Tuwhera Research Repository

The Tuwhera Research Repository provides sustainable open access and archiving to a broad range of AUT produced research - theses, research outputs and datasets.

Theses & dissertations

Deposit your thesis

Datasets

Deposit your dataset

Research outputs

Deposit your research output
via Research Elements

Recent Submissions

Access status: Open Access ,
SPAR-EEG: Selective Pass-wise Artifact Reduction for Wearable Single-channel EEG Denoising
(IEEE, 2026-09-16) Shaikh, Usman Qamar; Kalra, Anubha Manju; Lowe, Andrew; Niazi, Imran Khan
Single-channel electroencephalography (EEG) is attractive for wearable neurotechnology and brain-computer interface (BCI) applications, including assistive interfaces and clinical monitoring, but artifact suppression is difficult when auxiliary channels, artifact labels, or user-selected clean baseline segments are unavailable. We introduce SPAR-EEG, a self-contained framework for Selective Pass-wise Artifact Reduction in single-channel EEG. The framework applies three artifactspecific attenuation passes to each EEG epoch: a variational mode decomposition (VMD)-based pass for high-frequency electromyo-graphic (EMG) bursts, a singular spectrum analysis (SSA)-based pass for blink-like electrooculographic (EOG) transients, and an SSA-based pass for slow motion-related drift. Rather than rejecting components globally, each pass estimates artifact-dominant regions and attenuation strength directly from the input channel. SPAR-EEG was evaluated using controlled EEGdenoiseNet and PhysioBank benchmarks, pass ablations, task-locked event-related potential (ERP) preservation, runtime diagnostics, dry-electrode exercise EEG, and a downstream rapid serial visual presentation (RSVP)/P300 speller task using only FP1 and FP2. Across 26 EEGdenoiseNet input signal-to-noise ratio (SNR) levels, it obtained the largest average artifact-region SNR improvement among the tested wavelet, empirical mode decomposition (EMD), and artifact-label-guided wavelet quantile normalization (WQN) baselines for EMG, EOG, and combined EOG+EMG contamination (9.05, 8.28, and 8.00 dB, respectively). In exercise EEG, denoising reduced high-amplitude artifact burden and increased alpha and steady-state visual evoked potential (SSVEP) spectral-prominence metrics. In the P300 validation, the full SPAR-EEG sequence increased repetition-curve area under the curve (AUC) by 0.048 (Holm-adjusted p = 0.0069) and improved final Letter@15 accuracy by 7.8 percentage points. These results suggest that artifact-specific selective attenuation can provide a practical self-contained alternative for single-channel EEG denoising in low-burden and movement-prone settings.
Access status: Open Access ,
Engineering Perspectives on Kelp (Saccharina japonica) Cultivation Structures in China: Nearshore Practices and Transition to Offshore Environments
(MDPI AG, 2026-09-27) Huang, Chenxuan; Nguyen, Huu Phu; White, Lindsey; Wang, Chien Ming
China produces approximately 91.5% of the world’s Saccharina japonica (the species denoted by “kelp” herein) and its kelp farms collectively constitute one of the largest populations of floating marine structures anywhere. This review examines them as marine structures rather than as crops, asking what their geometry, load paths and standardization status imply for mechanization and for survival offshore. Nearshore cultivation rests on a single structural family, the floating single line, whose three variants, namely, horizontal, vertical and one-dragon kelp rope raft culture, differ in how the grow-out rope attaches to the backbone rather than in load path or design basis. Of these, the one-dragon configuration is the most promising for offshore adaptation, though on qualitative rather than measured grounds. The principal barrier to mechanized harvesting lies not in harvesting machinery but in a rope geometry dimensioned for manual operation and lacking standardized parameters. Integrated multi-trophic aquaculture, treated here as a structural modification of the kelp longline rather than as an ecological strategy, substantially increases suspended loading without any reported reinforcement of backbone, mooring or buoyancy. Recent offshore projects in Fujian and Shandong prove to be seaward extensions of nearshore rafts rather than purpose-designed systems, in contrast to international programmes which, although pre-commercial, begin from a stated design environment and return verifiable outcomes. We conclude that the defining obstacle to both mechanization and offshore expansion is the cultivation structure itself, and identify mechanization-compatible, load-verified structural standards as the critical and currently unaddressed research priority.
Access status: Open Access ,
The Limit of an Investor’s Capacity to Process Information
(Auckland University of Technology, 2026) Nguyen, Thang Ngoc; Nguyen, Nhut (Nick) Hoang; Phillips, James G.; Kumar, Saten
Standard models of financial markets typically assume that investors can process available information efficiently and incorporate it into prices without substantial delay. However, research in behavioural finance and psychology suggests that individuals face limits in attention and cognitive processing capacity. Although prior studies document the consequences of limited investor attention in financial markets, less is known about how specific features of information environments jointly influence investors' decision quality and cognitive workload. This thesis addresses this gap by examining how information quantity, task complexity, irrelevant data, and time constraints affect financial decision-making. The analysis is based on a controlled laboratory experiment in which participants perform stock valuation and trading tasks. Participants calculate the intrinsic value of firms using provided financial data and then decide whether to buy, hold, or sell the stock. Decision quality is measured using intrinsic value calculation accuracy and trading decision accuracy, while cognitive load is measured using self-reported mental effort. The thesis consists of three studies. Study 1 examines the joint effects of information quantity and task complexity. The results show that both factors significantly reduce decision accuracy and that their interaction intensifies the decline in performance as cognitive demands increase. Study 2 introduces irrelevant financial information to examine whether non-diagnostic cues influence decision outcomes. The results indicate that irrelevant information reduces performance when the information environment is otherwise simple, although this effect weakens when overall information load is high, suggesting selective filtering of information. Study 3 examines the role of time constraints and shows that reducing time pressure significantly improves decision accuracy, particularly in tasks with high complexity and large information sets. Overall, the findings provide experimental evidence that investor decision quality is strongly shaped by cognitive constraints. The thesis contributes to the literature on limited investor attention and cognitive load by demonstrating how multiple dimensions of information environments interact to influence financial decision-making. The results also have practical implications for financial disclosure design, suggesting that excessively complex or information-dense disclosures may hinder investors' ability to extract relevant signals.
