Fusion of Unobtrusive Sensing Solutions for Home-Based Activity Recognition and Classification Using Data Mining Models and Methods
AuthorsEkerete, Idongesit; orcid: 0000-0003-4072-7178; email: firstname.lastname@example.org
Garcia-Constantino, Matias; orcid: 0000-0002-3420-0532; email: email@example.com
Konios, Alexandros; email: firstname.lastname@example.org
Mustafa, Mustafa A.; email: email@example.com
Diaz-Skeete, Yohanca; email: firstname.lastname@example.org
Nugent, Christopher; orcid: 0000-0003-0882-7902; email: email@example.com
McLaughlin, James; email: firstname.lastname@example.org
MetadataShow full item record
AbstractThis paper proposes the fusion of Unobtrusive Sensing Solutions (USSs) for human Activity Recognition and Classification (ARC) in home environments. It also considers the use of data mining models and methods for cluster-based analysis of datasets obtained from the USSs. The ability to recognise and classify activities performed in home environments can help monitor health parameters in vulnerable individuals. This study addresses five principal concerns in ARC: (i) users’ privacy, (ii) wearability, (iii) data acquisition in a home environment, (iv) actual recognition of activities, and (v) classification of activities from single to multiple users. Timestamp information from contact sensors mounted at strategic locations in a kitchen environment helped obtain the time, location, and activity of 10 participants during the experiments. A total of 11,980 thermal blobs gleaned from privacy-friendly USSs such as ceiling and lateral thermal sensors were fused using data mining models and methods. Experimental results demonstrated cluster-based activity recognition, classification, and fusion of the datasets with an average regression coefficient of 0.95 for tested features and clusters. In addition, a pooled Mean accuracy of 96.5% was obtained using classification-by-clustering and statistical methods for models such as Neural Network, Support Vector Machine, K-Nearest Neighbour, and Stochastic Gradient Descent on Evaluation Test.
CitationApplied Sciences, volume 11, issue 19, page e9096
DescriptionFrom MDPI via Jisc Publications Router
History: accepted 2021-09-24, pub-electronic 2021-09-29
Publication status: Published
Funder: interreg VA; Grant(s): IVA5034
Showing items related by title, author, creator and subject.
Is the Morphological Subtype of Extra-Pulmonary Neuroendocrine Carcinoma Clinically Relevant?Frizziero, Melissa; email: email@example.com; Durand, Alice; orcid: 0000-0002-0193-9058; email: firstname.lastname@example.org; Taboada, Rodrigo G.; orcid: 0000-0001-5674-2669; email: email@example.com; Zaninotto, Elisa; email: firstname.lastname@example.org; Luchini, Claudio; orcid: 0000-0003-4901-4908; email: email@example.com; Chakrabarty, Bipasha; email: firstname.lastname@example.org; Hervieu, Valérie; email: email@example.com; Claro, Laura C. L.; email: firstname.lastname@example.org; Zhou, Cong; orcid: 0000-0002-6938-4685; email: email@example.com; Cingarlini, Sara; email: firstname.lastname@example.org; et al. (MDPI, 2021-08-18)Extra-pulmonary neuroendocrine carcinomas (EP-NECs) are lethal cancers with limited treatment options. Identification of contributing factors to the observed heterogeneity of clinical outcomes within the EP-NEC family is warranted, to enable identification of effective treatments. A multicentre retrospective study investigated potential differences in “real-world” treatment/survival outcomes between small-cell (SC) versus (vs.) non-SC EP-NECs. One-hundred and seventy patients were included: 77 (45.3%) had SC EP-NECs and 93 (54.7%) had non-SC EP-NECs. Compared to the SC subgroup, the non-SC subgroup had the following features: (1) a lower mean Ki-67 index (69.3% vs. 78.7%; p = 0.002); (2) a lower proportion of cases with a Ki-67 index of ≥55% (73.9% vs. 88.7%; p = 0.025); (3) reduced sensitivity to first-line platinum/etoposide (objective response rate: 31.6% vs. 55.1%, p = 0.015; and disease control rate; 59.7% vs. 79.6%, p = 0.027); (4) worse progression-free survival (PFS) (adjusted-HR = 1.615, p = 0.016) and overall survival (OS) (adjusted-HR = 1.640, p = 0.015) in the advanced setting. Within the advanced EP-NEC cohort, subgroups according to morphological subtype and Ki-67 index (55% vs. ≥55%) had significantly different PFS (adjusted-p = 0.021) and OS (adjusted-p = 0.051), with the non-SC subgroup with a Ki-67 index of 55% and non-SC subgroup with a Ki-67 index of ≥55% showing the best and worst outcomes, respectively. To conclude, the morphological subtype of EP-NEC provides complementary information to the Ki-67 index and may aid identification of patients who could benefit from alternative first-line treatment strategies to platinum/etoposide.
