- Research Article
- 10.1016/j.atech.2026.101862
Transparent deep learning for medicinal plant recognition: A hybrid CNN–ViT approach with explainable AI on BDHerbalPlants
- Aug 01, 2026
- Smart Agricultural Technology
- Sunzil Khandaker + 4 more +4
Publications from 2021 to 2026
Showing 10 of 2,990 papers
Transparent deep learning for medicinal plant recognition: A hybrid CNN–ViT approach with explainable AI on BDHerbalPlants
Transparent reporting is central to reproducible radiological AI research: A call to action.
Artificial intelligence (AI) is increasingly embedded in radiology research and practice, yet concerns about the reproducibility of AI studies remain a key barrier to regulatory acceptance and clinical translation. Transparent reporting across the analytic pipeline is essential for independent verification, evidence synthesis, and safe implementation. To examine major barriers to transparent reporting in radiological AI research and propose practical solutions to strengthen reproducibility. We identify four interrelated challenges undermining reporting quality. First, reporting is often treated as a final-stage manuscript requirement rather than a prospectively embedded component of study design, leading to loss of methodological detail. Second, inconsistent operationalization of reporting standards by journals and funders makes it difficult to assess reporting completeness. Third, fragmentation and overlapping complexity of AI-specific reporting guidelines increase interpretive burden for authors and reviews and contribute to heterogeneity in reporting. Fourth, limited availability and usability of data and code restrict independent verification, even when resources are nominally shared. These issues are amplified by the methodological complexity and heterogeneity of radiological AI pipelines. For each challenge, we outline actionable strategies including prospective documentation of methods, mandatory and consistently applied reporting standards, harmonized reporting principles, and clearer requirements for data and code availability with structured alternatives when sharing is constrained. Strengthening transparent reporting is a prerequisite for reproducible and clinically translatable radiological AI research. Embedding reporting within study design, aligning stakeholder expectations, and improving access to reproducibility-enabling resources are essential to enhance credibility, facilitate independent evaluation, and support safe clinical implementation of AI in radiology.
Read moreEstablishing an Australian Nurse Practitioner Research Agenda: A National Consensus Study.
To establish research priorities for Australian Nurse Practitioners and develop a robust research agenda. A descriptive, exploratory approach was used and conducted in two stages. Data were collected over two stages. Data for Stage 1 were collected from 14 December 2023 to 16 February 2024. For the Delphi rounds (Stage 2), data were collected from 11 March to 24 March 2024 for Delphi Round 1 and from 12 June to 26 June 2024 for Delphi Round 2. An exploratory survey was used in Stage 1 to identify clinical challenges and research themes perceived as important for Australian Nurse Practitioners. In Stage 2, a two-phased modified Delphi survey was conducted to prioritise the research themes identified in Stage 1. A total of 315 participants responded to the exploratory survey, with a majority being female (77%), aged between 30 and 75 years. Participants were primarily employed in the public healthcare sector (60%), the private sector (23%), while 17% practised across both sectors. A total of 1335 challenges facing Australian Nurse Practitioners were identified. Sixty-nine participants completed the first Delphi round, and 33 the second, giving response rates of 21% (69/315) and 48% (33/69), respectively. The first Delphi round yielded 11 research themes. Seven of these yielded Content Validity Indices of < 0.90. Four research priority areas remained and were ranked in order of importance. The identified Australian Nurse Practitioner research priorities will play a pivotal role in shaping policies, fostering the efficient integration of Nurse Practitioners into the healthcare system and advancing research capacity. Nurse Practitioners are established providers of high-quality care internationally; however, they face persistent integration challenges in Australia. This study delivers a nationally relevant, consensus-based research agenda that identifies key priorities across clinical, educational and leadership domains. No patient or public contribution.
Read moreNo evidence of sex differences in plastic ingestion by Sable Shearwater (Ardenna carneipes) chicks.
Staging Hot Weather and Fire
Facing Mass Extinction
Olive Components (Biophenols or Polyphenols) in Neurodegenerative Disease Models and Clinical Studies: A Systematic Review of Evidence and Translational Barriers.