Access status: Open Access ,
Determination of Asbestos Fibre Release Potential from Contaminated Craft Sand and Sand-Based Toys
(Auckland University of Technology) Berry, Terri-Ann
Over the last century, human exposure pathways for asbestos have changed from those associated with direct interaction with the minerals (e.g. mining) to secondary exposure from a variety of asbestos-containing materials, including those linked to para-occupational, domestic and environmental pathways. In 2025, the discovery of asbestos in children’s toy sand products caused panic, stress and fear amongst the community. Although there is no safe level of asbestos exposure, these products were deemed to be low risk by some due to uncertainty around whether fibres could become airborne during play. In this experimental study, asbestos contaminated sand samples were “played with” inside a Class A asbestos enclosure by an asbestos technician wearing protective equipment. Air samples collected were analysed using a combination of light and electron microscopy undertaken by two independent laboratories. 90% of tested craft sands released asbestos fibres into the air during play activities. Some fibres met the criteria for ‘respirable fibres’. Air samples taken during experiments using moulding sand samples did not show any airborne asbestos fibres; however, analytical detection limits for this study were high. This demonstrates that the potential risk from asbestos in the product can create a real risk from fibres in the air; however, the magnitude of this risk has yet to be determined by further research. Nevertheless, no amount of risk of asbestos exposure through toys is acceptable, so the most practical way to eliminate this risk would be to avoid the use of artificial sand products in toys or similar consumer products.
Access status: Open Access ,
Influence of an Electronic Medical Record with a Clinical Decision Support System on Inpatient Falls, Nurses' Clinical Experiences, and Decision-Making Processes for Fall Prevention: A Mixed Methods Study
(Auckland University of Technology, 2026) Murray, Paula Anne; Garrett, Nick; Foster, Mandie; Signal, Nada
Background: This thesis is informed by the researcher’s 25 years of operational management experience in New Zealand (NZ) private hospitals, including leading the design and implementation of an electronic medical record (EMR) with an integrated clinical decision support system (CDSS). International evidence shows that CDSS-enabled electronic health systems have advanced, offering evidence-based recommendations, workflow integration, and improvements in process measures and clinical performance, all with the potential to enhance fall-risk identification and outcomes. However, no NZ studies have explored the impact on inpatient fall management or nurses’ clinical decision-making in hospital settings. Objective: Determine whether replacing a patient management system with an EMR and a CDSS for fall risk assessment affected the incidence of falls, and to explore nurses’ clinical experiences and decision-making in fall prevention. Design: Employed a convergent mixed-methods approach consisting of three studies: a systematic review, a systems analysis, and a qualitative study, guided by the Donabedian Framework (Structure, Process, Outcomes). Methods Systematic review: Systematically synthesised empirical evidence on the impact of digital health systems, including EMR and CDSS, on inpatient fall outcomes and nurses’ clinical decision-making that followed a deductive analysis. Systems analysis: Examined whether replacing a patient management system with an EMR and CDSS for fall risk assessment influenced inpatient fall incidence at a private hospital using descriptive and inferential analyses of retrospective system data. Qualitative study: Explored nurses’ experiences with the electronic fall risk assessment CDSS and its influence on fall-related clinical decision-making. A hybrid inductive/deductive approach was used, combining inductive coding with deductive organisation. Integration of data: Using the mixed-methods synthesis, data from the systematic review, systems analysis, and qualitative study were integrated to develop a comprehensive understanding of how the EMR with CDSS may influence fall-prevention practices in private hospitals and to inform the overall discussion, conclusions, and recommendations. Findings: Structural factors, particularly system design, usability, hardware, training, and staffing, influenced nurses’ interactions with the EMR and its embedded CDSS, thereby affecting workflow efficiency and data quality. At a process level, the CDSS improved coordination by strengthening communication, shared awareness of patient risk, and fall-related documentation, though use varied. In terms of outcomes, electronic systems enhanced the quality and visibility of fall reporting but did not reduce fall incidence. Nurses primarily relied on clinical judgement, highlighting an ongoing tension between CDSS design and clinical relevance and emphasising the need for nurses’ involvement in system design. Conclusions: This thesis demonstrates that implementing an EMR with CDSS is a major organisational change that affects clinical practice, workflows, communication, and nurses’ experiences, indicating that technology alone does not improve patient safety. When systems misalign with workflows or lack clarity and usability, nurses develop workarounds that can compromise data quality, highlighting the need for co-design and ongoing training to support consistent documentation standards and confident use. While electronic systems offer benefits, fall prevention continues to rely on human judgement, vigilance, and engagement. Overall, the findings indicate that EMR with CDSS implementation influences fall prevention through complex interactions among system design, clinical processes, and human factors.