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.

Surgical aerosols in other words plumes are produced during thermal tissue destruction in medical operations. The cellular debris in the form of particulate matter may contain viruses and harmful chemical compounds which can lead to an infectious transmission in case of inhalation. The goal of this study was to design and produce an experimental setup to simulate Plasmajet (PJ) and ERBE experiments with minimal airflow disturbances. Such setup can lead to defining the lowest aerosol production conditions, investigating the production-affecting factors, and evaluating tissue effects to promote a safer and healthier surgical environment for both healthcare workers and patients. A clear correlation between the aerosol production affecting factors and particle counts was established for particle sizes 0.3, 0.5, 1.0, 2.0, 5.0, and 10.0 µm. The results of the experiments showed that among all experimental conditions, the PJ coagulation mode with fast operation yielded the lowest aerosol counts. In cutting mode, the lowest aerosol counts were also produced by PJ with fast operation speed. However, between speed and aerosol counts, no statistically significant correlation was found. Upon analyzing the correlation between aerosol counts and tissue effects, it was determined that, for ERBE device, higher aerosol counts were associated with darker tissue effects. In the case of the PJ device, this relationship persisted in the cutting mode, whereas no connection between tissue effect and particle counts was observed in the coagulation mode. Further investigation on the toxicity of the produced particulate matter and establishment of a clear minimal aerosol intake is recommended. Until then, preventive measures such as implementing local exhaust ventilation and using surgical N95 masks are strongly advised to minimize aerosol inhalation.

Brachytherapy is a type of internal radiation therapy that is used to treat cervical cancer. It involves the application of a radioactive source in close proximity to the tumour. This can be done either by directly inserting the source into (or close to) the tumour using interstitial hollow needles, or by loading the source into an intracavitary applicator that is placed in the vaginal cavity. Standard applicator types may lead to suboptimal radioactive source placement, possibly resulting in underexposure of target volumes and overexposure of organs at risk, especially in advanced cancers. Customised applicators and optimised needle channels based on the patient's MRI/CT data could enhance conformity between target volumes and prescribed isodose. Hence, the goal of this project was to develop and validate a design for a 3D-printed brachytherapy applicator which geometry is based on the patient's vaginal cavity and which features optimised interstitial needle channels based on the patient's anatomy and tumour location. Various analyses have been conducted which have led to the establishment of a list of requirements for the applicator design. Based on this list of requirements, two conceptual designs have been presented: one fully 3D-printed design and one design that is clicked on the Geneva ovoid tubes. Through the creation of prototypes, these conceptual designs have been refined into two final designs which were manufactured in PA12 using selective laser sintering. The dose attenuation properties of PA12 were evaluated and compared to that of water. Furthermore, the potential needle positions within the proximal end of both designs have been analysed. For both designs, a final prototype based on a phantom's vaginal cavity geometry has been created. The usability of these prototypes has been tested by three radiotherapist-oncologists, who also provided feedback on the designs. Upon analysing their feedback and the outcomes of the other evaluations, recommendations for future designs have been formulated. The conducted dosimetry experiment yielded a maximum difference of 0.8% between the average percent dose depth curves of water and PA12, which can be considered a water-equivalent response. This allows PA12 to be used as the material for the applicator. The result of the potential needle position analysis suggest that the first design provides more space for personalised needle channels in the top of the applicator compared to the second design. The three radiotherapist-oncologists validated the usability of both final prototypes. Two designs of a patient-tailored 3D-printed brachytherapy applicator containing optimised interstitial needle channels based on the patient's anatomy and tumour location have been presented, produced and validated. Based on the outcomes of the conducted evaluations, there can be concluded that the first concept shows the most promise to be used as a design for a patient-tailored 3D printed brachytherapy applicator. However, to ensure the proper functioning of the working principles, further development is required. If the recommended improvements are implemented, the design has the potential to be used as applicator in the treatment of cervical cancer.

