Browse Articles
Discover research articles across all indexed journals
Detection of human activities using multi-layer convolutional neural network
Abstract Human Activity Recognition (HAR) plays a critical role in fields such as healthcare, sports, and human-computer interaction. However, achieving high accuracy and robustness remains a challenge, particularly when dealing with noisy sensor data from accelerometers and gyroscopes. This paper introduces HARCNN, a novel approach leveraging Convolutional Neural Networks (CNNs) to extract hierarchical spatial and temporal features from raw sensor data, enhancing activity recognition performance. The HARCNN model is designed with 10 convolutional blocks, referred to as “ConvBlk.” Each block integrates a convolutional layer, a ReLU activation function, and a batch normalization layer. The outputs from specific blocks “ConvBlk_3 and ConvBlk_4,” “ConvBlk_6 and ConvBlk_7,” and “ConvBlk_9 and ConvBlk_10” are fused using a depth concatenation approach. The concatenated outputs are subsequently passed through a 2 × 2 max-pooling layer with a stride of 2 for further processing. The proposed HARCNN framework is evaluated using accuracy, precision, sensitivity, and f-score as key metrics, reflecting the model’s ability to correctly classify and differentiate between human activities. The proposed model’s performance is compared to traditional pre-trained Convolutional Neural Networks (CNNs) and other state-of-the-art techniques. By leveraging advanced feature extraction and optimized learning strategies, the proposed model demonstrates its efficacy in achieving accuracy of 97.87%, 99.12%, 96.58%, and 98.51% for various human activities datasets; UCI-HAR, KU-HAR, WISDM, and HMDB51, respectively. This comparison underscores the model’s robustness, highlighting improvements in minimizing false positives and false negatives, which are crucial for real-world applications where reliable predictions are essential. The experiments were conducted with various window sizes (50ms, 100ms, 200ms, 500ms, 1s, and 2s). The results indicate that the proposed method achieves high accuracy and reliability across these different window sizes, highlighting its ability to adapt to varying temporal granularities without significant loss of performance. This demonstrates the method’s effectiveness and robustness, making it well-suited for deployment in diverse HAR scenarios. Notably, the best results were obtained with a window size of 200ms.
Computational analysis of curved prestressed concrete box-girder bridges using finite element method
Spatio-temporal analysis of urban expansion and land use dynamics using google earth engine and predictive models
Artificial intelligence models predicting abnormal uterine bleeding after COVID-19 vaccination
Formation mechanism and evaluation of geothermal resources in Yanqi Qikexing town, Xinjiang
Mycoplasma pneumoniae MLST detected in the upsurge of pneumonia during the 2023 to 2024 winter season in the Netherlands
The neuropeptidomes of the sea cucumbers Stichopus cf. horrens and Holothuria scabra
Syntonic phototherapy versus part time occlusion for treatment of refractive amblyopia
Abstract To evaluate the effectiveness of syntonic phototherapy and compare it with partial time occlusion to improve visual acuity in cases of refractive amblyopia. This study is a prospective, comparative, and randomized study. It included 40 patients. Their mean age ± SD was 14.45 ± 10.03 years (Range: 6–45 years). Twenty patients were subjected to partial time occlusion of the sound eye, and 20 received syntonic phototherapy treatment.The study revealed that there was statistically significant improvement in visual functions, UCVA, BCVA, AOP, and functional visual field in patients who were subjected to syntonic phototherapy, whereas the improvement of UCVA and BCVA in patients who were subjected to conventional treatment was satisfactory but less than that reported by syntonic therapy. Visual acuity increased significantly in patients with amblyopia after syntonic phototherapy as compared to partial occlusion therapy.
Machine learning analysis of gene expression profiles of pyroptosis-related differentially expressed genes in ischemic stroke revealed potential targets for drug repurposing
Development of a novel postoperative adhesion induction model in cynomolgus monkeys with high reliability and reproducibility
Abstract Postoperative adhesions frequently occur following abdominal surgical interventions, leading to serious morbidities and requiring new therapeutic strategies. The development of new therapeutic agents to reduce postoperative adhesions needs animal models that closely mirror human pathophysiology. In this study, we established a novel surgical adhesion model in cynomolgus monkeys, which are characteristically similar to humans. Our model reliably and reproducibly developed adhesions. Histopathological analyses revealed that monkeys undergoing our novel surgery method exhibited changes consistent with those in monkeys that underwent open abdominal surgery. Furthermore, the cellular components of the adhesion tissue in our monkey model reflected those reported in human adhesion tissue. Furthermore, time-course transcriptomic analyses showed that our model accurately recapitulates the well-known progression cascade of postoperative adhesions. In addition, it identified the upregulation of gene that is absent in rodents. We expect our novel surgical method to be a promising tool for elucidating the detailed biology of postoperative adhesions and for assessing new therapeutic treatments with high translatability to human biology.