DNA data storage has attracted growing interest as a potential solution to the challenges posed by rapidly increasing global data generation. It offers distinct advantages over magnetic storage media, including high information density, low energy requirements, and long-term stability. Established DNA writing strategies typically rely on content-dependent biochemical reactions, such as de novo synthesis, molecular modification, or biological recording. Despite their demonstrated utility, these approaches can be limited by reaction kinetics, cost, and operational complexity.

This thesis introduces write-by-partitioning, a network-based DNA data storage strategy in which information is written through physical partitioning rather than content-dependent biochemical reactions. The method employs a pre-formed DNA barcode network in which spatial relationships among barcodes are preserved as proximity-dependent associations. Data are encoded by dividing the network into segments and assigning a symbol to each segment based on the information being stored. For readout, the DNA barcode associations are sequenced to reconstruct the network and infer the original symbol sequence.

This thesis establishes a workflow for data encoding, network reconstruction, and decoding. The relationship between network properties and data storage performance is investigated. Because decoding depends on network reconstruction, we introduce a method for evaluating the spatial coherence of the reconstructed network. This work demonstrates the feasibility of network-based data storage and provides a fundamental framework for the development, evaluation, and optimization of future systems.

By replacing content-dependent biochemical writing with the physical partitioning of a pre-formed network, write-by-partitioning could provide a faster, more scalable, and more accessible approach to DNA data storage. Beyond data storage, the concepts and analytical methods developed could support broader applications of DNA barcode network.

Carboxymethylation and oxidative carboxylation of four available lignins were investigated as routes to introduce carboxylic acid functionality for crosslinking with epoxidized linseed oil. Hardwood lignins showed consistently higher reactivity toward both routes, explained with DFT calculations and conformational modeling linking the molecular conformation of syringyl-rich lignin to greater hydroxy group accessibility. WAXS analysis showed that oxidative carboxylation disrupted supramolecular packing more extensively than carboxymethylation, correlating with the complete solubility of oxidized lignins and homogeneous mixing with the epoxide matrix.

A feasibility study confirmed crosslinked network formation but revealed phase separation and brittleness as key limitations. Incorporating oxidized lignin with PEG-400 yielded thermosets with approximately 90% bio-based content, gel contents of 88–90%, glass transition temperatures of 93–109 °C, and hydrophobic surfaces. Both thermosets degraded completely within 48 hours under alkaline conditions, while remaining stable under near-physiological conditions for 14 days, consistent with a surface-initiated erosion mechanism.

These findings establish a structure-property framework linking lignin molecular architecture to modification efficiency, thermoset performance, and controlled end-of-life degradation.

Hard carbon (HC) is currently one of the most promising anode materials for sodium-ion batteries (SIBs). For commercial applications, HC anodes require a high initial Coulombic efficiency (ICE) and a high reversible capacity, which are closely related to a low open-pore volume and a high closed-pore volume in the HC structure. In parallel, HC requires production routes with lower energy consumption and smaller environmental footprints in order to support a circular economy. However, current HC production predominantly relies on conventional resistance-heated carbonization, in which heat is supplied externally and transferred indirectly to the material. This leads to high energy consumption and typically requires extreme carbonization temperatures to induce closed-pore formation. Therefore, additional chemical treatments or post-modification processes are often required, further increasing energy demand and environmental impact. These limitations collectively restrict the scalable and sustainable application of HC anodes.

This thesis aims to develop feasible and energy-efficient modification and carbonization processes for HC production to enhance material resource circularity. Accordingly, fundamental studies combining data-driven modeling, laboratory-scale experiments, and process simulations are conducted. The thesis is based on four different studies that together establish a systematic understanding of the process–structure–performance relationships in biomass-derived HC and propose an energy-efficient pore-engineering strategy enabled by bio-oil modification and induction heating carbonization (IC).

First, through comprehensive literature data analysis and machine-learning modeling, carbonization temperature and HC structure are identified as the dominant factors for HC electrochemical performance. Low open-pore structures and surface defect densities are found to be critical for achieving high ICE, while large interlayer spacing and closed-pore volumes are beneficial for a high plateau capacity. Importantly, the analysis reveals the limitation of conventional carbonization in simultaneously achieving low open porosity and promoting closed-pore formation.

To overcome these limitations, a sustainable bio-oil surface engineering strategy is developed to suppress open pores and surface defects. This approach reduces the specific surface area of HC from 28 to 9 m²/g and increases the ICE from 84.4% to 89.9%. By combining bio-oil surface engineering, a novel IC route is further developed. IC enables direct volumetric heating through eddy currents, which simultaneously minimizes open porosities and promotes closed-pore formation. As a result, optimal IC-derived HC exhibits open-pore volume below 0.003 cm³/g and closed-pore volume of up to 0.23 cm³/g, with an ultra-high ICE value exceeding 95%, and plateau capacity above 260 mAh/g.

Finally, energy analysis and life cycle assessment demonstrate that IC reduces the carbonization energy consumption by approximately 60% and lowers overall environmental impacts by approximately 35% compared with conventional routes. Overall, this thesis demonstrates that the combination of bio-oil modification and IC provides an energy-efficient and low-carbon pathway for producing high-performance biomass-derived HC anodes, supporting the sustainable development of next-generation sodium-ion batteries.

The recycling and reuse of post-consumer resins remain challenging due to complex mixtures of additives and non-intentionally added substances (NIAS) introduced or formed during use and recycling. This thesis combines mass spectrometry-based method development for plastic additives with the safety assessment of recycled polyethylene terephthalate (PET) used for food contact materials (FCM).

Liquid chromatography mass spectrometry was used to investigate the solvent-associated species of a benzofuranone antioxidant (Paper I). Their formation influenced chromatographic retention, ion abundance, and fragmentation, particularly in positive ion mode. Negative ion mode provided more structure-dependent and interpretable fragmentation, demonstrating its value for complex additive analysis.

Characteristic ions in tandem mass spectroscopy enabled the identification of hindered amine light stabilizers (HALS)-related species, although their low sensitivity limited direct application in quantitative analysis (Paper II). Combining precursor ions and base-peak product ions improved the estimation of total HALS content in polymers.

