources to detect differences in reflected light intensity depending on the probing wavelength and phantom composition. However, the wavelengths of the available lasers were not significant for tissue discrimination, so the working principle of the measurement system was not tested in this configuration.

Overall, the results demonstrated that it is possible to use multi-wavelength optical measurements to detect differences between healthy and tumor tissue based on the fat and water concentrations. Further investigation is required to determine a robust method which allows to exactly quantify the tissue content based on the measured reflected intensity. Furthermore, following miniaturization and optimization, the sensor is suitable for integration in a hand-held surgical instrument. Despite the limitations, this study reveals the potential of multi-wavelength optical devices for tumor detection in breast, and encourages further research for their integration in the surgical practice.

BACKGROUND: Ovarian cancer is the number eight cause of cancer-related mortality in women around the world and has an incidence of more than 295 000 and mortality of almost 158 000. The most important prognostic factor for the survival of advanced-stage ovarian cancer is the complete resection of all microscopic tumour tissue during cytoreductive surgery. Still, even after complete CRS and removal of all visible and palpable tumours, more than two-thirds of women with advanced-stage ovarian cancer experience recurrence due to microscopic residual tumour. Therefore, an intraoperative visualisation technique is of great importance to detect these microscopic tumours. In this pilot study, the feasibility of NIR hyperspectral imaging for the detection of malignant ovarian cancer metastases was evaluated ex vivo. METHOD: Patients with suspected primary ovarian cancer planned for cytoreductive surgery were enrolled in the study. Hyperspectral images from the resected tissue were acquired in the wavelength range of 665-975 nm. All hyperspectral data was preprocessed by image calibration, min-max normalisation and noise filtering. The the most important features were selected before implementing the hyperspectral data in the classifier. The linear SVM, RBF SVM, and a k-NN classifier were used to discriminate malignant ovarian tissue from the surrounding tissue. The performance of the classifier was evaluated by leave-one-out cross-validation. RESULTS: Ten patients were included in the study. The linear SVM classifiers had the highest performance with a sensitivity of 0.81, a specificity of 0.75, AUC of 0.83 and MCC of 0.41. Thereafter came the RBF SVM classifier followed by the k-NN classifier. CONCLUSION: This pilot study shows that hyperspectral imaging with the use of a linear SVM classifier is a promising technique to discriminate malignant ovarian tissue from the surrounding tissue. Hyperspectral imaging can scan a whole area, is fast, non-contact, non-invasive and can be used inside the operation room. However, before HSI can be implemented in real practice, further validation is required in vivo and technical enhancements need to be made such as enlarging the training data set, evaluate the SWIR wavelength range, include a registration algorithm which accounts for elastic deformation of the tissue, evaluate several feature selection methods, and compare other classifiers.

