21436 research outputs found
Sort by
Unravelling domestic cat behaviour using accelerometers and machine learning : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Animal Science at Massey University, Palmerston North, New Zealand
No full textIt is estimated that there are at least 445 million companion cats worldwide. Despite this large number, little is known about how environmental factors (e.g., weather, presence of children, dogs or other cats) are associated with domestic feline behaviour and welfare. Traditional behavioural observation methods are limited by short observational windows, high labour demands, and subjective nature of observations, making them unsuitable for in-depth, longitudinal studies. This thesis leverages recent advances in accelerometer technology and machine learning to overcome these limitations to quantitatively analyse continuous domestic cat behaviour, providing new insights into the influence of environmental factors on their behaviour.
In the first phase of the thesis (Chapter 3), machine learning models were trained to classify cat behaviours from acceleration data. The models achieved a minimum accuracy of 70%, which improved as the number of behavioural classes was reduced, and with random forest models generalising better to new data than self-organising maps. Models for harness-mounted devices (accuracies 77% - 83%) performed slightly better than those for collar-mounted devices. The model for the collar-mounted device that classified eight different behaviours (active, eating, grooming, littering, lying, scratching, sitting, standing), performed adequately (accuracy 73%), and was considered the most practical due to cats being more likely to wear a collar than a harness.
The second and third phases of this thesis investigated how environmental factors influenced cat behaviour, both in semi-outdoor research cats (Chapter 4) and in pet cats within a home environment (Chapter 5). Weather variables, particularly daylength and the temperature humidity wind index, were significantly associated with domestic cat behaviour. Seasonal changes in grooming and scratching were closely linked to the natural hair growth cycle. Behaviours were influenced by individual differences and disruptions, such as the return of a cat to the group after an absence, affecting their behaviour. Among pet cats, those with outdoor access showed behavioural adaptations to seasonal weather changes, seemingly prioritising thermal comfort. Multi-cat households and those with at least one child present were associated with an increase in alertness-related behaviours of increasing sitting with decreasing lying.
By leveraging machine learning and accelerometer data, this research advances current methodologies in animal behaviour studies, providing an approach for continuous, minimally invasive animal behaviour and welfare assessment. This thesis contributed to bridging the gap in our understanding of how environmental factors can shape animal behaviour and welfare and highlighted the importance of taking interindividual differences and environmental factors into account when assessing domestic cat behaviour
Squeezing through the gut : micro-manufacturing of smart capsule : a thesis presented in partial fulfilment of the requirements for the degree of Master of Engineering in the School of Food and Advanced Technology at Massey University, Palmerston North, New Zealand
No full textCapsule robotics has been an important part of medical evaluation of conditions within the Gastro-intestinal (GI) tract since their inception in 1957. As technology improved, their functionality expanded to explore the entirety of the GI tract such as; taking images; measuring chemical and mechanical properties; delivering drugs; performing biopsies; and retrieving samples of microbiome. This has led to a significant increase in adoption with over 130,000 procedures conducted annually; however, approximately 1,400 of these procedures result in capsule retention, which requires surgery to remove the capsule. This risk is significantly increased, with patients suffering from inflammatory bowel diseases such as Crohn’s disease being approximately 8 times more likely to retain capsules. This thesis investigates how to improve the safety and motility of microrobotic capsules, especially for those with intestinal complications, to ensure equal access to this technology and improve patient outcomes, increasing access to the information needed to better treat these ailments. This project covers multi-disciplinary subject areas ranging from biomedical technology, mechanical characterisation, robotics, and electronics. Including the design and manufacture of capsule exteriors down to the microscale and the development of new testing equipment for them, such as a synthetic intestine tensile testing platform and robotic intestine testing system for quantifying capsule performance inside an intestinal-like environment. In addition, a precise testing procedure is provided with the created equipment so that experiments can be easily replicated and accurate data are collected. The best capsule design determined is a three-dimensional resin printed capsule using surgical guide resin with a six-ridged segmented design. Determined using force response data from pulling capsule designs and measuring the excess power draw to push them through a synthetic intestinal con striction. This also demonstrated the functionality of the testing equipment developed during this research project. In the future, these capsule design considerations are expected to be used to increase the adoption of this field of technology and improve patient outcomes. Also, it is hoped the testing equipment is used and developed further by my research group to improve their respective project outcomes; and by any external group looking to test their capsule prototypes
