The removal of tissues from the human body is a fundamental aspect of surgical interventions, especially in Minimally Invasive Surgery (MIS). In MIS, one or multiple small incisions (approximately 3 to 10 mm) are made, and the body is accessed by long slender devices. MIS offers benefits such as less postoperative pain, and shorter hospital stays, driving a trend toward miniaturisation: smaller medical instruments. Current suction-based instruments are limited in performance by clogging, shaft miniaturisation challenges, and dependency on tissue composition. To address these limitations, alternative mechanisms for transportation have arisen such as friction-based transportation. This study presents the design and experimental validation of a flexible friction-based tissue transporting device for MIS. The device is designed to achieve transportation efficiency independent of tissue elasticity within a Young's modulus range of 1-110 kPa, utilising a cylindrical conveyor mechanism with wires to transport tissue. Experimental validation demonstrated consistent transportation efficiency (75%) across various tissue types, shapes, and orientations, though shaft curvature significantly affected performance and total reliability (81%). The device addresses some of the limitations of suction-based instruments, including reduced clogging, enhanced potential for miniaturisation, and transportation efficiency independent of tissue elasticity. While further improvements to the device design are necessary, it has the potential to improve MIS tissue removal procedures and patient outcomes.

Providing adequate healthcare in low- and middle-income countries (LMICs) faces significant challenges due to the frequent failure of medical devices, including hospital beds. These failures arise from environmental stressors, resource limitations, and design mismatches that fail to address the specific needs of LMIC settings. This thesis examines these challenges and develops a robust and modular solution tailored to these contexts, focusing on designing an appropriate backrest movement mechanism as a foundational component of a Universal Modular Hospital Bed.

Hospital beds are essential medical devices that provide quality patient care, comfort, and safety. However, existing designs often need to be more suitable for the harsh environmental conditions and resource constraints of LMICs, leading to high failure rates, compromised patient safety, and increased operational costs. This research addressed these gaps by designing a backrest movement mechanism that prioritizes durability, affordability, and contextual adaptability while laying the groundwork for a Universal Modular Hospital Bed suitable for diverse healthcare environments.

The study employs a comprehensive methodology integrating a systematic literature review, fieldwork, and a structured design process. The literature review identified key causes of medical device failures in low- and middle-income countries (LMICs), including environmental challenges, maintenance issues, and inadequate design considerations, which also apply to hospital beds. Fieldwork in Nepal, Suriname, and the Netherlands assessed hospital bed performance in different contexts. Drawing from these insights, a structured design process was used to generate, evaluate, and refine several concepts for a backrest movement mechanism. The chosen concept was then validated through prototyping and testing to ensure it met performance criteria, including robustness, safety, and ease of maintenance.

The research revealed that hospital beds in LMICs frequently fail due to high humidity, fluctuating temperatures, and unreliable infrastructure, compounded by limited access to spare parts and maintenance expertise. These challenges necessitate a context-sensitive design approach. The designed solution for the backrest movement mechanism employs a counterweight-based system that is manually operated, minimizing dependency on electricity and ensuring usability in resource-constrained settings. The design's modularity allows for easy maintenance, scalability, and integration with more advanced features, making it suitable for LMICs and high-income countries (HICs).

The backrest movement mechanism developed in this study addresses the critical challenges of hospital bed performance in LMICs. Its modular design ensures adaptability and compatibility with various healthcare settings, while its robust construction enhances durability and reduces maintenance costs. By focusing on local manufacturing and cost-effective materials, the design promotes accessibility and sustainability.

Future research should focus on scaling the modular design to incorporate additional features such as height adjustment and lateral movement systems. Collaborations with local manufacturers and healthcare providers are recommended to facilitate adoption and implementation. Long-term field testing will further validate the design’s efficacy and provide insights for continuous improvement. This study contributes significantly to developing medical devices that are both contextually appropriate and globally adaptable, addressing a critical gap in healthcare infrastructure in LMICs.