Background Inter- and intra patient morphological differences and absence of visual feedback make the placement of pedicle screws technically very demanding. As the spine is surrounded by neurovascular tissue, complications can have severe consequences, including paralysis. Guidance techniques are often very expensive and can overexpose surgeons to radiation, while there is a high variability in reported accuracy rates. Diffuse reflectance spectroscopy (DRS) is a technique that allows for real-time tissue discrimination using an optical fiber which can e.g. be placed in a K-wire. It can detect difference in fat fraction between cortical and cancellous bone. Purpose This master thesis aims at investigating the capability of the DRS equipped K-wire to anticipate cortical breach by measuring a decrease in fat fraction, while being oscillated with a power hammering tool. Mechanical experiments investigate the amount of control over penetration depth during powered hammering insertion of a K-wire. Methods This thesis contains three experiments. Two focus on mechanical properties and one on optical properties. In the mechanical experiments, a 1.6 X 60 mm trocar tip K-wire without integrated optical fibers is used. In the first experiment, the K-wire and its oscillator are attached to a carriage on a horizontal guide rail. A mechanical bone phantom is placed at the end of the rail and is penetrated at varying oscillation frequencies (0 – 50 Hz). To move the carriage towards the bone phantom, masses are attached in discrete steps of 250 grams. Penetration depth is measured as insertion mass increases at constant oscillation frequency. In the second mechanical experiment, the same K-wire and K-wire oscillator are attached to a vertical linear stage. The K-wire penetrates the mechanical bone phantom at an oscillation frequency of 0 Hz or 35 Hz. Feed rates are varied (0.5 – 2.5 mm/s) and force on the bone phantom is continuously measured. For the optical experiment, a 1.6 X 300 mm fiber-optic K-wire is used with fiber distance equalling 1.22 mm. The K-wire with the oscillator are attached to a linear stage, and lowered at constant feed rate (0.5 mm/s) into an optical bone phantom. NIR spectra are acquired with a sampling frequency of 1.6 Hz. The effect of varying oscillation frequency (0 – 50 Hz) on measured fat fractions is quantified. Results The first mechanical experiment showed that for every oscillation frequency, penetration started at the first weight increment (0.1 kg). Without oscillations, phantom penetration initiated at a mass of 1.85 Kg. Regression analysis proved the penetration depth depends on the square root of applied axial force. In the mass range of 0 – 1.6 kg, 45 Hz led to the largest penetration depth (25 mm), with a relatively narrow 95% confidence interval (17 – 33 mm). The second mechanical experiment showed that measured mean force is lower when oscillating the K-wire (1.4 – 5.8 N vs. 9.5 – 12.3 N, p<0.0001). Effect of feed rate on measured mean force is found not significant (p=0.15 for oscillating and p=0.61 for not oscillating). In the optical test, it was found that look ahead distance (LAD) increases with increasing oscillation frequency and amplitude. 50 Hz was the only frequency where the LAD was larger than its oscillation amplitude (1.22 mm and 0.5 mm, respectively). Conclusion Powered hammering of a K-wire requires a lower axial force for penetration and therefore allows for more control over penetration depth. From the tested range, this effect is most significant for oscillations with high frequencies, adequate forces and low amplitudes. It is possible to measure a decrease in fat fraction with a fiber-optic k-wire during penetration. High oscillation frequencies with low amplitudes allow for earliest anticipation. An adequate sized sampling frequency is required to allow for breach anticipation.

Diagnostic imaging is a fundamental requirement for the effective treatment of about 25 % of patients globally. Ultrasound imaging is considered the safest, least expensive and most convenient diagnostic imaging modality. Its widespread adoption has been in the area of 2D imaging. However, the accuracy of diagnosis with 2D ultrasound imaging is highly dependent on the experience and expertise of the sonographer because the sonographer has to mentally create a 3D impression from multiple 2D images of the region of interest and this could lead to erroneous diagnosis. 3D ultrasound imaging addresses this concern, however, it drives up the cost of ultrasound imaging. Also, disadvantages such as difficulty in taking quality images, prolonged time in learning how to effectively take these images and human distress during the acquisition of these images have been reported. By leveraging the high level of precision, accuracy and maneuverability provided by robotic systems in conjunction with low-end ultrasound imaging platforms, 3D ultrasound images can be captured, the problem of human distress alleviated and subsequently reducing the cost of 3D ultrasound imaging. This project sought to explore the feasibility of designing a low-cost 3D ultrasound robotic system. The design followed a distributed system approach using ROS (Robotic Operating System) where the data acquisition decoupled from the processing and visualization unit. The performance of the design was tested by constructing 3D ultrasound images of custom made phantoms. These results were compared with 3D ultrasound images of the phantoms acquired using the Philips EPIQ 7 - a high-tier ultrasound machine. The results obtained with this design had a low resolution in comparison with those obtained with the Philips EPIQ 7 however, a 3D point cloud representation of the object embedded in the phantom can be seen. Also, a full 3D image of the phantom could not be acquired due the transducer movement limitations of the design. This design successfully demonstrates the feasibility of integrating low-end electronics with robotic systems to acquire 3D ultrasound images.