The effects of recycled content, processing history, and PET morphology on NIAS profiles, contaminant levels and uptake were investigated (Paper III and IV). Processing and manufacturing influenced volatile NIAS and contaminant levels. Acetaldehyde and benzene were found in PET materials but at low concentrations in simulants after migration tests, suggesting limited health concerns. Challenge tests assess decontamination by exposing PET to surrogates before recycling. Conditions used in challenge tests could alter PET morphology, in turn, affecting contaminant diffusion and complicating estimates of decontamination efficiency.

This thesis shows that analytical conditions, polymer properties, and processing histories are critical for interpreting chemical profiles in recycled plastics, and supports reliable additive and NIAS investigation, and safety evaluation of FCM.

Nuclear energy is expected to play an important role in the transition toward low-carbon energy systems. Light Water Reactors (LWRs) dominate the global nuclear fleet, and their safe, reliable, and long-term operation depends on understanding the interactions between radiation, water chemistry, and structural materials that govern corrosion and material degradation.

This thesis investigates the effects of direct gamma irradiation and gamma-induced radiolysis on polymers, aqueous systems, and metallic alloys relevant to LWR operation. Particular attention is given to hydrogen peroxide and other reactive species produced during water radiolysis and their interactions with materials commonly present in reactor systems and laboratory irradiation experiments. The results demonstrate that material interfaces strongly influence the chemistry of irradiated aqueous systems. Investigations of polymer sealing materials revealed that hydrogen peroxide accumulation depends significantly on material choice, highlighting how experimental design and material selection can affect radiation chemistry measurements and contribute to reproducibility challenges. Studies of reactor-relevant alloys further showed that hydrogen peroxide accumulation depends on alloy composition and is governed primarily by radiation-induced processes in solution. Surface analyses indicated only minor modifications of oxide films during irradiation, suggesting that changes in water chemistry are linked more closely to species released into solution than to extensive alterations of alloy surfaces under the investigated conditions. In situ electrochemical measurements under intermittent gamma irradiation revealed an immediate dose-rate-dependent response associated with radiolytically generated reactive species. The results distinguished the roles of short-lived radicals and longer-lived oxidants, identifying hydrogen peroxide as the primary driver of the longer-term electrochemical evolution of the investigated alloy surface. Overall, this thesis shows that the behaviour of irradiated aqueous systems is governed by the interplay between radiation, water chemistry, and material interfaces. The findings contribute to a deeper understanding of radiation chemistry and corrosion-related phenomena in LWR environments and provide guidance for the design and interpretation of future irradiation experiments.

Integrated sensing and communication (ISAC) has emerged as a key paradigm for future wireless networks, enabling communication infrastructures to support data transmission and environmental sensing within a unified framework. This integration, however, introduces new challenges in deployment, signal processing, and resource allocation. Cell-free massive multiple-input multiple-output (CF-mMIMO) networks, characterized by a large number of distributed access points (APs), provide a promising platform for ISAC. Through coordinated operation among distributed APs, CF-mMIMO systems can support flexible bi-static and multi-static sensing configurations, thereby avoiding the full-duplex requirement of conventional mono-static sensing systems. Moreover, the distributed AP architecture provides spatial diversity and multiplexing gains, making CF-mMIMO well-suited for advanced sensing and communication services.

Despite these advantages, integrating sensing into CF-mMIMO networks increases overall network power consumption and imposes additional demands on radio, fronthaul, and cloud-processing resources. Therefore, the effective realization of green CF-mMIMO ISAC requires joint system design and resource allocation frameworks that account for both sensing and communication requirements.

This thesis studies ISAC in CF-mMIMO systems, with a focus on efficient resource allocation, reliable sensing and communication, and end-to-end network power consumption. The main objective is to develop green CF-mMIMO ISAC frameworks that jointly design sensing and communication functionalities while addressing reliability, energy-efficiency, and scalability challenges.

The thesis first investigates power allocation for target detection in CF-mMIMO systems. By exploiting both communication signals and dedicated sensing signals, the work characterizes the trade-off between sensing and communication performance. Maximum a posteriori ratio test (MAPRT)-based detectors are developed to enable reliable target detection from signals received at distributed APs under both clutter-free and cluttered sensing environments.

Building on this foundation, the thesis extends the analysis to ultra-reliable low-latency communication (URLLC) scenarios, where sensing information is used to support target-aware actuation use cases. In such scenarios, sensing information must be delivered reliably and within stringent latency constraints. A joint power and blocklength optimization framework is proposed to minimize energy consumption across the radio and cloud domains. The results characterize the interplay among sensing performance, communication reliability, latency, and processing workload, highlighting the importance of jointly optimizing system parameters under strict quality-of-service requirements.

A central contribution of the thesis is the development of end-to-end network power models for CF-mMIMO ISAC systems. Unlike conventional approaches that focus primarily on transmit power, the proposed models incorporate radio, fronthaul, and cloud-processing power consumption. This enables a more comprehensive evaluation of energy efficiency in ISAC networks and reveals the impact of sensing-related processing and signaling overhead on the total network power consumption.

To address scalability challenges in large-scale deployments, the thesis further investigates distributed sensing architectures. Two sensing-information levels, namely fully-informed and partially-informed systems, are considered to capture different trade-offs among sensing accuracy, computational complexity, and fronthaul signaling overhead. A cross-layer optimization framework is then developed to jointly manage radio, fronthaul, and cloud resources, substantially reducing total network power consumption while maintaining reliable sensing and communication performance.

Overall, the results demonstrate that CF-mMIMO is a promising architecture for green ISAC by enabling flexible resource allocation, scalable sensing architectures, and significant energy savings. The proposed methods provide practical insights into the design of next-generation wireless networks that integrate sensing and communication in an energy-efficient and scalable manner.

Marine vessels operate as moving work environments, and vessel motions can affect the safe and effective performance of onboard tasks. By influencing crew balance, perception, control, and the handling of materials and equipment, vessel motions can compromise safety and operational performance. Such limitations should be addressed early in vessel design and procurement through seakeeping evaluation. However, conventional seakeeping evaluation practices often rely on generic motion criteria that do not adequately capture the relationship between vessel motions, onboard task demands, and crew performance. 