INTRODUCTION Artificial intelligence (AI) has great potential to optimise patient care and reduce the bur- den on healthcare. Despite numerous AI solutions being developed for hospitals, the clinical adoption rate remains low, largely due to challenges in evaluating their practical usefulness before implementation. This research addresses the gap by developing a decision-making strategy that hospitals can use to assess whether specific AI applications should be implemented. The strategy aims to ensure that AI solutions are effec- tively integrated, address genuine problems, and meet stakeholder needs, thereby facilitating better decision- making and more successful AI adoption in clinical settings.
METHODS The methodology involved three key phases: development, testing, and evaluation of the decision- making strategy. The strategy was initially developed through an analysis of existing guidelines, specifically the "Stappenplan Healthy AI (HAI)" document, and corresponding literature. Stakeholder evaluations, in- cluding input from hospital AI teams, medical officers, and clinical specialists, were used to refine the strategy. A practical case study was conducted in the Radiology department of the Noordwest Ziekenhuisgroep (NWZ) hospital to test the strategy’s applicability, followed by stakeholder feedback through a structured question- naire to evaluate its effectiveness and usability.
RESULTS The results showed that the iterative development process, involving multiple rounds of stake- holder feedback, substantially improved the decision-making strategy’s comprehensiveness and relevance. Stakeholders highlighted that the strategy effectively captured critical aspects of AI integration, such as tech- nical requirements, stakeholder needs, and workflow implications. Testing in the Radiology department re- vealed challenges in identifying responsible individuals for data collection, which initially delayed the pro- cess, but also underscored the need for well-defined roles. The feedback from stakeholders was largely pos- itive, indicating that the strategy was clear and practical for evaluating AI solutions, though some improve- ments were suggested for addressing technical integration and detailing follow-up actions. Stakeholders ap- preciated the structured format, which facilitated effective communication and collaboration among differ- ent departments. Overall, the decision-making strategy succeeded in creating a robust framework for evalu- ating AI applications, helping ensure that such technologies are implemented thoughtfully and effectively.
CONCLUSION The aim of this study was to develop a decision-making strategy for hospitals to determine whether AI applications should be implemented, as well as to test and evaluate the strategy. The iterative process proved effective in creating a practical and efficient tool that helps identify potential bottlenecks and clarifies resource needs for implementation. The involvement of ICT stakeholders was crucial, highlighting the importance of technical evaluation as a key factor in decision-making. Overall, the strategy provides a focused and manageable framework that allows hospitals to evaluate AI applications effectively, supporting informed decisions to improve healthcare efficiency and patient care quality.

Background: Vasovagal syncope (VVS) is the most common form of syncope, accounting for more than half of all syncope cases. At least 35% of people between the ages of 35 and 60 have experienced VVS at least once in their lives. The Head-Upright Tilt Test (HUTT) is commonly used to elicit VVS while monitoring clinical signs and changes in heart rate (HR) and blood pressure (BP). The current protocol is both uncomfortable and time-consuming with suboptimal diagnostic yield. This study aims to develop and evaluate a machine learning (ML) pipeline capable of providing an early prediction of whether a patient with suspected VVS will experience syncope during HUTT, while also identifying features that contribute to a better pathophysiological understanding of VVS.

Methods: The study included 434 adult patients with suspected VVS who underwent HUTT. A ML pipeline was developed for the early prediction of VVS, classifying patients into syncope and no syncope groups. Based on the results of the preliminary study, raw continuous BP, HR and Electroencephalogram (EEG) data were separated into 3-minute epochs before and after the tilt. Linear interpolation was selected as the most effective method to fill in missing data points. The Python package ‘tsfresh’ was used for feature extraction and the Boruta algorithm for feature selection. Three classification models were tested: Random Forest (RF), Support Vector Machine (SVM) and Extreme Gradient Boosting (XGBoost). Model performance was evaluated using 5-fold cross-validation with AUC as the primary metric. In addition, a SHapley Additive exPlanations (SHAP) analysis was performed to assess the importance of features, which provided insight into the features that contributed most to the model's predictions. These features were further analyzed to determine the significant difference between both the syncope and no syncope groups.