Introduction. In spinal surgery, the misplacement of spinal screws is a problem that causes (severe) pain, bleedings or even paralysis [2]. Screw misplacements are common as navigating in the spine is difficult due to the small vertebral dimensions and a lack of anatomical landmarks [3] [4]. In order to improve the navigational support of spine surgeons, this research focuses on the development of an optical sensing diffuse reflectance spectroscopy (DRS) surgical drill that identifies bone tissue boundaries. The developed drill concept introduces a stagnant optical fiber-equipped probe into a cannulated orthopaedic drill. To verify the clinical applicability of the developed system, the accuracy of the optical tissue boundary detection has been analysed under different tissue penetration speeds, as well as the axial drilling force increases due to the introduction of a stagnant probe into a drill.
Results. When increasing the drilling feed rate in the optical phantom, the drill overshoot (the difference between the DRS-derived tissue boundary location identification and the actual location of the phantom boundary) shows a larger spread. The maximum feed rate at which no overshoot takes place is 0,5mm/s. Increasing the sampling frequency –especially decreasing the inactive period between the measurements– can improve this.
None of the K-wire equipped drills can penetrate the used Sawbones® cortical bone phantom. The axial peak feed forces occurring in the Sawbones® cancellous bone phantom while using a regular 2,7mmØ orthopaedic drill is 38,2N. When using the 2,7mmØ orthopaedic drill with a 1,6mmØ K-wire, a peak of 57,6N is observed. Because the data from drilling in the Sawbones® cancellous bone is not normally distributed, a benchmark experiment on cheese is analysed. On average, the introduction of a K-wire increases the required drilling forces by a 296% (roughly a factor 3). Among the different feed rates and drill types, the force increase of introducing a stagnant K-wire varies between 16% and 575%.
Conclusion. To prevent orthopaedic screws from breaching the bone surface (an overshoot of 0mm) in spinal surgery, the established feed rate speed limit of 0,5mm/s is too low. To meet the observed feed rates applied by surgeons of up to 5mm/s, it is of interest to reduce the DRS sampling time – the inactive period between two measurements in particular. The feed force increase of approximately a factor 3 can be regarded as a challenge for surgeons, who indicated that they preferred feed forces to be kept low. Further testing on real (cadaveric) vertebrae can more give information about the DRS drill’s optical performance in pedicles, and whether the identified feed force increase proves to be problematic for clinical applications.

In spinal surgery, the misplacement of spinal screws is a common problem that causes (severe) pain, bleedings or even paralysis [1] [2] [3]. In order to improve the navigational support of spine surgeons, this research focuses on the development of an optical sensing diffuse reflectance spectroscopy (DRS) orthopaedic drill that identifies bone tissue boundaries, based on fat fraction. The developed drill concept introduces a stagnant optical fiber-equipped probe into a cannulated orthopaedic drill. To verify the clinical applicability of the developed system, the accuracy of the optical tissue boundary detection has been analysed under different tissue penetration speeds, as well as the axial drilling force increases due to the introduction of a stagnant probe into a cannulated drill. The maximum feed rate at which the drill consistently detects the tissue boundary before breaching it, is 0,5mm/s. The force increase due to the introduction of a stagnant probe is a factor 2,96 on average and varies widely between different feed rates and probe diameters (a factor 1,15–5,75). The established feed rate speed limit of 0,5mm/s is lower than drilling feed rates of up to 5mm/s, observed in the operation room. Increasing the sampling frequency –especially decreasing the inactive period between the measurements– is required. The feed force increase of approximately a factor 3 can be regarded as a challenge for surgeons, who indicated that they preferred feed forces to be kept low.

Previous studies show that the smart electrosurgical knife, which adds diffuse reflectance spectroscopy to the traditional electrosurgical knife, is a promising technique for breast-conserving surgery, namely it enables real-time tissue characterization while cutting. More specifically, based on the fat/water-ratio, it enables intraoperative healthy from malignant tissue discrimination, therefore potentially reducing the re-excision rate with breast-conserving surgery procedures. However, the smart electrosurgical knife cannot be used on patients yet since it has not been validated yet. Several studies suggest phantoms are ideal for validation of systems including imaging systems like diffuse reflectance spectroscopy. Hence, for this master thesis, the objective was to develop a breast phantom that enables validation of the smart electrosurgical knife.

