Article
Multidisciplinary Sciences
Tatpong Katanyukul, Pisit Nakjai
Summary: This study investigates the application of a new probabilistic interpretation of softmax output to Open-Set Recognition (OSR) in object recognition. By reinterpreting softmax inference and applying Bayes theorem, a method called Latent Cognizance (LC) is proposed for OSR, showing effectiveness in various scenarios.
Article
Engineering, Electrical & Electronic
Kasem Khalil, Ashok Kumar, Magdy Bayoumi
Summary: Hardware-based neural networks are gaining popularity due to their superior performance. This brief proposes an area-efficient implementation of an Artificial Neural Network (ANN) by reducing the number of layers through a novel use of hidden layers. The proposed design achieves high hardware utilization and cost reduction.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS
(2022)
Article
Engineering, Electrical & Electronic
Ze Li, Yong Qian, Hui Wang, Xiaoli Zhou, Gehao Sheng, Xiuchen Jiang
Summary: This study introduces a partial discharge image feature extraction method based on upright speeded-up robust features, and improves the accuracy of partial discharge pattern recognition by using an improved support vector machine. The results show that upright speeded-up robust features can effectively handle noisy data.
IET GENERATION TRANSMISSION & DISTRIBUTION
(2022)
Article
Computer Science, Information Systems
Jaechan Cho, Yongchul Jung, Seongjoo Lee, Yunho Jung
Summary: A BNN accelerator with adaptive parallelism is proposed, offering high throughput performance and higher area-speed efficiency. By analyzing target layer parameters and operating with optimal parallelism using reasonable resources, this accelerator can achieve high efficiency in all layers.
Article
Engineering, Biomedical
Yang Zhou, Chaoyang Chen, Mark Cheng, Yousef Alshahrani, Sreten Franovic, Emily Lau, Guanghua Xu, Guoxin Ni, John M. Cavanaugh, Stephanie Muh, Stephen Lemos
Summary: This study investigated the use of machine learning algorithms such as SVM, LR, and ANN for shoulder motion pattern recognition based on sEMG signals. Results showed that SVM method with a sliding time window of 270 ms achieved the highest accuracy.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2021)
Review
Chemistry, Multidisciplinary
Karolina Kudelina, Toomas Vaimann, Bilal Asad, Anton Rassolkin, Ants Kallaste, Galina Demidova
Summary: This paper reviews the fault diagnostic techniques based on machine learning, highlighting the increasing capability of using cloud computation for processing faulty data and the potential of utilizing mathematical models of electrical machines for training AI algorithms in the era of industry 4.0.
APPLIED SCIENCES-BASEL
(2021)
Article
Chemistry, Analytical
Fabian Hoitz, Laura Fraeulin, Vinzenz von Tscharner, Daniela Ohlendorf, Benno M. Nigg, Christian Maurer-Grubinger
Summary: Human movement patterns are as unique as fingerprints, with certain characteristics being more important for machine learning algorithms to distinguish individuals. By applying layer-wise relevance propagation to neural networks, it was found that movement patterns defined by important characteristics showed minimal overlap, while those defined by less important characteristics had substantial overlap. Elite runners were identified by specific sagittal plane movements during mid-stance and mid-swing, while generic characteristics were observed during early and late stance.
Article
Chemistry, Multidisciplinary
Giuseppe Ciaburro, Gino Iannace, Virginia Puyana-Romero, Amelia Trematerra
Summary: The study measured the noise emitted by a wind turbine near a sensitive receptor and used average spectral levels to identify the operating conditions of the turbine. Models based on support vector machines and artificial neural networks were developed and found to be useful tools for supporting the acoustic characterization of noise in environments close to wind turbines.
APPLIED SCIENCES-BASEL
(2021)
Review
Engineering, Electrical & Electronic
Xingan Yang, Meng Li, Xiaohua Ji, Junqing Chang, Zanhong Deng, Gang Meng
Summary: This article introduces the rapid applications of the smart electronic nose (E-nose) in various fields and emphasizes the role of recognition algorithms in its performance. The traditional algorithms and artificial neural networks (ANNs)-based algorithms are analyzed in detail, along with the evaluation metrics and challenges.