Vulvar cancer has significant implications for women's health and quality of life, emphasizing the need for prompt diagnosis and treatment. Surgical treatment aims to remove both visible and nonvisible malignant cells, incorporating a safety margin beyond the tumour perimeter. However, accurately assessing cancer spread and determining the optimal safety margin size remains challenging. Hyperspectral imaging allows non-invasive identification of vulvar tumour boundaries, capturing both spatial and spectral details of the region of interest. However, tissue type classification in hyperspectral imaging is complex due to the abundance of wavebands and limited training samples. Artificial intelligence (AI) methods show promise for accurate and automated tumour classification in hyperspectral imaging. In this study, four AI models (Support Vector Mechanism, Neural Network, 1D and 3D Convolutional Neural Networks) were trained to classify tumour tissue, skin tissue, and mucosa tissue. Additionally, the study aimed to explore the use of specific explanatory parameters (Tissue Water Index, Near Infrared Perfusion Index, and Tissue Haemoglobin Index) to reduce the spectral dimension, potentially improving acquisition time and interpretability without compromising classification accuracy. The study included 25 patients (mean age 71), with hyperspectral cubes captured from each patient. All cubes contained healthy skin tissue, while 22 cubes included tumour tissue and 10 cubes contained mucosa tissue. All models exhibited comparable performance for tumour detection, with F1 scores ranging between 0.91 and 0.93 and AUC-ROC scores ranging between 0.91 and 0.94. Moreover, the Neural Network trained only on the explanatory parameters achieved excellent results for tumour detection with an F1 score of 0.93 and AUC-ROC of 0.91. Overall, combining hyperspectral imaging with all models shows great potential for in vivo tumour detection, while leveraging physiological explanatory parameters with a Neural Network model can enhance acquisition time and data interpretability without compromising the classification accuracy.

PURPOSE. The purpose of this thesis is to investigate the possibility of implementing a prototype of a steerable needle in a medical procedure. Therefore the two main research questions are: “Can the intended user perform a steering motion with the steerable needle and reach the target area following a straight and curved trajectory?” and “Does the tip of the cannula maintain its position after the instrument change with a biopsy needle for a straight and curved trajectory?”. METHODS. Two studies were conducted to answer these research questions. A user study was designed and performed to test the steering properties of the steerable needle for a straight and curved trajectory. A second controlled study with automated insertions was designed and performed to investigate the cannula tip displacement after the instrument change. RESULTS. The results of both studies show that steering towards a specific target is highly achievable with the steerable needle for insertions with a straight and curved trajectory. Results also show a high success-rate, a low time consumption for the instrument change, and a high satisfactory level on the manual handling of the steerable needle by the interventional radiologists. The results of the second study show that the cannula tip maintains its position after the instrument change for a straight trajectory, and the cannula tip only displaces 1.4 mm back over a deflection of ±25 mm after the instrument change for a curved trajectory. CONCLUSION. Reaching the target with a steering motion is achievable with the steerable needle and only a small displacement of the cannula tip is seen after the instrument change for a maximum imposed curved trajectory. These results answer the main questions concerning the effectivity of the steerable needle and are promising for future implementation of this prototype in a real medical procedure, containing real patients, and real liver tissue. KEY WORDS. liver disease – liver biopsy – radiofrequency ablation – steerable needle – polyvinyl alcohol

Radiofrequency ablation (RFA) is a localized thermalintervention technique with the purpose to destroy small tumours (< 3 cm) byheating the tumour tissue. Due to various challenges with targeting andvisualisation during RFA, the clinical need of steerable needles is present forradiologists. Helwig, B.P. developed a new omnidirectional steerableneedle for this purpose. However, this needle is not yet ready forimplementation in the operation room (OR) because no clinical testing isperformed in vivo. Therefore, the aim of this thesis is to develop a novelneedle guide tool for percutaneous interventions with an omnidirectionalsteerable needle during real-time US-guidance. In order to examine the visibility of the steerable needle,an experimental study has been conducted. The steerable needle has a promisingvisibility in Polyvinyl alcohol and is expected to be well visible duringultrasound guided interventions. The developed needle guide tool is calledUShift and provides both the benefits of needle guided puncturing and free-handpuncturing. Which are for free-hand puncturing: freedom of movement of boththe probe and needle during puncturing and for needle guided puncturing:  a shorter procedure time, reduced needle manipulation and improvedneedle visualisation. A user test is performed to verify how the needle guide isperforming compared to the currently used needle guide while hypothetical usingthe Omnidirectional steerable needle. The difference between the two puncturetechniques is 0,9 seconds, the UShift is 23% faster. In conclusion, thisproject resulted in a novel needle guide tool that can be used in combinationwith the omnidirectional steerable needle during percutaneous interventions.The UShift has great benefits compared to currently used needle guides and istherefore stated as innovative. More gain is to be expected through future researchconcerning production methods, material options, the cost-effectiveness andother application fields of the UShift.