Firstly, it was found that a study which enables validation of the smart electrosurgical knife, should mimic a breast-conserving surgery procedure including a breast phantom, so that the potential of the intraoperative margin assessment technique ‘’ diffuse reflectance spectroscopy ‘’, added to the traditional electrosurgical knife, could sufficiently be tested. For such a study to take place, it was found that the phantom should have a similar size and shape as human breasts containing a tumour. With regard to pre- and postoperative margin assessment, the phantom should have a contrast between the tumour and healthy phantom, which enables size, border, and location assessment of the phantom tumour upfront and residual tumour inspection after surgery. Intraoperatively, the phantom should have a significant difference in fat/water-ratio between the tumour and healthy phantom. This enables us to assess diffuse reflectance spectroscopy with its capability in discriminating healthy from malignant tissue. Furthermore, visually and mechanically, there should be a minimal difference between the tumour and healthy phantom, which eliminates the possibility of using the intraoperative margin assessment techniques, palpation and visual inspection. Finally, the phantom should have similar mechanical- electrically conductive and thermo tolerance properties as real breast tissue. This will result in realistic haptic feedback and tissue effects with electrosurgery.

To develop such a phantom, various fat/water-ratios of water and lard, in combination with various additives such as guar gum, agar, gelatin and barium sulphate, were produced and tested. It turned out that agar in combination with water, lard, and the contrast agent barium sulphate, enables breast phantom production, that meets all the aforementioned phantom requirements. More specifically, the final phantom is a breast-shaped phantom with a realistic size and shape, consisting of healthy tissue with a tumour inclusion. The healthy tissue is composed of 50% lard, 50% water and then 5% agar by weight of water, whereas the tumour is composed of 20% lard, 80% water and then 3% agar- and 5% barium sulphate by weight of water. Since this phantom meets all requirements, it enables the design of a study that subsequently enables extensive testing and further validation of the smart electrosurgical knife.