Genomic characterisation of New Zealand working farm dogs : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Animal Science at Massey University, Manawatū, New Zealand
No full textThe majority of working farm dogs in New Zealand belong to either the Huntaway or Heading Dog breeds. These breeds are unique to New Zealand and are vital to efficient livestock farming, yet no large-scale DNA sequencing studies have been conducted on them to date. The Right Dog for the Job project aims to improve the health and performance of the farm dog population by analysing the genetic and phenotypic information from a sample of 2400 Huntaways and Heading Dogs, including 250 whole genome sequences. As the first use of this dataset, this project aimed to characterise the genetic variation in the sample of whole genome sequences. To this end, sequence bioinformatics tools were used to align sequence reads, then call, filter, and annotate 20 million genetic variants. The dataset was intersected with 395 OMIA-reported functional loci to identify 27 previously described Mendelian variants segregating in the population. Five of these (in the SOD1, VWF, CUBN, CLN8, and SGSH genes) were highlighted as compelling candidates for diagnostic selection. Next, the dataset was surveyed for high impact variants segregating within 132 genes previously shown to harbour phenotype causing variants. The aim here was to use variant effect prediction to identify novel causal variants within functionally relevant genes. This analysis yielded nine causal hypotheses (in the CNGB1, ABCA4, CNP, SLC3A1, CCDC66, GLB1, CYP1A2, and STK36 genes) for future association testing. The second stage of the project leveraged a sample of 299 dogs that were genotyped with the AxiomTM Canine HD Array. Of these dogs, 188 had also been whole genome sequenced. A linkage disequilibrium analysis was conducted to identify predictive markers on the array for several of the 27 Mendelian variants, with a view to supporting future marker-assisted selection. Next, the whole genome sequences were used as a reference to impute missing genotypes in the sample of 111 dogs genotyped with only the Axiom SNP chip, expanding the genomic dataset. Finally, GWAS were performed to identify genetic associations with four body size traits of interest: height, length, chest circumference, and muzzle circumference. Height was significantly associated with a region near LCORL, a gene known to regulate body size in various species. Despite representing one of the first analyses of the ‘Right Dog’ research programme, this project has identified genes and variants that should be of immediate practical use to working dog breeders and owners. The project has also helped to generate an extensive genomic dataset that will underpin future research as part of the broader programme, and ultimately contribute to the health, welfare, and performance of New Zealand’s iconic Huntaway and Heading Dog breeds
Saccharomyces cerevisiae Yeast-Based Supplement and Breast Milk Supply: A Randomised Placebo-Controlled Trial
No full textSaccharomyces cerevisiae yeast-based supplements (SCYS) are frequently used as galactagogues with limited evidence of their efficacy. This study investigates the effect of SCYS on human milk oligosaccharide (HMO) concentration and indicators of milk supply. Sixty-eight breastfeeding women with a healthy singleton infant aged 1–7 months were randomly assigned to consume a SCYS product (5 g/day) or placebo for 4 weeks. The primary outcome was the change in the total HMO concentration. The secondary outcomes included participants' perceptions of milk supply, intervention effectiveness, postnatal distress, infant feeding patterns, infant anthropometry, and adverse effects. Intention-to-treat analysis was performed. Multivariable linear regression analysis showed no significant effect of SCYS on individual or total HMO concentrations. However, 65% of women in the SCYS group, compared to 35% in the placebo group, perceived an increase in milk production (p < 0.05). No significant differences were found for other secondary outcomes. However, mothers in the SCYS group had a small but significant improvement in perception of their milk quantity and quality (p < 0.05). SCYS use was also associated with significantly lower formula use at 6 months postpartum (4% vs. 27%, p < 0.05). While SCYS does not impact HMO concentration, it may improve women's perceptions of milk supply. A larger randomised controlled trial is needed to assess its potential effects on actual milk production and composition and address issues of perceived insufficient milk.
