Santoso Wibowo

Early and accurate cancer detection from medical imaging remains challenging because clinically relevant evidence is distributed across local image appearance, structural relationships between suspicious regions, and ordered imaging context. This study proposes a graph-aware and sequence-aware deep learning framework that combines a convolutional neural network (CNN) backbone for spatial feature extraction, a Graph Attention Network (GAT) for lesion-structure modelling, and a Bidirectional Long Short-Term Memory (BiLSTM) module for ordered-view or slice-sequence representation learning. Cross-attention-based multimodal fusion is evaluated exclusively for the RSNA mammography task, where the metadata branch is restricted to patient age and implant status, both available before diagnosis. In contrast, the primary LIDC-IDRI experiment is conducted as an image-only analysis because radiologist malignancy scores and semantic nodule attributes are annotation-derived variables and are not treated as independent clinical predictors. The framework is evaluated on the RSNA Breast Cancer Detection dataset and the LIDC-IDRI lung CT dataset using accuracy, precision, recall, F1-score, specificity, and area under the receiver operating characteristic curve (AUC). Additional ablation experiments assess the contribution of graph learning, sequence-aware modelling, and leakage-safe RSNA metadata fusion, while SHAP analysis quantifies the influence of the included RSNA metadata variables on multimodal predictions. For the RSNA multimodal experiment, the best held-out run achieved 95.2% accuracy and an AUC of 0.978, while three repeated runs yielded 94.0 ± 0.3% accuracy and an AUC of 0.970 ± 0.007 under the internal patient-wise benchmark protocol. These results should be interpreted as public-dataset benchmark outcomes rather than evidence of real-world clinical performance; external multi-centre and prospective validation is required before clinical deployment.

The efficient management of solid waste remains a persistent challenge worldwide, particularly in developing countries, due to increasing waste generation driven by rapid population growth and urbanisation. Artificial Intelligence (AI) has emerged as a transformative approach to improving solid waste management (SWM) by enhancing waste collection, sorting, recycling, and resource recovery processes. Several countries have integrated AI into their SWM systems; however, the extent of AI-driven automation across the SWM value chain remains insufficiently understood. This study examines the implementation of AI-powered systems across various areas of SWM in ten selected countries to identify progress, benefits, challenges, and opportunities for advancing sustainable SWM practices. A systematic literature review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. A total of 268 studies published between 2005 and 2025 were identified and analysed. The findings indicate that the selected countries have made significant progress in adopting AI-powered systems across different areas of the SWM value chain, contributing to improved efficiency, resource recovery, and sustainability. However, challenges related to data availability and quality, privacy, cost, infrastructure, and ethical considerations remain. Addressing these challenges requires collaboration among governments, waste management agencies, technology providers, private companies, and researchers. The study provides best-practice insights and recommendations for leveraging AI to optimise resource efficiency and support improved economic, environmental, and social outcomes in sustainable solid waste management.

Early and accurate cancer detection from medical imaging remains challenging because clinically relevant evidence is distributed across local image appearance, structural relationships between suspicious regions, and ordered imaging context. This study proposes a graph-aware sequence-aware deep learning framework for cancer analysis from medical imaging, combining a convolutional neural network (CNN) backbone for spatial feature extraction, a graph attention network (GAT) for lesion-structure modelling, and a bidirectional long short-term memory (BiLSTM) module for ordered-view or slice-sequence representation learning. Multimodal fusion is evaluated in a leakage-controlled setting only for the RSNA mammography task, where the metadata branch is restricted to patient age and implant status, both available before diagnosis. In contrast, the primary LIDC-IDRI experiment is conducted as an image-only analysis because radiologist malignancy scores and semantic nodule attributes are annotation-derived variables and are not treated as independent clinical predictors. The framework is evaluated on the RSNA Breast Cancer Detection dataset and the LIDC-IDRI lung CT dataset using accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (AUC). Additional ablation experiments assess the contribution of graph learning, sequence-aware modelling, and leakage-safe metadata fusion, while SHAP analysis is used to quantify the influence of the included RSNA metadata variables on the final multimodal predictions. The proposed framework provides an interpretable and leakage-aware architecture for integrating spatial, relational, and ordered imaging information, and demonstrates how limited pre-diagnostic metadata can be incorporated without overstating multimodal novelty or clinical realism.