Cardiovascular disease is the number one cause of death in the USA and number two in the Netherlands(1, 2). In addition, it is expected that in 2030 40.5% of the population in America will be inflicted with a form of cardiovascular disease(2). The growing volumes and thereby health care costs will put a strain on hospitals and health care providers, as more individuals need to be treated with the same resources. Workflow and efficiency optimization are topics that are gaining momentum over the last years in the field of medicine (3-6). Health care providers see this as a viable solution to treat more patients and reduce the costs, i.e. increasing efficiency. This research group intend to automatically evaluate the workflow efficiency in the cath lab with use of two sensors: the X-ray imaging system and video cameras. In order to facilitate automatic workflow monitoring and optimization, prior research should be conducted, as “the initial step for improving the process is to acquire knowledge of how things are done”(7). To continue in this line of reasoning, in this research the current peri-operative workflow of the CAG procedure is formalized and quantified.METHODWe used the SPM (surgical process model) methodology to formalize the peri-operative workflow of the CAG. The role activity diagram (RAD) was chosen as the model representation. The model approach was top-down. The metrics were determined and designed based on a previous conducted literature study, the formalization of the workflow and the design criteria. The metrics were divided in two groups: 1) the high-level metrics and 2) the procedure and CAG specific metrics. To evaluate metrics two datasets were used: 1) the manually acquired dataset through observations and 2) the dataset retrieved from the X-ray imaging system. To validate the methodology, the outcomes of the metrics were compared with the experience of the cath lab team. The generalizability was tested by utilizing the phases and metrics in different endovascular procedures and in a different hospital.RESULTSIn total 19 metrics were created of which 11 evaluated the high-level workflow, three the procedure specific process and five the CAG specific steps. Prior the first procedure in the morning, the LA waiting time showed an average of 37 minutes and on average seven minutes are used for lab preparation. In the afternoon similar results are found: the LAs waiting time is longer than the actual lab preparation. The average percentage on-time start cases was 43%. A significant mean difference of 3.3 minutes (p=0.041) regarding the turn-over time was found between the two LA and three LA group. Delivery time showed a mean difference of 3.7 minutes (p=0.004), comparing inpatients (M = 8.1± 5.43 minutes) to outpatients (M = 4.4 ± 4.4 minutes). There was significant positive correlation between delivery times and turn-over times, r=0.66 (p<0.001). Longer delivery times resulted in longer turn-over times. The procedure time presented the highest mean and variance, M=38.0 ± 21.1 minutes. The patient preparation time and post-care duration presented low variances, respectively M=11.8 ± 3.8 minutes and M=4.6 ± 3.3 minutes. The variability in the procedural time could be attributed to the operative phase, M=22.4 ± 18.8 minutes. Within the operative phase, the duration of P1 and P2 accounted for 70% of the variability in the operative length. The X-ray imaging system was able to correctly measure consolidation of metrics describing P2, P3 and P4. The non-procedure specific metrics have the potential to be generalized to any type of procedure. Only P1 and P5 showed potential to be utilized in other endovascular procedures. The analysis in Haga hospital suggests that phases and the metrics could be generalized and therefore used in more cath labs throughout the Netherlands.DISCUSSIONThis research has paved the way for in-depth efficiency assessment and workflow optimization in the cath lab. The cost of delay in an OR is $25 /minute(8). In the event that turnover time can be shortened by five minutes, 20 minutes of scarce lab time would be saved each day. Assuming the cost of delay in the cath lab is similar to the cost of delay in the OR, 500 $/day can be saved by optimizing the turn-over times. Moreover, the LAs wait in the morning for 20 minutes and in the afternoon for 15 minutes. In case the waiting time could be reduced with 50%, 18 minutes of wasted time would be utilized and 450 $/day would be saved. Enhancing starting time and the turn-overtime can result in saving 38 minutes that account for 950 $/day, which is approximately the scheduled time of one CAG procedure. As a result of optimization, an extra patient could be treated every day and money could be saved.

Maximisation of treatment efficiency in hospitals can lead to significant growth in terms of patient satisfaction, staff productivity and hospital revenue. Scheduling for the operating department is currently done manually, using standard values for duration that vary with procedure type. This implies that no patient characteristics nor historical data are used to personalise predicted surgical duration. Furthermore, schedules are not updated when procedures are delayed. A relevant step towards dynamic scheduling is the realisation of real-time analysis of surgical workflow, based on intraoperatively acquired data. As of yet, the amount research performed into the use of automatically generated online data for duration prediction is limited. This research project involves an analysis of the usability and characteristics of interventional X-ray data, for predicting the total duration of coronary angiograms intraoperatively. A random forest classification algorithm is used to analyse each acquisition within the dataset and classify the total duration of the corresponding procedure as being below 10 minutes, 10 to 20 minutes, 20 to 30 minutes or 30 minutes and over. An additional 22 features were generated to add data from prior acquisitions of the same patient. Recursive feature elimination was used to determine the optimal feature set, to be used in the final model. Based on out-of-bag validation, the overall accuracy of the classification model was found to be 92.8\%. Considering an average procedure duration of 11 minutes, the interventional X-ray data shows exceptional capabilities of classifying both standard and delayed procedures. Further analysis with respect to procedure time shows that some Class 1 acquisitions are overestimated as Class 2 acquisitions, but that overestimation of procedure duration rarely occurs for the other three phases. This implies that a prediction made beyond 10 minutes into the procedure can be perceived as the absolute minimum duration class of the procedure. Additionally, Class 3 and Class 4 predictions are found to always correspond to the minimum procedure time, independent of procedural progress. Class 2 and Class 3 acquisitions are underestimated up until 10 minutes and 20 minutes into the procedure, respectively. Class 4 acquisitions are correctly classified at a relatively earlier point with respect to progress and are never incorrect beyond 20 minutes. Given the fact that overestimation does not occur beyond 10 minutes, nor when Class 3 or Class 4 is predicted, the first prediction of a new duration class is concluded to be a reliable reference point. Further investigation has shown that most Class 3 and Class 4 detection occur within the first five minutes of a procedure. Therefore, the model is successful at predicting a total duration of 20 minutes or more at an early stage of the procedure. This could significantly benefit the hospital in terms of procedure planning and knowing when to request the next patient. For a deeper understanding of the model's potential, individual procedures were analysed to gain insight into its additional value in terms of procedure scheduling. In terms of prediction features, the acquisition frequency, longitudinal position of the operating table, cumulative procedure duration, patient age and cumulative cine acquisition time were found to be the most important. The implementation of these features in further research on CAG phases is recommended. Movement of the operating table seems to be particularly informative for workflow analysis. In terms of clinical application, further optimisation is required in order to enable accurate prediction for the duration of shorter procedures. Nevertheless, the model provides an accurate tool for the real-time monitoring of procedural workflow and detecting significant delay, which proves that the usability of the data goes far beyond machine maintenance and service only.