Trial Registration: This trial was registered at the Australian New Zealand Clinical Trials Registry (trial registration number: ACTRN12619000704190).fals
An exploration of factors that have facilitated and constrained access, adoption and availability of mind-body therapies as adjunctive interventions to treat trauma-related conditions in Aotearoa, New Zealand : a thesis submitted in partial fulfilment of the requirements for the degree of Doctor of Clinical Psychology at Massey University - Te Kunenga ki Pūrehuroa, Albany, Aotearoa New Zealand
No full textThis qualitative study explores factors that facilitate and constrain the use of mind-body interventions to treat trauma-related conditions in Aotearoa New Zealand (NZ). The phenomenon explored references mind-body therapies as an adjunctive treatment option with a focus on trauma-sensitive yoga. This study is intended to examine mind-body interventions as a complementary treatment approach and augmentation of interventions such as cognitive behavioural therapies and other validated, evidence-based approaches to treating trauma-related conditions. There is a significant body of literature that supports the use of talk therapies, such as cognitive behavioural therapy, to treat symptoms associated with trauma-related mental health conditions. Recent research has revealed that in many cases, trauma cannot be resolved through interventions that utilise talk therapies alone, as trauma, according to some theorists, is located not only in the core of the brain but also within the body. A practical treatment approach to mitigate individual experiences of trauma is to integrate Western psychological talk therapy approaches with those that focus on calming the nervous system, such as trauma-sensitive yoga, romiromi, mindfulness, and somatic experiencing. The current study highlights factors across the social system that have facilitated and constrained the access, adoption, and availability of mind-body therapies as complementary approaches for treating trauma-related conditions in the NZ context. Findings in the current study highlight that in the NZ context, few factors have facilitated access, adoption and availability of mind-body therapies as adjunctive interventions to treat trauma-related mental health conditions. A more significant number of factors, it appears, have constrained access, adoption and availability of mind-body therapies. Findings illuminate an overall positive attitude toward the utility of mind-body therapies as adjunctive treatments for trauma-related conditions, which highlights the potential for greater use of such interventions in the NZ setting
Comparison of growth, relative abundance, and diet of three sympatric Hemiandrus ground wētā (Orthoptera, Anostostomatidae) in a New Zealand Forest
Get PDFEcology is an essential discipline for understanding the biology and behavior of organisms. This study increased knowledge of three sympatric species of ground wētā (Hemiandrus spp.). Hemiandrus ground wētā are nocturnal Ensifera that live in burrows during the day, and for these reasons, there is limited knowledge of their activity, development, and diet. We examined the diet of Hemiandrus electra, Hemiandrus ‘disparalis’, and Hemiandrus nox by examining the crop contents of specimens caught in malaise traps set in a native forest (St Arnaud) over seven months during two spring-summer-autumn seasons (2004/2005 and 2005/2006). The three species investigated varied in the plant and invertebrate fragment proportions in their diets. Hemiandrus electra and H. ‘disparalis’ were predominantly herbivores, while H. nox was primarily carnivorous, although plant matter constituted 20% of its diet. We identified the species and sex of 966 Hemiandrus wētā that were intercepted by the same malaise traps. Our results showed that H. electra was the most abundant species, with 701 individuals, while 157 and 109 were identified as H. ‘disparalis’ and H. nox, respectively. Surprisingly, the species with maternal care (H. electra) was the only one of the three sympatric Hemiandrus to have an even sex ratio; for the other two species, more males were caught in the traps. Using hind leg dimensions, we categorized each female H. electra specimen using naïve Gaussian mixture model clustering, which identified five size clusters (putatively corresponding to instars). Based on the month of collection and the growth category, we observed no seasonality in the development of this species of ground wētā, as almost all instars were found in each month of the sampling period. This study found no evidence that these nocturnal forest species synchronize their growth or reproduction with seasons.fals
Navigating scientific modelling and uncertainty: Insights from hazard, risk, and impact scientists in disaster risk management (DRM)
Get PDFScientific models have long been used as an important tool in assisting in decision-making in Disaster Risk Management (DRM). However, it is commonly understood that uncertainty in these models significantly influences the integration of model outputs into decision-making processes, presenting a challenge for effective uncertainty communication. This paper explores how scientists in DRM approach modelling and uncertainty.
We conducted seventeen in-depth qualitative interviews in Aotearoa, New Zealand, with scientists working on DRM. We provided an overview of the varied approaches and key factors influencing their processes for model characterisation and communicating uncertainties. We used Reflective Thematic Analysis to construct key themes, including (a) model development and characterisation, (b) accountability and opinion, (c) communication approach and challenges, and (d) collaboration for effective uncertainty communication.