IEEE SENSORS JOURNAL
(2023)
Article
Energy & Fuels
Abdul Manan, Khurram Kamal, Tahir Abdul Hussain Ratlamwala, Muhammad Fahad Sheikh, Abdul Ghani Abro, Tayyab Zafar
Summary: In this study, gas pipeline incidents from 2002 to 2020 in the US were analyzed to predict different types of failures using Artificial Neural Networks and Support Vector Machine. The study identified that the Medium Gaussian SVM integrated with ANOVA and Holdout cross-validation performed better than other algorithms with 74.8% accuracy.
Article
Computer Science, Interdisciplinary Applications
Chih-Jui Yu, Hsing-Jung Yeh, Chun-Chao Chang, Jui-Hsiang Tang, Wei-Yu Kao, Wen-Chao Chen, Yi-Jin Huang, Chien-Hung Li, Wei-Hao Chang, Yun-Ting Lin, Herdiantri Sufriyana, Emily Chia-Yu Su
Summary: A machine learning system was developed to detect and localize gallstones, and to detect cholecystitis, with acceptable discrimination and speed. The system showed good performance with acceptable judgment and speed using specific algorithms.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2021)
Review
Computer Science, Information Systems
Zengchen Yu, Ke Wang, Zhibo Wan, Shuxuan Xie, Zhihan Lv
Summary: Due to its automatic feature learning ability and high performance, deep learning has gradually become the mainstream of artificial intelligence in recent years, playing a role in many fields. This paper introduces several deep learning algorithms such as Artificial Neural Network (NN), FM-Deep Learning, Convolutional NN and Recurrent NN, and explains their theory, development history, and applications in disease prediction. The paper also analyzes the current defects in the disease prediction field and provides some current solutions. Furthermore, it discusses two major trends in the future disease prediction and medical field - integrating Digital Twins and promoting precision medicine.
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS
(2023)
Article
Neurosciences
Leonie Borne, Denis Riviere, Arnaud Cachia, Pauline Roca, Charles Mellerio, Catherine Oppenheim, Jean-Francois Mangin
Summary: The study explores the relationship between local cortical folding patterns and psychiatric illnesses as well as cognitive functions. While manually classifying local sulcal patterns is time-consuming and challenging, the development of automatic classification algorithms is proposed to improve efficiency and reliability. Three different methods are tested on challenging patterns, showing promising results for ACC patterns and PBS classification.
Review
Computer Science, Artificial Intelligence
Himanshu Kumar, A. Martin
Summary: This paper discusses the application of emotion recognition in the field of artificial intelligence, focusing on its limitations and challenges. It compares the latest machine learning and deep learning algorithms and evaluates their accuracy and effectiveness in emotion recognition. The study demonstrates that hybrid classification techniques achieve a balance between accuracy and efficiency in speech emotion recognition.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Engineering, Mechanical
Julian N. Heidenreich, Maysam B. Gorji, Dirk Mohr
Summary: The use of micromechanics and homogenization theory allows for predicting the effective mechanical properties of materials based on microstructural information. The microstructural information is encoded in images using convolutional neural networks (CNN), through which geometric information is reduced to ten characteristic features. A fully connected neural network model is introduced to predict the effective yield surfaces based on the encoded information, resulting in a computationally efficient CNN-FCNN model.
INTERNATIONAL JOURNAL OF PLASTICITY
(2023)
Editorial Material
Nutrition & Dietetics
Antonio Narzisi, Gabriele Masi, Enzo Grossi
Article
Clinical Neurology
Antonella Gagliano, Monica Puligheddu, Nadia Ronzano, Patrizia Congiu, Marcello Giuseppe Tanca, Ida Cursio, Sara Carucci, Stefano Sotgiu, Enzo Grossi, Alessandro Zuddas
Summary: The study highlights the high prevalence of sleep disorders in patients with PANS and their connections with other symptoms such as periodic limb movement disorder and REM-sleep without atonia. Disordered sleep plays a significant role in PANS patients and could potentially be included in major diagnostic criteria considering its impact on diagnosis and prognosis.