The aim of this thesis is to develop a systematic approach to seakeeping evaluation that aligns seakeeping performance with functional requirements, onboard tasks, and end-user needs. The research combines literature reviews, semi-structured interviews, focus group discussions, onboard observations, and numerical simulations. It identifies key limitations in conventional seakeeping evaluation practices and develops a systems-oriented, vessel-specific, and task-based framework for seakeeping evaluation. The framework is demonstrated through two case studies of the Swedish Coast Guard vessel KBV 202 and the Swedish Maritime Administration fairway maintenance vessel Fyrbjörn, showing that task-based evaluation yields substantially different operability assessments from conventional generic approaches. A task-informed motion-induced interruption estimator is also developed, enabling the prediction of interruption rates without prescribed stance geometry or orientation.

The findings show that seakeeping evaluation should extend beyond routine generic procedures towards vessel-specific and task-based approaches that better support the safe and effective performance of onboard work. The proposed framework provides a structured and defensible basis for translating functional requirements and end-user needs into relevant seakeeping evaluation criteria.

With the rapid development of wireless communication systems driven by the demand for higher data rates and sufficient coverage, a factor that may limit communication performance is the compliance requirements for radio frequency (RF) electromagnetic field (EMF) exposure. This thesis presents a holistic study on the EMF exposure in modern antenna systems, facilitating improvements in RF device communication performance while fulfilling EMF exposure compliance. The thesis provides comprehensive studies from an evaluation process to the EMF solutions for sub-6 GHz indoor base stations, and early research for Frequency Range 2 (FR2) and FR3 RF equipment.

For sub-6 GHz indoor base stations (BSs), two main contributions are made. First, a holistic evaluation framework is established to assess power-related multiplexing efficiency, considering antenna radiation characteristics, indoor propagation scenarios, and the power reduction resulting from ensuring EMF compliance at any distance from the device (i.e. touch compliance). The proposed evaluation framework gives antenna designers a valuable basis for comparing and optimizing MIMO antenna systems while considering specific absorption rate (SAR) touch compliance and various deployment scenarios for indoor BSs. Second, passive and active EMF solutions are developed to achieve improved communication performance while ensuring EMF touch compliance. Regarding passive designs, two antenna are proposed, including a dedicated monopole antenna and a patch antenna. These designs spread the SAR distributions and thus lower the peak SAR levels, achieving touch compliance for indoor BSs without power reduction. In addition, a novel active EMF solution is proposed by reusing the communication antenna as a proximity sensor. By detecting human proximity through variations in the antenna's reflection coefficient, the BS can maintain high transmission power during normal operation and only trigger power back-off mechanisms when a human body enters the EMF exclusion zone.

As 6G systems can shift toward FR2 and FR3 bands, where the exposure metrics are absorbed power density (APD) and incident power density (IPD), three further contributions are made. First, a comprehensive review of state-of-the-art IPD and APD assessment methodologies is conducted. It aims to identify open challenges and potential future directions for accurate assessment of EMF exposure. Second, a novel metasurface-based conformal human phantom is proposed, with the potential to serve as a new test equipment for EMF assessment. Finally, an analysis of the implications of APD limits on 6G user equipment (UE) is conducted, establishing system design benchmarks, such as maximum allowed transmitted power and equivalent isotropically radiated power (EIRP) levels, for both single- and dual-antenna systems within the 6-15 GHz spectrum.

In conclusion, the contributions of this thesis provide a set of tools related to EMF exposure research. The work also helps the future FR2/3 device development and EMF assessment equipment.

Tumors are highly heterogeneous populations of cells that evolve dynamically, acquiring mutations as they divide. Reconstructing this evolutionary history is essential for understanding cancer progression, metastasis, and therapy resistance. With the advent of single-cell DNA sequencing (scDNA-seq), we can now profile the genomic landscape of individual cells, which provides an unprecedented window into the formation and evolution of tumor cell populations. Specifically,single-cell whole-genome sequencing allows us to detect structural variations called \textit{copy number aberrations} (CNAs), which are known to play a critical role in cancer development and progression. Inferring the evolutionary trees, i.e., phylogenies, from single-cell DNA sequences presents immense computational challenges due to both the sparsity of the data inherent to the current state of sequencing technologies and the complexity of the underlying mutational processes.

In this thesis, we develop novel probabilistic models and algorithms to infer the evolutionary history of tumors from single-cell DNA sequencing data. The research addresses core methodological bottlenecks in tumor phylogenetics, moving from foundational distance-based methods to comprehensive joint Bayesian inference frameworks.

First, we introduce a method to estimate biologically meaningful evolutionary distances between single cells directly from noisy read counts, employing an original Hidden Markov Model to accommodate the unique noise profile of scDNA-seq and the interdependence of copy number states across the genome. Next, we extend distance-based tree inference from classical phylogenetics by presenting a scalable algorithm specifically designed for rooted trees, leveraging the biological premise that tumor evolution originates from a known healthy diploid ancestor. To enable rigorous uncertainty quantification over tree topologies, we then tackle the problem of sampling directed trees (arborescences). We present a stable, polynomial-time sampling algorithm capable of generating arborescences even on weakly connected graphs, which commonly arise when performing inference from single-cell sequences. Finally, we integrate these advancements into a comprehensive variational inference framework. This framework efficiently achieves joint inference over clonal tree structures, branch lengths, copy number profiles, and cell-to-clone assignments.

Collectively, this thesis contributes a suite of statistically grounded, highly scalable tools that bridge the gap between noisy single-cell sequencing reads and robust insights into cancer evolution, offering a foundation for future clinical applications and oncological research.

Artificial intelligence has become part of higher education, yet its future role remains unsettled. Its growing presence raises questions about what counts as learning, how assessment can remain meaningful, what teachers and students should be expected to do, and which forms of educational life should be protected. This thesis examines these questions through education fiction: short speculative stories written within or about (future) educational settings. The thesis uses education fiction to study how AI futures are imagined, contested, and made meaningful by researchers and higher education stakeholders.

This research is grounded in a social constructivist understanding of futures as shaped through social interactions, cultural narratives, and shared expectations. Drawing on the theory of Social Construction of Technology (SCOT), it approaches the current moment of AI in education as one with high interpretative flexibility, where multiple and competing meanings of AI coexist. Fiction is positioned as a distinctive site for studying this process because stories give form to possible futures through conflicts, metaphors, omissions, and other narrative choices.