The Interdependency and Co-Regulation of the Vitamin D and Cholesterol MetabolismWarren, Tara; email: email@example.com; McAllister, Roisin; email: firstname.lastname@example.org; Morgan, Amy; email: email@example.com; Rai, Taranjit Singh; email: firstname.lastname@example.org; McGilligan, Victoria; email: email@example.com; Ennis, Matthew; email: Ennis-M4@ulster.ac.uk; Page, Christopher; email: Page-C7@ulster.ac.uk; Kelly, Catriona; email: firstname.lastname@example.org; Peace, Aaron; orcid: 0000-0001-9556-7509; email: Aaron.Peace@westerntrust.hscni.net; Corfe, Bernard M.; email: Bernard.Corfe@newcastle.ac.uk; et al. (MDPI, 2021-08-06)Vitamin D and cholesterol metabolism overlap significantly in the pathways that contribute to their biosynthesis. However, our understanding of their independent and co-regulation is limited. Cardiovascular disease is the leading cause of death globally and atherosclerosis, the pathology associated with elevated cholesterol, is the leading cause of cardiovascular disease. It is therefore important to understand vitamin D metabolism as a contributory factor. From the literature, we compile evidence of how these systems interact, relating the understanding of the molecular mechanisms involved to the results from observational studies. We also present the first systems biology pathway map of the joint cholesterol and vitamin D metabolisms made available using the Systems Biology Graphical Notation (SBGN) Markup Language (SBGNML). It is shown that the relationship between vitamin D supplementation, total cholesterol, and LDL-C status, and between latitude, vitamin D, and cholesterol status are consistent with our knowledge of molecular mechanisms. We also highlight the results that cannot be explained with our current knowledge of molecular mechanisms: (i) vitamin D supplementation mitigates the side-effects of statin therapy; (ii) statin therapy does not impact upon vitamin D status; and critically (iii) vitamin D supplementation does not improve cardiovascular outcomes, despite improving cardiovascular risk factors. For (iii), we present a hypothesis, based on observations in the literature, that describes how vitamin D regulates the balance between cellular and plasma cholesterol. Answering these questions will create significant opportunities for advancement in our understanding of cardiovascular health.
Hermetia illucens (L.) (Diptera: Stratiomyidae) Odorant Binding Proteins and Their Interactions with Selected Volatile Organic Compounds: An in Silico ApproachScieuzo, Carmen; email: email@example.com; Nardiello, Marisa; email: firstname.lastname@example.org; Farina, Donatella; email: email@example.com; Scala, Andrea; email: firstname.lastname@example.org; Cammack, Jonathan A.; email: email@example.com; Tomberlin, Jeffery K.; orcid: 0000-0002-0560-4466; email: firstname.lastname@example.org; Vogel, Heiko; orcid: 0000-0001-9821-7731; email: email@example.com; Salvia, Rosanna; orcid: 0000-0002-6661-7164; email: firstname.lastname@example.org; Persaud, Krishna; email: email@example.com; Falabella, Patrizia; orcid: 0000-0003-0304-6867; email: firstname.lastname@example.org (MDPI, 2021-09-11)The black soldier fly (BSF), Hermetia illucens (Diptera: Stratiomyidae), has considerable global interest due to its outstanding capacity in bioconverting organic waste to insect biomass, which can be used for livestock, poultry, and aquaculture feed. Mass production of this insect in colonies requires the development of methods concentrating oviposition in specific collection devices, while the mass production of larvae and disposing of waste may require substrates that are more palatable and more attractive to the insects. In insects, chemoreception plays an essential role throughout their life cycle, responding to an array of chemical, biological and environmental signals to locate and select food, mates, oviposition sites and avoid predators. To interpret these signals, insects use an arsenal of molecular components, including small proteins called odorant binding proteins (OBPs). Next generation sequencing was used to identify genes involved in chemoreception during the larval and adult stage of BSF, with particular attention to OBPs. The analysis of the de novo adult and larval transcriptome led to the identification of 27 and 31 OBPs for adults and larvae, respectively. Among these OBPs, 15 were common in larval and adult transcriptomes and the tertiary structures of 8 selected OBPs were modelled. In silico docking of ligands confirms the potential interaction with VOCs of interest. Starting from the information about the growth performance of H. illucens on different organic substrates from the agri-food sector, the present work demonstrates a possible correlation between a pool of selected VOCs, emitted by those substrates that are attractive for H. illucens females when searching for oviposition sites, as well as phagostimulants for larvae. The binding affinities between OBPs and selected ligands calculated by in silico modelling may indicate a correlation among OBPs, VOCs and behavioural preferences that will be the basis for further analysis.