NATURE AND SCIENCE OF SLEEP
(2021)
Article
Neurosciences
Enzo Grossi, Elisa Caminada, Michela Goffredo, Beatrice Vescovo, Tristana Castrignano, Daniele Piscitelli, Giulio Valagussa, Marco Franceschini, Franco Vanzulli
Summary: This study used video recordings to analyze stereotypic behavior in individuals with Autism Spectrum Disorder, finding a correlation between the number of stereotypies and ASD severity. Individuals with fewer stereotypies were associated with lower ASD severity, while there were no significant differences in ASD severity between individuals exhibiting simple versus complex stereotypies. The study provides a data-driven approach to better understand the pathophysiology and management of restricted, repetitive behaviors.
Article
Endocrinology & Metabolism
Emanuele Garzia, Valentina Galiano, Giovanni Marfia, Stefania Navone, Enzo Grossi, Anna Maria Marconi
Summary: The aim of this study was to identify reliable predictors of response to metformin therapy for weight loss and reduction in plasma androgen levels using artificial neural networks (ANNs). The results showed that in women with PCOS, menstrual pattern imbalance and ovarian androgens excess are the best predictors of metformin response. Baseline plasma testosterone level can serve as a sensitive marker to predict treatment compliance.
REPRODUCTIVE BIOLOGY AND ENDOCRINOLOGY
(2022)
Article
Economics
Paolo Massimo Buscema, Francesca Della Torre, Giulia Massini, Guido Ferilli, Pier Luigi Sacco
Summary: This paper studies the interdependences between stock market indexes in 30 different countries, revealing complex dynamic properties not previously found in existing literature. A toolkit of ANN-based methods is presented for analyzing various aspects of these dynamics.
COMPUTATIONAL ECONOMICS
(2023)
Article
Genetics & Heredity
Andrea Stoccoro, Roberta Gallo, Sara Calderoni, Romina Cagiano, Filippo Muratori, Lucia Migliore, Enzo Grossi, Fabio Coppede
Summary: This study used artificial neural networks to examine the connections among blood gene methylation levels, sex, maternal risk factors, and symptom severity in children with autism spectrum disorder (ASD). The results showed that the methylation levels of MECP2, HTR1A, and OXTR genes were associated with females, while the methylation levels of EN2, BCL2, and RELN genes were associated with males. High gestational weight gain, lack of folic acid supplements, advanced maternal age, preterm birth, low birthweight, and living in a rural context were the best predictors of a high ADOS-2 score.
Article
Pediatrics
Giulio Valagussa, Giulia Purpura, Alessandra Nale, Rita Pirovano, Miryam Mazzucchelli, Enzo Grossi, Cecilia Perin
Summary: Atypical sensory processing and motor impairments are commonly observed in individuals with autism spectrum disorders (ASD). Tip-toe behavior (TTB) is a possible clinical finding, but its cause is not well understood. This pilot study found that individuals with ASD and TTB showed a specific pattern of under responsive/seeks sensation on a sensory profile assessment, suggesting that sensory-motor features may be important when rehabilitating TTB in individuals with ASD.
Article
Medicine, General & Internal
Monica Puligheddu, Patrizia Congiu, Michela Figorilli, Ludovica Tamburrino, Patrizia Pisanu, Roberta Coa, Maria Giuseppina Mascia, Davide Fonti, Rosamaria Lecca, Enzo Grossi, Antonella Gagliano
Summary: This study aimed to assess whether sleep instability affects cognitive functions in patients with Disorder of arousal (DOA) or sleep-related hypermotor epilepsy (SHE). The results showed that the SHE group had reduced sleep efficiency and increased wake after sleep onset (WASO); both the SHE and DOA groups showed increased % of N2 and REM sleep compared to the healthy control (HC) group. Neuropsychological and behavioral evaluations showed a different cognitive profile in the SHE group. These findings support the discriminative power of cognitive and psychiatric assessment in SHE and DOA.
JOURNAL OF CLINICAL MEDICINE
(2023)
Article
Biochemistry & Molecular Biology
Fernanda Velluzzi, Andrea Deledda, Mauro Lombardo, Michele Fosci, Roberto Crnjar, Enzo Grossi, Giorgia Sollai
Summary: This study explores the connection between olfactory function and clinical and nutritional aspects related to overweight or obesity. The results show that 65% of patients exhibit hyposmia. Olfactory scores are negatively correlated with systolic blood pressure, fasting plasma glucose, and triglycerides levels, but positively correlated with Mediterranean diet scores. Olfactory function is associated with obesity, metabolic disorders, and male gender, while normosmia is linked to adherence to the Mediterranean diet, normal blood pressure, lipids, and glucose levels.