The thesis comprises four interrelated papers. Paper 1 reviews 100 published speculative fictions to map recently published visions of AI futures in education. Paper 2 develops a three-lens analytical framework for analysing education fiction by integrating literary analysis with educational research. Paper 3 uses a web-based story-crafting tool to elicit original education fictions from 69 stakeholders in Swedish higher education, generating empirical data about their hopes, concerns, and values regarding AI futures. Paper 4 deepens this analysis through semi-structured interviews with 17 of these participants, examining how and why their anticipations of AI futures are both generative and constrained. 

Across the studies, the thesis shows that AI futures in education are plural yet patterned. Stakeholder-written fictions clustered into four configurations: Enhancement, Transformation, Displacement, and Resistance. These configurations were structured by three recurring tensions: the human remainder, the assessment paradox, and the efficiency-depth trade-off. The thesis also develops the concept of bounded anticipation to explain why many imagined futures preserved familiar educational structures, roles, and relationships. These limits are interpreted as meaningful expressions of the educational values stakeholders sought to protect, even as they may narrow what becomes thinkable.

The thesis makes three types of contributions. Methodologically, it advances education fiction as a research approach and offers analytical tools for studying fiction as data, method, and narrative knowledge. Theoretically, it shows how education fiction participates in the social construction of AI futures and extends SCOT by highlighting the narrative and value-bounded character of interpretative flexibility. Practically, it provides a vocabulary for more values-explicit conversations about AI in higher education, showing how educators, students, researchers, institutional leaders, and designers may negotiate competing visions of technological change.

Modern datasets are increasingly complex and heterogeneous. Analyzing such data raises fundamental questions about the mathematical spaces in which data should be represented, how objects in these spaces should be compared, and which computable invariants preserve information relevant to a given task. This thesis studies these questions with topological data analysis as a central framework, using homology as a language for describing the geometry of data.

The first part of the thesis develops distances and invariants for spaces of persistence modules. The starting point is categorical: we regard data as objects living in categories equipped with enough algebraic structure to define a notion of size. In particular, abelian categories provide kernels, cokernels, and exact sequences, allowing distances between objects to be constructed from the failure of morphisms to be isomorphisms. Persistence modules form a central example: they encode how homological features appear and disappear along a filtration, and may be viewed as functors from a partially ordered parameter space to vector spaces. Within this framework, we study distances induced by contours, which provide a flexible way of specifying the geometry of the parameter space. This setting leads to compactness results for families of multidimensional persistence modules. In the one-dimensional setting, where a barcode decomposition is available, we develop algebraic Wasserstein distances based on ℓ_p norms of contour-dependent bar lifetimes. Using these distances, we define Wasserstein stable ranks, stable and computable invariants whose interpretable parameters can be learned for a given task.

The second part of the thesis moves from the mathematical framework to applications in neuroscience, where cellular morphologies provide natural examples of structured geometric data. Microglia and other branched cells can be represented as rooted trees embedded in three-dimensional space, and their morphology can be characterized using topological morphology descriptors. In the morphOMICs pipeline, such descriptors are combined with vectorizations, bootstrapping, dimensionality reduction, and classification in order to map microglial morphology across brain regions and sexes, and through development, disease progression, and experimental perturbations. This gives a data-driven atlas of microglial morphology that avoids relying on preselected scalar morphometric features.

We further introduce the chromatic topological morphology descriptor (chromatic TMD) to study intracellular organization in branched cells. Here a microglial cell is represented by a rooted tree, while CD68-positive and mitochondria organelles are represented by subgraphs of that tree. The inclusion of the organelle subgraph into the cell tree induces a morphism of persistence modules, and the image, kernel, and cokernel of this morphism describe complementary aspects of organelle organization: where organelles occupy branches, where they co-localize within branch structures, and where they are absent. An efficient tree-based algorithm is developed for computing these descriptors. Applied to retinal microglia, the method reveals organelle-specific spatial programs: CD68-positive organelles reorganize in a layer- and injury-dependent manner, while mitochondrial organization remains more closely coupled to the underlying branching morphology.

The third part of the thesis studies how stable homological invariants can be used in machine learning. Stable ranks provide a bridge from persistence modules to function spaces or finite-dimensional vector spaces, making persistence-based information accessible to kernel methods and neural networks. We introduce stable rank kernels, in which the choice of distance on persistence modules determines the stable rank and, consequently, the similarities encoded by the kernel. Varying this distance through contours can improve supervised learning performance. We also study subsampling-based stable ranks, in which probability distributions on a reference dataset are used to draw many subsamples, compute persistent homology and the corresponding stable ranks, and average the resulting functions. Different choices of distribution yield global descriptors of datasets or relative descriptors of points in the ambient space with respect to a reference object.

Finally, we investigate robustness in persistence-based learning. Persistent homology is stable with respect to suitable metrics, but these guarantees need not be preserved when persistence modules are processed by neural networks. We therefore introduce a stable rank network, combining stable rank vectorizations with Lipschitz neural network layers. This architecture has a controlled Lipschitz constant and yields sample-wise certificates of robustness in Wasserstein or bottleneck distance. This shows that topological stability can be preserved through a learning pipeline and used to certify robustness against adversarial perturbations.

Quantitative computed tomography (QCT) enables objective assessment of lung structure and may provide information complementary to spirometry in chronic airflow limitation (CAL). Inspiratory and expiratory chest CT can assess lung volume, lung density, low-attenuation volume, and ventilation-related changes. However, their role in population-based cohorts remains less extensively studied, particularly in individuals without established lung disease or with mild airflow limitation. Furthermore, regional analysis requires accurate lung lobe segmentation, and the value of radiomic and deep learning-derived features for identifying CAL remains incompletely understood.

The overall aim of this thesis was to extract, analyse, automate, and regionalise quantitative measures from inspiratory and expiratory chest CT images, and to study their relationship with normal lung structure, ventilation-related changes, and spirometry-defined CAL. Data were obtained from the Stockholm cohort of the Swedish CArdioPulmonary bioImage Study (SCAPIS), a population-based cohort of men and women aged 50–64 years. Inspiratory and expiratory chest CT, post-bronchodilator spirometry, and questionnaire data were used. CAL was defined as a post-bronchodilator FEV1/FVC ratio < 0.70.

