Massey Documents by Type
Permanent URI for this communityhttps://mro.massey.ac.nz/handle/10179/294
Browse
Search Results
Item Evaluation of a text-mining application for the rapid analysis of free-text wildlife necropsy reports(PLOS, 2025-11-25) Saverimuttu S; McInnes K; Warren K; Yeap L; Hunter S; Gartrell B; Pas A; Chatterton J; Jackson BThe ability to efficiently derive insights from wildlife necropsy data is essential for advancing conservation and One Health objectives, yet close reading remains the mainstay of knowledge retrieval from ubiquitous free-text clinical data. This time-consuming process poses a barrier to the efficient utilisation of such valuable resources. This study evaluates part of a bespoke text-mining application, DEE (Describe, Explore, Examine), designed for extracting insights from free-text necropsy reports housed in Aotearoa New Zealand’s Wildbase Pathology Register. A pilot test involving nine veterinary professionals assessed DEE’s ability to quantify the occurrence of four clinicopathologic findings (external oiling, trauma, diphtheritic stomatitis, and starvation) across two species datasets by comparison to manual review. Performance metrics—recall, precision, and F1-score—were calculated and analysed alongside tester-driven misclassification patterns. Findings reveal that while DEE (and the principals underlying its function) offers time-efficient data retrieval, its performance is influenced by search term selection and the breadth of vocabulary which may describe a clinicopathologic finding. Those findings characterized by limited terminological variance, such as external oiling, yielded the highest performance scores and the most consistency across application testers. Mean F1-scores across all tested findings and application testers was 0.63–0.93. Results highlight the utility and limitations of term-based text-mining approaches and suggests that enhancements to automatically capture this terminological variance may be necessary for broader implementation. This pilot study highlights the potential of relatively simple, rule-based text-mining approaches to derive insights natural language wildlife data in the support of One Health goals.Item Enabling near real time use of wildlife necropsy data: Text-mining approaches to derive interactive dashboard displays(PLOS, 2025-09-19) Saverimuttu S; McInnes K; Warren K; Yeap L; Hunter S; Gartrell B; Pas A; Chatterton J; Jackson B; Panter CManual review of necropsy records through close reading and collation is a time-consuming process, leading to delays in knowledge acquisition, communication of findings, and subsequent actions. Text-mining techniques offer a means to reduce these barriers by automating the extraction of information from large volumes of free-text clinical reports, minimizing the need for manual review. Additionally, interactive dashboards enable end users to interrogate data dynamically, tailoring analyses to their specific needs and objectives. Here, we describe the principles underlying an application designed to extract and visualize information from free-text necropsy records within the Wildbase Pathology register. Reflecting the structure of a traditional necropsy review—where each record is examined in detail to identify and collate key observations—the application is divided into three sections. The first allows a user to upload a dataset in comma separated value format as downloaded from the Wildbase Pathology Register. A user can then filter and interrogate selected signalment variables of the population within this dataset. The second section uses established text-mining calculations of word correlations and Latent Dirichlet Allocation to generate visualisations to give a user a subjective sense of common themes found within the uploaded data. The third and final section uses a custom rule-based algorithm to identify and quantify positive occurrences of clinicopathologic findings as input by an end user. The foundational methods employed in this application have the potential for broader application in veterinary and medical pathology, facilitating more efficient and timely access to critical insights.Item What's in a database? Insights from a retrospective review of penguin necropsy records in Aotearoa New Zealand(PLOS, 2025-04-30) Saverimuttu S; Hunter S; Gartrell B; McInnes K; Warren K; Pas A; Chatterton J; Yeap L; Jackson B; Raverty SWildlife necropsy databases often provide data for morbidity and mortality studies of free-ranging species, with implicit relevance for conservation goals, as well as domestic animal and human health. Retrospective reviews are a common way to derive insights from such opportunistic data, despite the methodological difficulties of performing these analyses, alongside findings being prone to bias. This study reviews morbidity and mortality data from Sphenisciformes of Aotearoa New Zealand, using records extracted and manually refined from submissions to the national Wildbase Pathology Register. The review corroborates the broader consensus that hoiho (yellow eyed penguin, Megadyptes antipodes) are most commonly diagnosed with infectious/inflammatory disease (43.1%, 422/978 diagnoses), kororā (blue penguin, Eudyptula minor) with traumatic injuries (42.9%, 156/364 diagnoses), and emaciation being a common finding across both species (33.9%, 393/1463 diagnoses). Further, there are marked spatiotemporal trends in submissions, driven primarily by the affected species and the submitting organisations, highlighting the biases within such databases that must be factored into the application of results. Typographical errors, redundancies from synonymous terms, and missing data are captured as barriers to performing manual reviews of free-text data. Overall, this study highlights strengths and limitations of storage and review of wildlife necropsy data while providing insight into threats faced by the penguins of Aotearoa.Item Ovariectomy as treatment for ovarian bacterial granulomas in a Duvaucel's gecko (Hoplodactylus duvaucelii)(Taylor & Francis, 18/06/2015) Barry M; Le Souëf AT; Brunton DH; Jakob-Hoff R; Jackson BCASE HISTORY: An adult female Duvaucel’s gecko (Hoplodactylus duvaucelii) from a threatened species breeding programme presented due to a prolonged gestation period and distended abdomen. CLINICAL AND PATHOLOGICAL FINDINGS: The gecko was in lean body condition with an irregularly-shaped, firm mass in the coelomic cavity. Radiographically there was a diffuse radioopacity within the coelomic cavity with cranial displacement of the right lung field. Ultrasonography revealed a round homogenous abdominal mass of medium echogenicity with an echogenic capsule. Haematology showed a leucocytosis with a moderate left shift in heterophils and toxic changes. Bilateral ovariectomy was performed to remove two ovarian granulomas and Salmonella enterica subspecies houtenae (IV) was cultured from the ovarian tissue. The gecko recovered well from the surgery, regained weight and remained in good health 3 years following the surgery. DIAGNOSIS: Pre-ovulatory stasis and ovarian granulomas associated with infection with Salmonella enterica subsp. houtenae. CLINICAL RELEVANCE: The surgery described in this case resulted in recovery of the gecko, which despite its loss of reproductive capability is of value as an education animal. This is the first report of pre-ovulatory stasis and ovarian granulomas associated with infection with Salmonella enterica in a Duvaucel’s gecko and is also the first reported case of pre-ovulatory stasis in a viviparous lizard species. The case adds to knowledge regarding potential reproductive pathology in lizards, which is particularly important information for managers of captive lizard breeding programmes.