Article
Genetics & Heredity
Andrea Stoccoro, Vanessa Nicoli, Fabio Coppede, Enzo Grossi, Giorgio Fedrizzi, Simonetta Menotta, Francesca Lorenzoni, Marta Caretto, Arianna Carmignani, Sabina Pistolesi, Ernesto Burgio, Vassilios Fanos, Lucia Migliore
Summary: Exposure to environmental stressors during pregnancy can influence susceptibility to chronic diseases through epigenetic mechanisms. This study used artificial neural networks (ANNs) to explore the links between gestational environmental exposures and DNA methylation in placental and buccal cells. The results showed associations between birth weight and placental H19 methylation, maternal stress during pregnancy and methylation levels of NR3C1 and BDNF in placenta and buccal cells respectively, and air pollutant exposure and maternal MGMT methylation. Associations were also found between placental metal concentrations and methylation levels of OXTR, HSD11B2, MECP2, and MTHFR in different cells. Dioxin concentrations were associated with methylation levels of RELN, HSD11B2, and H19 in placenta and neonatal buccal cells. These findings suggest that environmental stressors during pregnancy can lead to abnormal methylation levels in genes related to embryogenesis and may serve as biomarkers of environmental exposure.
Article
Geosciences, Multidisciplinary
Paolo Massimo Buscema, Weldon A. Lodwick, Masoud Asadi-Zeydabadi, Francis Newman, Marco Breda, Riccardo Petritoli, Giulia Massini, David Buscema, Donatella Dominici, Fabio Radicioni
Summary: The Twisting Theory (TWT) and Crown Clustering Algorithm (CCA) are innovative adaptive algorithms that can determine the shape of a landslide and predict its future evolution based on the movement of position sensors located in the affected area. These algorithms were thoroughly explained and applied to real-life cases, achieving a high correlation between model estimates and expert's measurements. The results demonstrated the accuracy of TWT in identifying landslide shapes and predicting progression, while CCA effectively represented complex cause-and-effect relationships among sensors. The operational feasibility of this model, which only requires the installation of GNSS sensors in a predetermined grid, is crucial for its wider application to secure landslide-prone regions.
Article
Chemistry, Multidisciplinary
Alfredo Raglio, Enzo Grossi, Luca Manzoni
Summary: Music listening is widely used in therapeutic interventions, but understanding the relationship between musical elements and their effects is complex. This study analyzed data from a relaxation-focused music listening experiment and found potential relationships between variables. The research identified different relaxing effects for different music types and populations.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Artificial Intelligence
Paolo Massimo Buscema, Giulia Massini, Giovanbattista Raimondi, Giuseppe Caporaso, Marco Breda, Riccardo Petritoli
Summary: This paper proposes a vessel type recognition method based on the TOCAT classification strategy, which uses a collection of adaptive models trained with radar data. The models consider factors such as identifiers, velocity, and heading to achieve an average accuracy of 83% on a 6-class classification task.
Article
Rheumatology
E. Grossi
Summary: Despite medical research being predominantly male-focused, efforts have been made to overcome this gender bias. Retrospective examination of 21 datasets found that gender information was included in the predictive model 19 out of 21 times, highlighting its importance even in highly adaptive tools like Artificial Neural Networks. The study also showed that gender information significantly improved the accuracy of predicting psoriatic arthritis diagnoses using ANNs.
CLINICAL AND EXPERIMENTAL RHEUMATOLOGY
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Fabio Massimo Ulivieri, Luca Rinaudo, Carmelo Messina, Luca Petruccio Piodi, Davide Capra, Barbara Lupi, Camilla Meneguzzo, Luca Maria Sconfienza, Francesco Sardanelli, Andrea Giustina, Enzo Grossi
Summary: The study applied an artificial intelligence model to predict fragility fractures in postmenopausal women, finding that femoral BSI is a useful DXA index for identifying patients at lower risk for lumbar VFs.
EUROPEAN RADIOLOGY EXPERIMENTAL
(2021)