Globally, Australia is ranked the fourth-top producer of e-waste. While Australian states and territories are increasingly recognising the e-waste problem and developing policies and regulations that deal with e-waste, huge amounts of e-waste still go undocumented. Currently, regional councils are concerned with the existing recycling practices owing to the escalating quantities of e-waste generated yearly. This paper examines the e-waste management practices of Victorian and Queensland regional councils, compares the current practices in managing e-waste, identifies the problems and challenges, and recommends innovative approaches to address e-waste collection and recycling practices in Australia. This study was conducted in regional Victoria and Queensland councils. A qualitative research method employing semi-structured interviews is adopted in this study. The study interviewed 49 employees of Victorian and Queensland regional councils. The findings of the study identified a number of areas that needs improvement in the management of e-waste such as: (a) lack of education and awareness about e-waste for the residents in the communities, (b) absence of knowledge about e-waste recycling (c) poor attitude towards e-waste recycling, (d) lack of sufficient funding; (e) limited number of e-waste recyclers to engage, (f) lack of specific policies and legislation on e-waste management (g) inability of the current National Television Computer Recycling Scheme (NTCRS) and mobile muster national product stewardship initiatives and schemes to cover all categories of e-waste. These results suggest that e-waste management practices in Victorian and Queensland Regional Councils need improvement. The findings of this study show that there are more similarities and fewer differences in how the Victorian and Queensland regional councils manage their e-waste. This study provides numerous practical and workable suggestions for addressing the e-waste management practices of regional councils in Australia. From a real-world scenario, this study will be useful to regional councils, environmentalists, and policymakers in identifying areas of improvement in e-waste management practices and developing impactful solutions.

Purpose: The main purpose of this study is to investigate the critical factors influencing Generation Zers′ acceptance of service robots for mental health and well‐being application. This leads to the development of a conceptual framework within the theoretical background of the theory of personality traits and health belief model. Design/Methodology/Approach: This study adopted a qualitative approach via semistructured interviews to address the research question. It was conducted in West Java, Indonesia, involving 25 participants. There were 11 women and 14 men among the participants, and the participants were 21 years old on average. Social media sites like Facebook and LinkedIn were used to recruit these participants. Findings: The results show that robot design, personality traits, and user perception have some influence on Generation Zers′ use of service robots for mental and health well‐being. Generation Zers are willing to use service robots if the service robots are user‐friendly and cheap to operate for efficiently providing the necessary information. They highlighted the importance of service robots′ features and capabilities in understanding, interpreting, and responding to their needs. This study also reveals that individuals who have these personality traits such as agreeable and open to new experiences are likely to use service robots for mental and health well‐being. In terms of health‐related behavior, Generation Zers′ use of service robots is dependent on the severity of their health condition. From the theoretical side, this study provides a validated research model for exploring the Generation Zers′ acceptance of service robots for mental health and well‐being. Practically, this study provides useful insights for enhancing user engagement, reducing barriers to care, and fostering trust in automated mental health solutions. Moreover, the findings may contribute to shaping best practices, policy frameworks, and ethical standards for the integration of service robots into mental health delivery systems.

Purpose: The performance of digitally transformed businesses relies on user interaction, with engagement’s and stickiness’s contribution to market performance well recognised. However, the negative impacts of user engagement and stickiness on consumers’ health and well-being remain largely unexplored. This study examines the role of user engagement, stickiness and self-regulation in users’ problematic usage of food delivery applications (FDAs), one of the fastest-growing smartphone apps. Design/methodology/approach: Based on a survey of 428 Chinese consumers, this study examines the relationships between user engagement, stickiness and both FDA addiction and food addiction using a PLS-SEM analysis. Findings: The results confirm the significant effects of user behavioural engagement and stickiness on FDA addiction as well as the relationship between FDA addiction and food addiction. It also identifies the insignificant moderating effect of self-regulation. Originality/value: For more ethical marketing strategies, this study highlights the importance of exploring the under-researched adverse effects of promoting user engagement and stickiness.