Study I evaluated global inspiratory and expiratory CT-derived lung volumes, mean lung density, low-attenuation volume, and density gradients. Inspiratory CT lung volumes were lower than literature-based total lung capacity reference values, whereas expiratory CT lung volumes exceeded residual volume reference values. Participants with CAL had higher inspiratory and expiratory lung volumes, lower mean lung density, and greater low-attenuation volume than participants without CAL. Expiratory CT measures showed better discriminatory performance than inspiratory measures, with the highest performance observed for expiratory low-attenuation measures. A dorsal–ventral attenuation gradient was observed during expiration in participants without CAL but not in participants with CAL.

Study II trained and evaluated a deep learning-based lung lobe segmentation method incorporating prior anatomical information from lung vessel connectivity. The method was evaluated in both inspiratory and expiratory CT, including expiratory scans, in which segmentation is typically more demanding. Prior anatomical information mainly improved boundary accuracy, with the clearest benefit in expiratory CT. The best overall segmentation performance was achieved using a multitask model segmenting both lung lobes and fissures. Complementary downstream analyses showed that the choice of segmentation method had a limited effect on CAL discrimination in the full cohort, although variability was greater in smaller cohorts.

Study III extended the global analyses in Study I to lobar CT measures in a larger cohort. Participants with CAL had higher inspiratory and expiratory lung volumes, higher low-attenuation volumes, and more negative mean lung density than participants without CAL. Expiratory measures again showed stronger discriminatory ability than inspiratory measures. Lobar analyses demonstrated regional heterogeneity in lung volume, density, and low-attenuation measures, while complementary thesis analyses examined dorsal–ventral variation. In regularised logistic regression models, combined inspiratory–expiratory measures contributed to CAL discrimination; performance improved when expiratory measures were added to inspiratory measures and showed a slight additional improvement with lobar variables. The best Study III model achieved an area under the ROC curve (AUC) of 0.81.

Study IV assessed whether lobe-level radiomic and deep learning-derived features improved CAL discrimination beyond established quantitative CT measures. Classical radiomics, SegResNet, and 3D U-Net feature models were compared using inspiratory, expiratory, and combined inspiratory–expiratory data. Expiratory feature models consistently outperformed inspiratory models. The best performance was achieved by the classical radiomics model combining inspiratory and expiratory features, which outperformed the handcrafted quantitative CT model from Study III. Feature selection showed that classical radiomics relied mainly on expiratory texture features, whereas deep learning-derived models showed more balanced contributions from inspiratory and expiratory images. Combining radiomic, deep learning-derived, and handcrafted feature sets did not improve performance beyond the best individual model.

In conclusion, inspiratory and expiratory chest CT provide quantitative measures associated with spirometry-defined CAL in a population-based cohort. Expiratory CT was consistently more informative for discriminating CAL than inspiratory CT, supporting its value for assessing ventilation-related abnormalities and CAL-related lung changes relevant to early COPD. However, regression analyses showed that combinations of inspiratory and expiratory measures contributed to CAL classification, suggesting that the two respiratory phases provided complementary information. Automated lung lobe segmentation enabled scalable regional analysis, while lobar and radiomic approaches added information beyond whole-lung measures. Together, the findings support QCT as a complementary imaging-based approach for describing lung structure, regional heterogeneity, and CAL-related abnormalities in population-based research.

Several genetic diseases can now be treated using gene therapy. Adeno-associated viruses (AAVs) are widely used in these types of therapies as gene delivery vehicles, where the AAV genome is replaced with a therapeutic gene. However, AAV gene therapies are expensive due to challenging recombinant production. Additionally, AAVs are not organ-specific, increasing the dose required to reach efficacy. This drives up drug prices and risks of side effects, especially towards the liver. This thesis presents methods to improve AAV production and to make AAVs organ-specific.

AAV production requires plenty of plasmid DNA and in drug development phases of AAV-therapies, plasmid purification is highly manual. In the first article, automated plasmid purification is compared to a manual, industry-standard kit. The automated system lowered manual labour by up to 95 %, with similar or up to 14-fold higher AAV titres depending on the AAV production system. Transfection parameters were studied, showing larger transfection complex formation when using automatically purified plasmids.

The next two articles describe the main work of the thesis: the development of a modular platform for organ-selective AAVs. By attaching a type of affinity scaffold protein called Affibody molecules to the viruses, the AAVs became infectious only towards cells displaying the target receptor of the Affibody. The Affibody-AAVs demonstrated selective cell uptake in vitro and modified organ uptake in vivo, avoiding the liver. Cryo-electron microscopy (Cryo-EM) studies of Affibody-AAVs yielded a high-resolution protein structure at 1.67 Å, along with the first structural data of an affinity scaffold integrated into an AAV.

In the final paper, AAV-production was systematically compared between two common AAV production cell lines: adherent HEK293T and suspension HEK293F. HEK293T produced higher AAV titres than HEK293F (3 to 633- fold higher) and viruses secreted from HEK293T up to 316 times more. By creating AAV-hybrids, secretion was increased from HEK293F. Cell expression level studies of heparan sulphate proteoglycans (HSPGs), targeted by the AAV-variant AAV2, showed lowering of HSPG levels may improve AAV2 secretion.

In summary, this thesis describes methods for creating more efficient AAV gene therapies and for improving AAV production. The presented Affibody-AAV platform can also broaden the use of AAV gene therapies, by enabling targeting of AAVs to organs which naturally-occurring AAVs do not reach.

This thesis investigates the numerical modeling and design of gradient-index (GRIN) lens antennas and radome-integrated phased array antennas (PAAs) at millimeter-wave frequencies. The research encompasses the analysis of periodic structures, the development of efficient ray-tracing and physical-optics (RT-PO) tools, and the design of high-performance GRIN lens and radome antennas.

First, a comprehensive methodology is established to obtain the dispersion diagrams of structures periodic in all three spatial dimensions and characterized by various lattice arrangements. Analysis shows that symmetries in body-centered and face-centered cubic lattices lead to improved isotropy and maintain low-dispersion characteristics over a wider frequency range than conventional simple cubic lattices. These properties facilitate the implementation of high-performance GRIN lenses based on suitable lattice arrangements, as demonstrated through fully metallic Luneburg lens and dielectric truncated virtual image lens designs.