Atrial fibrillation (AF) is the most common form of arrythmia. Without treatment, the heart rhythm usually becomes permanently disrupted and can lead to various heart related complications as stroke or heart failure. Cardiac phantoms are needed for the development of electrophysiological instruments and for training and education of electrophysiologist. A tissue mimicking phantom is an accurate model of organs or tissues mimicking the real life like scenario for a desired goal. Currently, no cardiac phantoms are available containing all requirements needed. Therefore, as a first step towards an applicable cardiac phantom for development of electrophysiology studies, this Master Thesis focused on the design of an anatomically anthropomorphic left atrium (LA) phantom mimicking the dielectric properties of human heart tissue in the frequency range of 10 kHz – 20 kHz. To assess this research goal, the objective was divided into 3 subgoals: 1. Evaluate different tissue mimicking materials with additives to meet the main phantom requirements. Three promising types of phantoms were examined: agar-, gellan gum- and polyvinyl alcohol-cryogel-based (PVA-C) phantoms. The dielectric properties and stiffness of the phantoms were manipulated by varying the concentrations of materials and by using additives. Graphite powder, charcoal powder, salt and sodium acetate were examined for increasing the electrical conductivity and oil, sugar and PVC for decreasing its relative permittivity.2. Extensively evaluate the most promising cardiac phantom materials. The most promising hantom tissue materials, based on the main phantom requirements, was used for the final design of a LA phantom. The cardiac phantom was prepared four times for extensive examination based on all phantom requirements.3. Develop a phantom design with comparable anatomy.
As a last step, the phantom was fabricated with comparable anatomy. In this part, a suitable mould was prepared in order to get a hollow shaped phantom.

Results
A 13,0wt% PVA, 0,15wt% salt and 0,05wt% sodium acetate and tap water solution was used for the final LA phantom design. The flexible cardiac phantom is easily made, affordable and has a shelf life of at least 1 month. The average difference between the target values and the mean of all four prepared cardiac phantom samples is 21.9% and 15.0% for permittivity and conductivity, respectively. Storing the phantoms in an aqueous liquid results in diffusion of salt and sodium acetate and water absorption of the phantom material. Stability is reached after 1 week. Finally, a cylinder-shaped hollow cardiac phantom is produced, with the use of a 3D printed mould, as comparable shape of the LA.
Producing a cardiac phantom of PVA is proved to be a good model. This research shows the design of a low cost, easily produced hollow shaped cardiac phantom consisting of 13.0wt% PVA, 0.15wt% salt and tap water. Even though more future work needs to be done for the development of a cardiac phantom, this research functions as a proof of concept for the development of a flexible phantom with possibilities to manipulate the dielectric properties and to create a hollow shaped phantom. The use of PVA as tissue mimicking material (TMM) and the use of 3D printed PVA for creation of the hollow shapes shows promising results for future development.