Activities of daily living and falls recognition and classification from the wearable sensors data

Ivașcu, Todor; Cincar, Kristijan; Diniş, Adriana; Negru, Viorel

doi:10.1109/ehb.2017.7995502

Cited by 16 publications

(11 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the ADL and fall recognition classifier model training we used several machine learning algorithms: Support Vector Machine (SVM) with both linear kernel and radial basis function (RBF) kernel, Random Forest Classifier (RF), Decision Tree Classifier (DT), Gaussian Naïve Bayes (GNB), Adaptive Boosting Classifier (AB) and K-Nearest Neighbours (KNN) algorithm with k=5. Some of these methods were selected based on results achieved by Ivascu, Cincar, Dinis and Negru implementing the same machine learning algorithms, although using different features [5]. The datasets were divided into training data (70% for the PIV dataset, and 75% for UniZg activ2 dataset) and testing data (30% for the PIV dataset, and 25% for UniZg activ2 dataset).…”

Section: Resultsmentioning

confidence: 99%

Physical Activity Recognition Based on Machine Learning

Jurčić¹,

Magjarević²

2022

Proceedings of the 29th Minisymposium

View full text Add to dashboard Cite

The following paper presents a comparison study of various machine learning techniques in recognition of activities of daily living (ADL), with special attention being given to movements during human falling and the distinction among various types of falls. The motivation for the development of physical activity recognition algorithm includes keeping track of users' activities in real-time, and possible diagnostics of unwanted and unexpected movements and/or events. The activities recorded and processed in this study include various types of daily activities, such as walking, running, etc., while fall activities include falling forward, falling backward, falling left and right (front fall, back fall and side fall). The algorithm was trained on two publicly available datasets containing signals from an accelerometer, a magnetometer and a gyroscope.

show abstract

Section: Resultsmentioning

confidence: 99%

Physical Activity Recognition Based on Machine Learning

Jurčić¹,

Magjarević²

2022

Proceedings of the 29th Minisymposium

View full text Add to dashboard Cite

show abstract

“…In medicine, predictive data analytics is a crucial challenge to improve the diagnosis and the monitoring of patients. Machine learning for predictive data analytics in medicine is now used in many fields: oncology (Adegoke et al., 2017; Ammad‐Ud‐Din et al., 2017; Armero et al., 2016; Borisov et al., 2017; Coley et al., 2017; Hoogendoorn et al., 2016; Kim & Cho, 2015; Nagarajan & Upreti, 2017; Schwartzi et al., 2015), neurology (Ertuğrul et al., 2016; Jeon et al., 2017; Khan et al., 2014; Kim et al., 2015; Kramer et al., 2017; Tripoliti et al., 2013; Xia et al., 2015; Yuvaraj et al., 2014), geriatric (Deschamps et al., 2016; Fabris et al., 2016; Ivascu et al., 2017; Kabeshova et al., 2016); Wan et al., 2015), epidemiology (Khanna & Sharma, 2018; Modu et al., 2017; Wang et al., 2016), pharmacology (Bakal et al., 2018; Bendtsen et al., 2017; Huang et al., 2017; Luo et al., 2015; Oztaner et al., 2015), … (Alghamdi et al., 2016; Delibašić et al., 2018; Hu et al., 2017; Jarmulski et al., 2018; Jing et al., 2016; Montoye et al., 2017; Oztekin et al., 2018; Saleh et al., 2017; Sanz et al., 2017).…”

Section: Introductionmentioning

confidence: 99%

Machine learning for predictive data analytics in medicine: A review illustrated by cardiovascular and nuclear medicine examples

Jamin

Ábrahám

Humeau-Heurtier

2020

Clin Physio Funct Imaging

View full text Add to dashboard Cite

With the evidence-based medicine, doctors have settings and data from which they can evaluate the risk-benefit ratio of a treatment. They can also analyse different information to make risk-based choices. With the emergence of new technologies in medicine, a large amount of data is recorded offering interesting perspectives with machine learning for predictive data analytics. However, the collection, storage and use of data raise ethical and security issues. Internet of things (IoT) is a field that groups all the systems that can be identified, perform measurements and do tasks while being open on the Internet. It is an emerging field (Stojkoska & Trivodaliev, 2017; Whitmore et al., 2015) that leads to various IoT applications in healthcare (Islam et al., 2015; Yin et al., 2016). Pavel et al. (2013) proposed to divide the future of IoT health care into four parts: patient-based applications, physician-based applications, treatment and prevention

show abstract

“…Benefiting from mobile applications and largescale data, technologies such as Blue Button (website of the Office of the National Coordinator for Health Information Technology, n.d.) and IBM Watson (IBM Watson, n.d.) help track treatment progress and reduce wrong diagnoses. In healthcare, AI covers a wide range of applications, including screening (Lin, Chang, Lin, Tsai, & Chen, 2017), monitoring (Ivascu, Cincar, & Negru, 2017), and diagnosis (Islam Chowdhuryy, Sultana, Ghosh, Ahamed, & Mahmood, 2018). Of the applied AI methods, deep learning techniques (Gharehbaghi & Lindén, 2017;Loh & Then, 2017) seem to be more adaptable, accurate, and robust in a wide range of applications and biological signals, such as lung sound classification (Chen, Zhang, Tian, Zhang, Chen, & Lei, 2016), cardiac auscultation (Amiriparian, Schmitt, M., Cummins, N., Qian, K., Dong, F., & Schuller, 2018), phonocardiography (Thomae & a https://orcid.org/0000-0002-6805-166X b https://orcid.org/0000-0003-1819-6200 Dominik, 2016), and vital sign evaluation (Jones, 2013).…”

Section: Introductionmentioning

confidence: 99%

Two-stage Artificial Intelligence Clinical Decision Support System for Cardiovascular Assessment using Convolutional Neural Networks and Decision Trees

Pasha

Lundgren

Carratù

et al. 2020

Proceedings of the 13th International Joint Conference on Biomedical Engineering Systems and Technologies

View full text Add to dashboard Cite

This paper describes an artificial-intelligence-assisted screening system implemented to support medical cardiovascular examinations performed by doctors. The proposed system is a two-stage supervised classifier comprising a convolutional neural network for heart murmur detection and a decision tree for classifying vital signs. The classifiers are trained to prioritize higher-risk individuals for more time-efficient assessment. A feature selection approach is applied to maximize classification accuracy by using only the most significant vital signs correlated with heart issues. The results suggest that the trained convolutional neural network can learn and detect heart sound anomalies from the time-domain and frequency-domain signals without using any user-guided mathematical or statistical features. It is also concluded that the proposed two-stage approach improves diagnostic reliability and efficiency.

show abstract

Activities of daily living and falls recognition and classification from the wearable sensors data

Cited by 16 publications

References 3 publications

Physical Activity Recognition Based on Machine Learning

Physical Activity Recognition Based on Machine Learning

Machine learning for predictive data analytics in medicine: A review illustrated by cardiovascular and nuclear medicine examples

Two-stage Artificial Intelligence Clinical Decision Support System for Cardiovascular Assessment using Convolutional Neural Networks and Decision Trees

Contact Info

Product

Resources

About