Second, an efficient RT-PO model is developed to provide a rapid tool for analyzing lens radiation properties. This framework applies geometric optics to calculate ray trajectories and determine the corresponding phase distributions, followed by amplitude evaluation via ray-tube power conservation theory and radiation-field computation using the field equivalence principle. For azimuthally symmetric three-dimensional GRIN lenses with directive feeds for collimation or focusing, RT is reduced to a two-dimensional (2D) axial cross section, while PO is performed on a circular integration surface at the lens output. If the field on this surface is uniformly linearly polarized, the PO formulation is further simplified from vector to scalar integration. This approach enables accurate evaluation of both quasi-nondiffracting near-field and collimated far-field distributions, with computation times significantly shorter than those required by commercial full-wave simulators.

Lastly, this research explores lens-based radomes integrated with PAAs to enhance directivity within the desired scanning range without modifying the existing arrays. The initial dielectric lens (DL) designs are based on full-wave simulations; however, the high computational cost of these simulations motivates the development of efficient RT-PO models for radome analysis and design. Using the 2D RT-PO model, a DL is designed and validated through prototype measurements, confirming improved gain and beam-pointing accuracy for a realistic PAA at wide scanning angles. Beyond standalone DLs, a hybrid dome combining DLs and metadomes (MDs) is introduced to exploit the benefits of both structures, achieving lower weight and volume than standalone DLs and a broader bandwidth than standalone MDs. The capability of the hybrid dome to extend the PAA scanning range is demonstrated in both single-beam scanning scenarios and steerable flat-top beam applications.

This thesis studies computational optimal transport methods for modelling, estimation and control problems in networked systems. The focus is on scenarios where the available information is incomplete, aggregated, or constrained by the physical structure of the network. Such settings arise naturally in critical infrastructure systems, including water distribution networks, where contaminant flows must be inferred from sparse measurements, and transportation networks, where vehicles or agents must be steered under physical constraints. The thesis formulates these problems using multi-marginal entropy-regularized optimal transport and Schrödinger bridge methods, taking into consideration both theoretical and computational aspects, developing algorithms based on Sinkhorn-type iterations and entropic proximal schemes.

The first paper considers the problem of estimating the spread of contaminants in water distribution networks from sparse sensor measurements. The water flow is modelled as a time-varying Markov chain, and the pollutant evolution is recovered as a Schrödinger bridge problem with partial marginal observations. A dual formulation and a Sinkhorn-type algorithm are derived, and the method is illustrated on simulated water-network data.

The second paper extends this formulation to the case where the first marginal is also only partially observed, corresponding to an unknown contamination source. This leads to an incomplete-information problem with unknown total mass and possible non-uniqueness. The paper characterizes the optimal solution set in terms of observability of an associated time-varying linear system, and proposes an algorithm that combines an entropic proximal scheme with Sinkhorn-type iterations. The method is validated on experimental data collected at a water distribution laboratory.

The third paper studies the steering of electric fleets over networks with origin-destination, battery-charge, and capacity constraints. By augmenting the physical network with discrete charge states, the routing problem is formulated as a structured multi-marginal optimal transport problem. A dual coordinate ascent algorithm is developed, exploiting the structure and sparsity of the problem, and the method is demonstrated with numerical simulations on a grid network.

The fourth paper addresses the estimation of Markov transition matrices from aggregate observations of indistinguishable particles. The problem is formulated as a convex inverse optimal transport problem, where transport plans and the transition matrix are estimated jointly. The paper provides existence, uniqueness and duality results, proposing an entropic proximal algorithm for computing the solution. Numerical experiments show that the method can recover the underlying dynamics when the observations sufficiently excite the state space.

The brain enables organisms to perceive the world, learn from experience, generate complex behavior, and give rise to cognition. Understanding the information processing principles underlying brain computation remains a key challenge in computational neuroscience and cognitive science. Elucidating these principles has the potential to provide a foundation for developing intelligent machines and energy-efficient, scalable, and robust artificial intelligence paradigms.

This thesis investigates neural network models incorporating key brain-like design principles, focusing on unsupervised representation learning and the formation of robust associative memory. While modern deep learning approaches achieve strong performance in representation learning tasks, they rely on backpropagation-based optimization, which lacks biological plausibility and depends on globally coordinated computations. There remains a need for neural network models that can demonstrate complex functionalities while remaining grounded in biological principles.

To address this gap, this thesis develops a class of brain-like models based on the Bayesian Confidence Propagation Neural Network (BCPNN) framework. The proposed models incorporate key brain-like design principles, including localized Hebbian synaptic plasticity, structural plasticity, activity and connection sparsity, and a modular architecture derived from neocortical columnar organization. Furthermore, the models integrate feedforward, recurrent, and feedback connectivity to support both representation learning and associative memory.

This work comprises three main lines of investigation. First, the unsupervised representation learning capabilities of a feedforward model are examined, demonstrating that structured internal representations can be learned directly from unlabeled data using localized synaptic and structural plasticity. Second, the formation of associative memory is studied by integrating recurrent connectivity in the representation learning model, demonstrating robust recall from partial, noisy, or corrupted inputs when tested on pattern completion, prototype extraction, and noise robustness tasks. Third, the framework is extended to spiking neural networks, where neurons communicate via stochastic spike events, and the results show that, through synaptic short-term filtering and appropriate temporal scaling, the spiking models approximate the behavior of their rate-based counterparts while preserving functionality and performance. 

Overall, the results establish that brain-like design principles can support scalable representation learning and robust associative memory, demonstrating a bridge between neuroscience and artificial intelligence within the emerging field of NeuroAI. This work further offers a pathway toward energy-efficient, brain-like neuromorphic systems capable of operating in real-time and in dynamic environments.