Introduction: Olives have been used in traditional Mediterranean medicine for thousands of years to address the causes of inflammation, ageing and cognitive health. Traditional preparations of olive include olive oil and olive leaf extract, which are major components of diets that contribute to maintaining cognitive function and reducing neurodegenerative disease risk. Aims of the study: This systematic review aimed to synthesise experimental and limited human evidence on olive biophenols in neurodegenerative disease models, identify the most studied compounds, characterise their mechanisms of action, and evaluate key translational barriers. Materials and methods: Following PRISMA 2020 guidelines and registered with PROSPERO (CRD420251252252), primary studies investigating the effects of well-characterised olive biophenols in neurodegenerative relevant in vitro, in vivo, or human models were systematically reviewed. Each study was assessed for its design, experimental model, mechanistic outcomes and reported limitations. Risk of bias was evaluated using validated tools (SYRCLE/OHAT/ToxR) appropriate for preclinical and experimental study designs. Results: Among the 25 studies, 7 (28.0%) examined oleuropein or oleuropein aglycone, 10 (40.0%) focused on hydroxytyrosol or its derivatives, and 9 (36.0%) investigated oleocanthal. Most studies employed in vivo animal models (57.7%), predominantly transgenic mouse models of AD and toxin-induced PD models. Oleuropein-based studies reported inhibition of amyloid-β and α-synuclein aggregation with behavioural improvements. Hydroxytyrosol primarily exerted antioxidant and anti-inflammatory effects with modest cognitive benefits. Oleocanthal showed the most consistent anti-amyloid and anti-tau activity, including enhanced amyloid-β clearance across the blood-brain barrier. Most studies show a moderate risk of bias due to incomplete reporting, randomisation and blinding. Conclusions: Olive biophenols demonstrate consistent neuroprotective effects in preclinical models; however, translation to clinical application remains limited by pharmacokinetic constraints, methodological heterogeneity, and insufficient human evidence.
Read moreCaregiver-Implemented Hanen Programs®: A Narrative Review.
Hanen programs® have been used by clinicians to support caregivers of young children with communication needs to learn optimal ways to respond in everyday interactions and facilitate child-oriented strategies that promote interaction and language. This review aimed to evaluate peer-reviewed literature regarding Hanen programs® for children with speech, language, and communication needs regarding the credibility, nature, quality, scope of the research, and outcomes of the research findings. The Preferred Reporting Items for Systematic reviews and Meta-Analysis protocol was used. Of 196 papers documenting Hanen interventions, 57 papers met inclusion criteria. Studies were synthesized and evaluated to rate (a) credibility/levels of evidence; (b) nature of existing research: efficacy, effectiveness, and perceived effectiveness; (c) quality of the studies; (d) scope of evidence; and (e) outcomes for each Hanen program® for children with speech, language, and communication needs. Credibility: Studies of Hanen programs® were reviews (7%), randomized controlled trials (17.5%), nonrandomized controlled trials and mixed methods (31.6%), case-control or cohort studies (26.3%), qualitative reviews (1.8%), or employed qualitative methods (15.7%). Nature: The majority were efficacy studies (57.9%), others were perceived effectiveness (24.6%), and others were effectiveness studies (8.8%). Quality: Few studies were rated as high quality. Scope: Most studies related to two Hanen programs® (It Takes Two to Talk® and More Than Words®) conducted across 11 countries (Canada, the United States, Spain, Turkey, Republic of Ireland, the Netherlands, the United Kingdom, Australia, Hong Kong, Singapore, and Malaysia), and a few studies evaluated adaptations of Hanen programs® (e.g., telehealth). The studies reported on dyad interactions; child language, behavior, and autism; and caregiver perceptions, stress, and self-efficacy. Outcomes: Evidence supporting significant communication outcomes for different programs exists; however, mixed results demonstrate a variety of language, interaction, acceptability, feasibility, and maintenance outcomes. A range of evidence supports the use of Hanen programs® in its original format, with acceptability reported by families from target communities (e.g., middle-class families) and educators (e.g., Westernized countries). Effectiveness studies in real-world contexts conclude that future research could evaluate content flexibility and adaptations to meet a diversity of families and communities. A precision family initiative has been suggested to shift future research from "Does Hanen work?" to "How can Hanen-based intervention optimize outcomes for individual families?"
Read moreK–12 teachers’ perspectives on AI use in education through the lens of activity theory
The rapid proliferation of AI systems in the past several years has pointed to a potential revolution in education. However, there is currently a lack of theoretical framework to explain how educators understand and view AI and its impact on education. The present study draws on qualitative data from Australian teachers at K–12 levels to explore their perceptions on using AI in school settings. Data were collected in April/May of 2023, approximately six months into the sudden global recognition of AI’s potential impact on society, including education. Our thematic analysis suggested six major themes representing in-service teachers’ views that AI is unsuitable, impersonal, imperfect, uncertain, assisting, and inevitable. This study also presents theorizing of the activity theory in an AI world, highlighting the complex and nuanced nature of teachers’ sentiments regarding the use of AI in school settings.
Read moreHeparin Anticoagulant Therapy and Its Monitoring.
Heparin remains a foundational parenteral anticoagulant across both acute and chronic care settings. This narrative review summarizes clinical indications and dosing of unfractionated (UFH) and low-molecular-weight heparin (LMWH). It also details laboratory monitoring using activated partial thromboplastin (APTT), anti-factor Xa (anti-Xa), activated clotting time (ACT) and viscoelastic testing (VET), including common pitfalls and interferences. We provide considerations for specific populations as well as complications including heparin resistance, heparin-induced thrombocytopenia (HIT) and heparin reversal strategies. Future research directions include harmonization of therapeutic ranges, mitigation of assay interference and prospective evaluation on monitoring, particular in extracorporeal membrane oxygenation (ECMO), pregnancy and cardiac surgical settings.
Read more