Results: The best performing model was the RF model with 61% AUC, 70% sensitivity, and 45% specificity. All selected features contributed to the classification of the syncope and no syncope groups across all folds. Five features were selected repeatedly during cross-validation, including three stroke volume index (SVI) features and two HR features. The two most frequently selected features were the after SVI minimum and the before HR partial autocorrelation features. In particular, the after SVI minimum feature was identified as a valuable feature in every fold. Both features showed a significant difference, suggesting that higher values for these features were associated with a greater likelihood of not experiencing syncope during HUTT.

Conclusion: This study introduced a novel approach to the early prediction of VVS during HUTT. By developing an automated ML pipeline, we demonstrated that with hemodynamic features from the 3 minutes before and after the tilt can provide predictive insight into the occurrence of syncope 20 to 30 minutes later, although with limited sensitivity and specificity. Future research should focus on methods for handling artifacts and outliers and improving model robustness and accuracy. Ultimately, early detection of syncope has the potential to make HUTT more efficient, reduce patient discomfort, and avoid unnecessary testing.

Open Wedge High Tibial Osteotomy (OWHTO) is an extensively used, effective treatment option for medical conditions such as medial knee osteoarthritis and varus malalignment. The current developments in the 3D printing industry facilitated using 3D printed patient-specific guides (PSSGs), making OWHTO a desirable treatment option. Although the PSSG improves the outcomes (e.g., high accuracy, lower radiation) of the OWHTO, lateral hinge fractures and posterior soft tissue injuries of the tibia are two main risks still present. Therefore, this study aims to facilitate an optimal 3D printed PSSG design with a lateral hinge protecting K-wire and a soft tissue protecting sleeve for OWHTO that can prevent or reduce these two critical risks. For this purpose, first, 3D models were constructed from pre-operative CT images of the cadaver specimens. Afterward, a monopolar osteotomy cut was created at a location where the orthopedic surgeon Tom Piscaer decided on the 3D-constructed models to develop Finite Element (FE) analysis. Using the FE analysis and 3D constructed models, different intact lateral hinge lengths (5 mm, 6 mm, 7 mm, 8 mm, 9 mm, and 10 mm) without apical drill holes were analyzed for relative comparison between six different simulations. The intact hinge length was chosen between 10 mm and 5 mm according to the information in the literature. Von Mises stress was used as an outcome measure. The decrease in distribution and magnitude for the von Mises stress at the lateral side of the tibia and the lateral hinge was used to point out the overall impact of the hinge length reduction. As the intact hinge length decreased, the von Mises stress on the lateral hinge generally decreased. According to the results from FE analysis using a high-quality mesh, leaving a 6 mm intact hinge was determined to be the optimum case among other cases. It reduced the stress around the hinge, maintained sufficient cortical and cancellous bone, had a higher margin of error, and likely reduced the fraction risk. The results of this FE analysis were used to develop the hinge K-wire location on the PSSG. Several PSSG models with soft tissue-protecting sleeves were created. Among these designs, the second and third prototypes (3 prototypes in total) were tested on the cadaver specimens. After the tests, post-operative CT images of the cadavers were taken. From these images, 3D models of the tibia were constructed. Pre- and post-operative 3D models were registered onto each other for measuring the deviation between the planned and obtained osteotomy. This was used to analyze the accurate placement of the PSSG and its effect on the sleeve function. While the translation deviations ranged from 5.2 mm, 2.05 mm, and 7.81 mm, the rotational deviations ranged from 7.04°, 1.55°, and 9.16° for cases 1, 2, and 3, respectively. Two PSSG designs passed the acceptable region for translational malpositioning (lower than 5 mm). Although the malpositioning of the PSSG affected the sleeve position slightly, the cadaver test results showed that the sleeve still protected the tibia's posterior side from soft tissue injuries by stopping the cutting saw at the cutting trajectory due to its lengthy coverage at the posterior side of the tibia. Since the cadavers did not have the necessary soft tissues, it wasn't easy to give a concrete answer. Still, according to the results, the sleeve is expected to ensure soft tissue protection. Using the PSSG with a hinge-protecting K-wire and a soft tissue-protecting sleeve will likely provide lower radiation exposure, shorter operation time, fewer complication risks and lateral hinge fractures, less posterior soft tissue damage, and increased healing.