This dissertation investigates the emulation of complex wave dynamics, topologicalphases, and non-Hermitian systems using integrated photonic circuits,addressing the limitations of traditional static and energy-conservingoptical systems through passive and reconfigurable architectures.The experimental foundation relies on two complementary platforms: passivesilicon nitride circuits for simulating static spatial propagation, and programmablesilicon-on-insulator Mach-Zehnder interferometer (MZI) meshes.Utilizing singular value decomposition, the active meshes synthesize discretetimeevolution operators to dynamically tune coupling amplitudes and onsitepotentials.Within the Hermitian framework, the passive circuits provide observationalevidence of the Topological Anderson Phase and a reentrant metal-insulatortransition driven by spatially correlated disorder. Additionally, the activeMZI mesh is employed to emulate the Bloch-Siegert shift. By mapping thediscrete time evolution of a driven two-level system onto spatial propagation,the system demonstrates resonance jumps and the conversion of bidirectionalRabi oscillations into unidirectional transport.The experimental scope is subsequently expanded into the non-Hermitianregime. Employing the unitary dilation method, ancillary modes on the programmablemesh introduce controlled dissipation to synthesize non-unitaryoperators. This approach enables the realization of a non-Hermitian Hamiltonianon a Klein bottle parameter space, yielding experimental signaturesconsistent with paired exceptional points. Furthermore, implementing rowstochasticMarkov matrices demonstrates dissipation-induced multimode phasesynchronization, with synchronization rates governed by the spectral gapand operating independently of absolute optical attenuation.In summary, this thesis explores Hamiltonian engineering on integrated photonicplatforms as a method to simulate topological and non-Hermitian phenomena.The investigated architectures and dissipative mechanisms offera framework for studying fundamental wave physics, suggesting potential pathways for future applications in exceptional-point-enhanced optical neuralnetworks and high-bandwidth optical interconnects.

Recent advances in microfluidics and optical fiber technologies have enabled the development of miniaturized platforms for investigating, diagnosing, and treating disease. By precisely manipulating fluids and biological components at small scales, it is possible to recreate complex biological environments, while providing improved control over parameters such as flow conditions, chemical gradients, and cellular interactions. Microfluidic organ-on-chip systems provide in vitro environments that recapitulate key aspects of human physiology and pathology, offering opportunities to improve disease modeling and drug development. Similarly, optical fiber-based platforms provide compact platforms for improved biosensing, while point-of-care microfluidic technologies enable rapid and cost-effective diagnostic analysis. The integration of these technologies offers new opportunities for developing multifunctional systems that combine complex analytical capabilities with minimally invasive operation.

This thesis explores the use of microfluidic and optical fiber technologies for applications in pancreatic ductal adenocarcinoma (PDAC), the most common form of pancreatic cancer. The work focuses on two complementary technological approaches: Lab-on-Chip platforms for disease modelling and molecular diagnostics, and Lab-in-a-Fiber platforms for targeted cell detection, capture, and analysis.

First, a microfluidic pancreatic tumor-on-chip model was developed to reconstruct the desmoplastic organization of PDAC. In parallel, multifunctional fiber-based platforms were developed for the detection and capture of specific cell populations, spatial identification of target cells, and subsequent analysis. These approaches were further extended to fiber-based flow cytometry, providing a pathway towards integrated analysis of captured cells and other biological targets. Finally, a complementary point-of-care microfluidic platform was developed for sensitive nucleic-acid detection and genotyping, paving the way for decentralized molecular diagnostics.

Together, these platforms provide a technological foundation for studying complex tumor–microenvironment interactions and for exploring less invasive approaches to diagnosis and treatment. Ultimately, the combination of these technologies aims to enable improved disease characterization and treatment.

Spatial biology has attracted growing interest because it measures two types of information at the same time: which molecules are present in a tissue and where they are located. Both are needed for understanding biology, as the organization of tissues, cells, and molecules drives biological processes in health and disease. Most leading spatial technologies recover this spatial organization from an external reference, either by imaging the tissue directly with optics or by using a substrate containing barcoded coordinates that were decoded in advance. Sequencing-based microscopy, the approach studied in this thesis, instead obtains spatial information by sequencing the interactions inside the sample. DNA-barcoded molecules interact with their neighbors, sequencing reads the corresponding interaction products,  and a reconstruction algorithm recovers coordinates from the resulting network, much like assembling a jigsaw puzzle. Thus, the spatial organization of a tissue can be recovered without optics and without a predefined spatial reference. This is particularly attractive as a natural solution to three-dimensional imaging. However, a sequencing-based microscopy experiment produces a network rather than an image, and how to reconstruct coordinates from that network is the computational problem at the center of this thesis.

Article I addresses part of this problem, which is how to recover coordinates when the interaction rule is unknown and the graph is large. We developed STRND, a graph reconstruction method that uses random walks and manifold learning to reconstruct images from proximity networks without assuming how physical distance determines edge formation. Its computational complexity is approximately linear in the number of  nodes, which allowed reconstruction of large graphs.

Article II addresses how to assess and denoise a DNA barcode network without reference ground-truth coordinates. We introduced spatial coherence, which uses three topology-based metrics to evaluate if shortest-path distances in a network follow the geometric properties expected from physical distances. The metrics are able to detect distortions in topology caused by false interactions, and also provided an objective function to guide denoising.

Article III applies these tools to an existing spatial transcriptomics method that was initially not meant for sequencing-based microscopy. This method, Slide-tags, diffuses bead-carried DNA barcodes into tissue sections and normally assigns cell positions using an optically decoded bead map. However, we found that the sequencing data also contained a cell-bead proximity network produced via diffusion during the experiment, and that this network preserved enough spatial information for a reconstruction that only uses interaction data. Applying STRND to a human tonsil dataset recovered tissue coordinates without the optical decoding step, including approximately 2,000 cells that the original analysis had excluded.

Article IV reexamines the assumption that spatial reconstruction requires networks dominated by short-range interactions. Longer-range interactions are usually treated as noise, but we found that densely connected networks could still be reconstructed accurately even after individual hop distances collapsed into a few discrete values. The reason is that each node had a distinct shortest-path profile that changed smoothly with position, thus containing spatial information. Proximity interactions are therefore sufficient for spatial reconstruction, but they are not always necessary. These results suggest that increased network connectivity can be beneficial for reconstructions under certain regimes.

Article V extends the sequencing-based microscopy logic to an application unrelated to spatial imaging, DNA data storage. Write-by-partitioning stores information in the spatial order of a DNA barcode network that forms in a hydrogel before the message is known, like a blank page that can be written on. The message is written by cutting the hydrogel into labeled sections and recovered by reconstructing the network order and detecting the partition boundaries from sequencing data. The writing chemistry is therefore independent of the message content. We showed that this strategy stored and recovered information, and spatial coherence confirmed the expected quasi-one-dimensional structure of the barcode networks.

These five articles show that DNA barcode networks contain spatial and ordinal information. The network structure can be reconstructed, evaluated, and denoised computationally without optical coordinates, and useful geometry is present in a more diverse range of interaction mechanisms and connectivity regimes than what was previously assumed. These results establish DNA barcode networks as a measurable source of spatial information and show that the same reconstruction principles can also support applications outside microscopy.