In recent years, more and more medical operations are done minimally invasive. Intervention radiology is a medical specialty which uses minimally invasive techniques to diagnose, or treat diseases. Instruments like needles and catheters are used by radiologists to enter the network of veins and arteries guided by image modalities. A complex treatment in the interventional radiology is the Transjugular Intrahepatic Portosystemic Shunt (TIPS) procedure. This treatment is developed for people who suffer from liver cirrhosis which are not eligible for liver transplantation. The problem which arise with a liver affected by cirrhosis is that it can not transmit enough blood. Without transplantation or a treatment this will eventually lead to death. During the TIPS procedure a connection is made between the right hepatic and the portal vein by a shunt. As a result, blood pressure reduction in the portal system since the blood can flow back to the right atrium of the heart, bypassing the liver. The hardest part in a TIPS procedure is the intrahepatic puncture between the hepatic and portal vein. The interventional radiologist tries to enter the portal vein by puncturing a small stylet from the hepatic vein through the liver tissue. Due to cirrhosis the liver tissue is very stiff and stylet deflection will occur. To reduce the uncertainty of entering the portal vein, this thesis is focused on designing a stylet that is more stiff and able to steer. It is expected that the complexity of the procedure will be reduced and a higher hit rate to enter the portal vein will be achieved.\ \ The prototype of the steerable stylet has been evaluated through various experiments, a visibility test and with procedures in a test liver model. During these experiments the steering characteristics, the stiffness of the stylet, the maximum lateral forces exerted by the tip while steering, the influence of the stylet orientation and the visibility are obtained. Afterwards, an evaluation was done in a liver model, made of PVA, to determine whether the stylet is capable to reduce the complexity of the intrahepatic puncture step in the TIPS procedure. With the prototype made in this graduation project, based on a steerable ablation needle, the complexity of the TIPS procedure is not reduced yet. The steerable stylet was not able to enter the portal vein. It was already hard to enter the right hepatic vein since the pre-bent stiffening cannula was adapted with a smaller angle which was necessary since the steerable stylet was too stiff to push through the pre-bent angle. According to this prototype, possibilities are shown to use a mechanical steering mechanism in instruments with a long thin shaft. The transmission in combination with the joint mechanisms fits within 1.3mm diameter, was able to bridge 60cm from distal end to proximal end and had only 4 components, the stylet, the rigid cannula, the key to fix the stylet to the rigid cannula and the transmission. By translating the stylet in a push or pull direction relative to the rigid cannula, steering angles could be achieved. With further research and development this steering mechanism must be able to steer the required amount of degrees without any extra components, is well visible with ultrasound, and is good resistant against lateral forces.

The technique of ultrasound-guided needle insertion is commonly employed in various clinical fields, including biopsy, anesthesia, brachytherapy, and ablation. However, the visibility of the needle in ultrasound (US) images remains a persistent challenge. To improve the guidance accuracy of needle insertion during interventions, it is crucial to develop a reliable technique that enhances the visibility of ultrasound needles and accurately detects their position and orientation. Recently, Deep Learning (DL) based segmentation methods have drawn attention because of their high efficiency and accurate results. In order to improve the model performance on challenging datasets, previous researchers modified the segmentation models by introducing spatial attention mechanism and temporal information. However, whether the approaches are effective for ultrasound images, especially for the needle insertion tasks remains unclear. This thesis aims to investigate whether deep learning models are able to increase segmentation accuracy as well as localization accuracy in 2D ultrasound images, specifically focusing on introducing spatial attention and optical flow information into U-Net backbone. Spatial Mask Attention U-Net (SMA-UNet) and Optical Flow Attention U-Net (OFA-UNet) were therefore proposed. The hierarchical experiments were designed to evaluate the effects of training loss, mask width and optical flow methods, and then select an optimal configuration for the segmentation models. Furthermore, U-Net, Attention U-Net and two proposed models were validated on datasets collected from pork and beef phantoms, as well as patients. The evaluation results indicate that OFA-UNet has significant improvement in terms of segmentation metrics and geometrical errors compared to the U-Net baseline and the U-Net only considering the mask attention. Specifically, the model achieved Dice of 86.7%, IoU of 88.2%, Precision of 88.6%, tip error of 2.7 mm and angular error of 0.002 radians on the pork dataset. Furthermore, the OFA-UNet shows robustness and consistency in evaluation metrics across three different datasets, indicating its ability to adapt to varying complexities of US datasets.