Abstract—Background: Vacuum extraction is a way to prevent unnecessary maternal deaths, by using a vacuum extractor to help deliver the baby in the second stage of labour. Vacuum extraction is still little-used in low- and middle-income countries (LMIC) compared to high-income countries, due to the lack of sufficient devices on the market that fit the local context. For this purpose, a novel reusable and affordable vacuum extractor was designed. This study compares this device, the VelaTM, with a commonly used vacuum extractor, the Kiwi Omnicup®, in terms of performance and user experience in a non-inferiority mannequin study. Methods: Obstetricians and residents from LUMC and CWZ hospitals performed a vacuum extraction on a birthing mannequin using both devices. During simulated “traction windows” of 25 seconds, the participant was allowed to pull. The primary endpoint was the traction window in which successful vacuum extraction occurred. Secondary endpoints were the total number of pop-offs of the cup, the exact success time within the successful traction window, the pumping time before the successful traction window, the satisfaction rate and the comparison between the two devices on placement, pumping and pulling. Finally, the participants were asked for their opinions about the Vela in an open-ended question. Results: A total of 47 participants were included in this study, of which three never achieved success with the Vela. For the primary endpoint, the Vela was non-inferior to the Kiwi in both LUMC and CWZ. Furthermore, no significant differences were found between the Vela and the Kiwi regarding any of the secondary endpoints, with the exception of the satisfaction rate in the LUMC. This was significantly lower for the Vela. Themes raised by participants concerned the stiffness of the Vela tube, the comfort of the Vela during usage and the need for familiarization with a new device. Conclusions: The results of this study indicate that the Vela is non-inferior to the Kiwi Omnicup when used by clinicians who had success with both devices, however there is room for improvement regarding satisfaction of the Vela. More familiarization is needed and the tube of the Vela should be reduced in stiffness. With this initial validation, the Vela can be seen as a promising alternative for LMIC, and can be further optimized into a high-quality fully reusable vacuum extractor.

The quality of rigid endoscopes deteriorates during clinical use due to the sterilization process, mechanical forces and wear and tear during regular use. Regulations and standards on ensuring the quality of these instruments are currently non-existent or contain only qualitative measures, resulting in subjective examinations . Defective rigid endoscopes still reach the operating room, resulting in direct and indirect patient risks. An experimental test set-up has been developed to quantify the sharpness, contrast, distortion, light transmission, vignetting and colour correctness of the optical system
of rigid endoscopes. Results are given for 85 measurements performed on 33 endoscopes, including 7 high quality endoscopes and 26 low quality endoscopes. 37 measurements have been performed on a reference rigid endoscope. The results for sharpness, contrast and distortion provide valuable insights but the current design of the test set-up proved not stable enough to draw significant conclusions. The results for light transmission, vignetting and colour correctness produce stable results and display the expected values for damaged lens systems such as loose or broken lenses. Although the current design requires significant optimization, this study has both given an account of the need for an objective method to evaluate the quality of rigid endoscopes as well as provided a promising step towards this new method and greatly encourages further research.

The goal of this study was to design a maternal phantom model specialized for experiments in which the stresses in the amniotic sac can be determined during fetoscopic procedures. These stresses caused by surgical procedures are believed to be the leading cause of post surgical pre-birth. Measuring these stresses indicates the quality of fetoscopic instruments and surgical procedures.
Founded with literature and experiments with amniotic tissue, the requirements for the design of the phantom model were formulated. With these requirements as a framework a design was made which was validated by medical specialists.
Measuring the stresses in the amniotic sac failed, therefore measuring the translations of the amniotic tissue was used as a measure of trauma. This proved a success as medical specialists were positive about the design of the phantom model.
The findings in this study support the idea that there is a need for a maternal phantom model for the development of fetoscopic procedures. Measuring the translations of the instruments through the amniotic sac is a successful method for indicating trauma. The resulting design showed potential as a training tool as well.
The developed phantom model is a successful first step in making a maternal phantom model to contribute to the quality of fetoscopic procedures.