We found that different DRM scientists have different disciplinary and experiential training, experience, and interaction with decision-makers. These factors greatly influence their choices regarding scientific modelling and communication of uncertainty. The lack of a generally accepted guideline for a best practice approach to uncertainty communication is a key barrier to successfully incorporating uncertainties into DRM decision-making. We suggest that a better collaboration between scientists and decision-makers throughout the lifecycle of the model development process is a way forward for effective communication of uncertainty in DRM decision-making.fals
Exploring the Use of Verjuice for Reduced Sodium Pickle Production: Determination of Hedonic and Rejection Thresholds
Get PDFFood industry has been searching for strategies to reduce sodium in foods. Verjuice, an acidic juice made from thinned grapes, appears to enhance perceived saltiness in pickles. This study aimed to determine the affective thresholds (Consumer Rejection Threshold: CRT, Compromised Acceptance Threshold: CAT, Hedonic Rejection Threshold: HRT) of pickles produced with verjuice and vinegar, in order to assess whether salt concentration can be reduced. Consumers (n = 103) tasted eight pairs of samples (four per acidifier), comparing the control (14 g/L salt) with one of the treatments (either 0, 3.5, 7.0 or 10.5 g/L). Results showed that it is possible to reduce more salt by using verjuice as an acidifier compared to vinegar, without impacting preference in relation to the control. CAT values were similar for both acidifiers; however, HRT values for vinegar pickled cucumbers were lower compared to those with verjuice, suggesting that vinegar allowed for a greater salt reduction without sensory rejection. This work used hedonic and rejection thresholds to study whether salt concentration in pickle production can be reduced. This method has wide applicability in the food industry, offering a process to identify when the level of an ingredient exceeds or does not meet consumer expectations. Obtaining data on affective thresholds of pickles preserved with vinegar and verjuice (proposed herein as an acidifier alternative ingredient) is crucial for product development purposes, supporting quality control and guiding formulation development. The use of verjuice may be a first step in product innovation given its health and sustainable credentials. Data from this study show that food producers can reduce salt content in pickles down to ~4 g/L by preserving them with either vinegar or verjuice, without impacting liking compared to levels currently available in the market (14 g/L). This reduction can be implemented in the production of vinegar-based pickles, leading to healthier products without compromising consumer acceptance.fals
Multimodal Deep Learning for Android Malware Classification
Get PDFThis study investigates the integration of diverse data modalities within deep learning ensembles for Android malware classification. Android applications can be represented as binary images and function call graphs, each offering complementary perspectives on the executable. We synthesise these modalities by combining predictions from convolutional and graph neural networks with a multilayer perceptron. Empirical results demonstrate that multimodal models outperform their unimodal counterparts while remaining highly efficient. For instance, integrating a plain CNN with 83.1% accuracy and a GCN with 80.6% accuracy boosts overall accuracy to 88.3%. DenseNet-GIN achieves 90.6% accuracy, with no further improvement obtained by expanding this ensemble to four models. Based on our findings, we advocate for the flexible development of modalities to capture distinct aspects of applications and for the design of algorithms that effectively integrate this information.falsearticle-number: 2
Psychosocial Work Environment Risks Among Danish Fishers
Get PDFOBJECTIVE: Our paper focuses on psychosocial risks that are under-researched within the fishing industry. The study was conducted in Denmark. Firstly, we investigated how Danish fishers perceive their psychosocial work environment. Secondly, we formed a group of industry partners and stakeholders with the purpose of designing ways to mitigate the reported stressors.
METHODS: The study design is action research, employing qualitative co-creative methods for data collection in both the first and the second stages. The methods include chronicle workshops and design thinking workshops. Participants in the workshops came from different stakeholder groups, including fishers. We ensured that the focus remained on centering the voices of the fishers to understand the psychosocial risks within the industry context.
RESULTS: In the first stage, we used thematic analysis, which led to the identification of ten themes: (1) discouragement, (2) stress, (3) poor economy, (4) negative atmosphere, (5) distrust, (6) injustice, (7) fear, (8) unpredictability, (9) powerlessness, and (10) lost pride. This study explored the key risks and highlighted that a lack of influence and meaning, reduced rewards, and increased unpredictability are likely to lead to increased stress in an environment that lacks social support.
CONCLUSION: Existing research into the psychosocial working environment has established six dimensions where an imbalance leads to negative health effects. They are sometimes referred to as "the 6 gold nuggets" where an imbalance leads to negative health effects. We compared our themes to these gold nuggets and identified a clear overlap, which gave us confidence in the relevance of our data. Through the workshops, we co-created six recommendations to address the reported risks and provide policymakers with actionable recommendations.fals