The goal of this study is to create a phantom which can be used to test strategies for sealing the vaginal cavity, so that ultrasound gel can be used to fill the vaginal cavity improve the visibility of the vaginal cavity in MR-images. Silicone rubber and PolyVinyl Alcohol (PVA) were tested, to check whether their mechanical properties can be used to mimic the bio-mechanical behaviour of Human Anterior Vaginal Wall (HAVW)-tissue. Both materials were used to produce a phantom. Contrast differences in MR-imaging of PVA-hydrogel and silicone rubber with ultrasound gel and air were measured, to verify proper contrast within a phantom. Both phantoms were used to test sealing strategies. Young's moduli for HAVW-tissue were obtained from literature, and formed a range (6.7 - 14.4 MPa). Silicon proved a closer match to the HAVW-tissue, compared to PVA-hydrogel. Calculated Young's moduli were between 1.4 - 4.2 MPa (silicone rubber) and 0.04 - 0.16 MPa (PVA-hydrogel). Contrast differences in MR-images were measured between 177.8 and 1585.7, and proved sufficient enough to be spotted by an untrained eye. The sealing strategies using a probe cover surrounding a catheter, showed an increase of 54.3% (silicone rubber) and 119.9% (PVA-hydrogel) of the filling percentage measured. Indentation testing was confirmed as a possible alternative to tensile testing for the verification of mechanical properties. In this study the production process of a thin-walled double-layered phantom of the vaginal cavity has been developed and tested. The developed phantoms have been used to test strategies for sealing the vaginal cavity. The concept using a catheter surrounded by a probe cover, is considered to improve the sealing of the vaginal cavity.

Brachytherapy (BT) is an essential component in the treatment of cervical cancer as it allows for locally delivering a high dose to the tumour with minimal trauma to surrounding tissues and organs at risk (OARs). However, in advanced cervical cancer patients commercially available BT applicators are particularly ill-adapted and therefore result in suboptimal local control and frequent occurrence of substantial tissue morbidity. Additionally, cervical cancer BT is associated with large dosimetric uncertainty which has been shown to significantly impact the delivered dose and the occurrence of normal tissue complications. The clinical outcomes of treatment may be improved through combined efforts in sophisticated applicator design and robust treatment optimisation. Patient-tailored BT applicators have been introduced to improve dose conformity, but currently rely on manual indication of needle channels. Automated needle channel planning software exists, but does not account for OAR dose constraints or uncertainty in planning. Robust treatment optimisation potentially improves the dose conformity of plans in the presence of uncertainty, but relies on the questionable presumption that optimisation of dwell times can fully correct for suboptimal dwell positions. In this thesis, the freedom of source placement that 3D-printed applicators allow and the principle of robust treatment optimisation are leveraged to develop robust needle channel planning software for personalised applicators. This thesis was accordingly divided into three parts. In the first part, literature was reviewed to establish the dosimetric uncertainty budget and evaluate geometric uncertainty of OARs. Inter and intra-fraction uncertainty are likely the greatest contributors to the uncertainty budget, possibly increasing the delivered dose to OARs with up to 4.0±20% (k = 1). Using dose-response models it was established that this may realistically increase the occurrence of moderate to severe morbidity of the bladder or rectum by 1.5 and 3.7% respectively. The BT needle channel planning problem under uncertainty was accordingly defined as the problem of computing multiple feasible, non-intersecting curvature- constrained channels under probabilistic or bounded spatial uncertainty of OARs. In the second part, a tool termed motion-planning quality function deployment (MP-QFD) was developed to select a suitable motion planning class. Using the results from a pilot study among nine medical specialists, this tool substantiated the preferred choice for an incremental sampling-based motion planning algorithm. In the last part, robust variants of sampling-based planners were introduced that are capable of computing trajectories for non-holonomic agents in environments under uncertainty. In a two-dimensional simulated patient case, it was shown that these planners were able to generate near-optimal trajectories that (probabilistically) guaranteed not exceeding OAR dose constraints. Subsequent dose-based optimisation showed that (robust) trajectory planning could theoretically yield treatment plans with improved dose conformity over those generated for conventional applicators. Due to modelling assumptions, robust motion planning did not result in improved dose conformity over a nominal motion planning approach in a worst-case scenario. Future work should therefore focus on improving our understanding of OAR movement in and during BT treatment and validating this theoretical work in a patient case series.