This dissertation proposes designing with discomfort as a feminist approach for interaction design research on intimate care technologies. Contemporary healthcare technologies increasingly promise empowerment through seamless tracking and monitoring of diverse biomarkers. While new data-driven insights may be unlocked, their interactive experience often instigates users to attribute epistemic authority to the technology over the felt and lived knowing of their body. Furthermore, (re)productivity-oriented inclinations and ableist assumptions embedded in these technologies implicitly promote curative, disembodied, individualistic, and disciplinary relations towards the body. Against this backdrop, uncomfortable bodies that are norm-misfitting, in pain, and disabled become further marginalised. Through four design projects on reproductive and menstrual health, I explore discomfort, pain, and misfitting as generative design sites rather than treating them as noise, pathology, or usability problems. I show how such an approach can engender new design repertoires of intimate care technologies that diversify the kinds of relations we can have towards our own and others’ bodies. 

This work explores what it means to design care technologies with “discomfort” and how we may harness situated knowledge of pain/discomfort. First, I contribute with a feminist somaesthetic notion of discomfort. Grounding in critical disability studies and feminist epistemology, this work conceptualises discomfort both (1) as a situated, everyday, embodied experience that care technologies should better account for and (2) as a feminist epistemic standpoint. The analysis of design projects surfaces four modes of discomfort, namely affective, sociocultural, lived, and evocative, that can be taken by designers and researchers to see, unpack, and design with the complexities and multiplicities of the human body. The second contribution is designing with discomfort as a feminist research and design praxis. Building on the principles of Research-through-Design (RtD) and soma design methodology, I articulate four tactics—narrative, discursive, multi-temporal, and emergent—and methods for designing with different modes of discomfort, generatively and ethically. The third contribution is a feminist crip design space for intimate care technologies vii constituted by design strategies of ambivalence, pain-centred design provocations, and two design exemplars, Touchware and Kuddi. They showcase alternative interactions of technology-facilitated care that promote practices of reparation, rest, and compassion towards imperfect and uncomfortable menstruating bodies. I conclude by discussing the theoretical and material implications of designing with discomfort for the intimate care context and beyond. 

The water and wastewater infrastructure of the built environment is currently operated as a linear system that delivers potable water and removes wastewater, carrying away the energy and resource potential embedded in the streams that pass through buildings. Recovering this potential through integrated water-energy systems is technically feasible, but adoption remains limited by fragmented assessment methods, divergent stakeholder priorities, and institutional structures that do not align with the resource-efficient configurations such systems require. This thesis investigates how water and energy systems in the built environment can be assessed and developed to be more sustainable, energy and resource-efficient, through four interconnected studies.

A comparative characterization of numerous water streams in a typical building was developed, integrating volume, quality, and energy content into a single reference for source-first system selection. Two rounds of semi-structured interviews with thirteen stakeholders spanning regulatory agencies, municipalities, water utilities, property owners, and consultants identified evaluation criteria covering resource efficiency, socioeconomic, and resilience dimensions. These were synthesized into a modular, decision-support framework structured around five stages: conceptualization of the project, description and classification of the site, technical-solution library, system assessment, and multi-criteria decision analysis. The framework was demonstrated through a proof-of-concept application and shown to produce consistent rankings across a wide range of weight configurations.

Two integrated case studies were numerically modelled, with key components validated against pilot installations in Gothenburg. A rainwater-based system combining evaporative cooling with toilet flushing was simulated. Using the suggested rainwater catchment area and tank size, energy savings of 19–28 % in commercial buildings during cooling periods were achieved, with residential savings falling as low as 1–7 % annually with negative energy savings for certain sizes. This was mainly due to limited cooling demand and high parasitic losses. An integrated greywater reuse and heat recovery system at the HSB Living Lab delivered approximately 22–23 % energy savings and 92 % water savings for toilet flushing, with the main trade-off appearing not as competing water- and energy-saving objectives but as a moderate increase in treated-water residence time.

Across the four studies, good technical performance emerges as necessary but is not the only constraint for implementation. Stakeholder priorities, such as environmental, cost and health aspects, are essential, but institutional barriers, such as split incentives, regulatory uncertainty, and the cost-price principle must be aligned to form an optimal environment for system implementation realization. In addition, the decentralization-centralization trade-off for downstream resource recovery has to be considered to avoid suboptimization. Further institutional alignment across climate adaptation, water management, and energy efficiency policy is a precondition for translating technical systems and stakeholder priorities into implementation at scale.

The built environment is a major contributor to global greenhouse gas emissions, resource depletion, and waste generation. Concrete, as the most widely used construction material, has a substantial environmental impact. Reusing structural concrete elements offers a pathway toward circularity by reducing demand for new materials and avoiding upfront CO₂ emissions. However, widespread implementation remains limited by technical challenges, such as remaining service life and the lack of standardized information flows throughout reuse processes. This dissertation investigates the reuse potential of structural precast concrete elements recovered from existing buildings that were not originally designed for disassembly. The research establishes and validates integrated digital workflows that combine engineering assessment, environmental evaluation, and digital information management to support systematic decision-making for concrete reuse in new buildings.

The work first examines the capabilities and limitations of existing digital tools for circular construction before developing BIM-based information management guidelines based on the Level of Information Need (LOIN) framework. To link reclaimed physical elements with their corresponding BIM representations, component tracking workflows are developed, supporting their identification and traceability. The research further demonstrates how openBIM standards, including the Information Delivery Specification(IDS), can enable automated validation of reuse-specific information requirements and improve interoperability within circular construction workflows. The thesis also develops an integrated computational workflow for assessing the technical and environmental viability of reuse by combining service life prediction, carbonation modelling, and embodied carbon assessment across multiple life cycles. The findings show that direct reuse provides substantially greater climate benefits than carbonation alone. A probabilistic performance-based framework is further introduced to assess the service life of reclaimed concrete elements for reuse. It focuses on carbonation-induced corrosion, as a key durability concern in concrete buildings, modelling both corrosion initiation and propagation phases. Through parametric analysis and Monte Carlo simulations, results show that reclaimed precast concrete elements can achieve a second 50-year service life, under appropriate exposure conditions and repair strategies. This research provides methods to support the safe, scalable, and climate-efficient implementation of structural concrete reuse within a circular built environment.