Jul 2026· International Conference on Signal Processing and Communications· pp. 1-6· 0 citations· 29 references
Abstract
Alzheimer's disease affects millions of older adults worldwide, with wandering and fall-related injuries ranking among the most preventable yet dangerous consequences of cognitive decline. Current assistive technologies are largely reactive, alerting caregivers only after an incident occurs rather than anticipating the behavioral patterns that precede it. This paper presents a wearable IoT system that addresses this gap through continuous nine-axis inertial sensing, low-power Lo-RaWAN transmission, and a cloud-hosted inference pipeline for real-time fall detection. We evaluate four offline deep learning architectures (LSTM, GRU, 1D-CNN, and Transformer) alongside three online adaptive classifiers on the KFall benchmark dataset comprising 189,987 sliding windows with a 7.56% fall ratio. The stateless LSTM achieves the highest F1-score of 0.8893 with zero missed falls, while the 1D-CNN records the best precision at 0.9507 and the lowest false-alarm count. Beyond detection, the system incorporates a Retrieval-Augmented Generation (RAG) module backed by a Pinecone vector store and BART-Large generation model, enabling caregivers to query patient history and clinical guidance through natural language. All sensor payloads are protected with AES-256 encryption prior to transmission. The results show that pairing non-invasive wearable sensing with sequence models and a conversational caregiver interface produces a practical platform for keeping Alzheimer's patients safer.
Falls and irregular heart rhythms are the main causes of injury among kids and the elderly, always overwhelming healthcare systems, and thus making privacy-aware, real-time monitoring a necessity. This work unveils a TinyML wrist-worn prototype based on Arduino Nano 33 BLE Sense, which combines the MPU6050 IMU for motion-based fall detection and activity (walking, sitting, running, lying) recognition with the MAX30102 PPG for heartbeat, SpO2, and HRV anomaly detection over generations.
The device, tested on 7 subjects (3 children 8-12 years, 2 adults 25-40, and 2 seniors 65-75) for 140 real-life sequences in a lab in Kerala, uses Butterworth-filtered data, 56 temporal features extracted from 256-sample windows, and the optimized hybrid CNN-LSTM model (65% structured pruning, 8-bit QAT) to perform inference on the edge under 217KB flash. Dual-threshold triggering (fall confidence >0.9 plus HR anomalies or SpO2<92%) allows BLE alerts within 100ms to caregiver apps, and cancellation via 30s haptic/button helps reduce the false alarms.
Field experiments demonstrated the device performance with 94.3% accuracy, 0.95 fall F1-score, 38ms latency, 0.7mW power, and 2.1% false positives, showing a significant improvement of 15% F1 when compared against unimodal baselines, while being fully processed on the edge, GDPR-compliant, and with a multi-day battery life, the device is ready for wide deployment in homes, schools, and care facilities. This work is a step forward in TinyML across demographics, thus opening the gate to multimodal extensions such as cry detection.
P. Prathap, H. Hibafathima, Sufaira Shamsudeen et al.· ITEGAM- Journal of Engineeri...· 0 citations
Elderly fall detection and human activity recognition represent important components of smart healthcare monitoring systems due to the increasing aging population and the growing incidence of fall-related injuries among older adults. Falls frequently result in fractures, mobility impairment, hospitalization, and long-term health complications, thereby increasing the demand for intelligent monitoring technologies capable of continuous supervision and rapid emergency identification. Conventional elderly monitoring approaches primarily rely on wearable sensors, vision-based surveillance systems, inertial sensing devices, and Internet of Things (IoT)-enabled healthcare frameworks for recognizing daily activities and detecting abnormal fall events. Although these systems provide automated monitoring capability, several frameworks experience limitations associated with environmental variation, occlusion, computational complexity, communication instability, privacy concerns, and reduced generalization across real-world healthcare environments. Recent advancements in Artificial Intelligence (AI), Machine Learning (ML), Deep Learning (DL), hybrid AI, and ensemble learning significantly improved automated activity recognition and fall detection performance through intelligent spatial-temporal feature extraction and multimodal data analysis. This study presents a comprehensive review of recent ML-based, DL-based, hybrid, and ensemble AI-driven frameworks developed for elderly fall detection and activity recognition in smart monitoring environments. The review analyzes wearable, vision-based, IoT-enabled, edge-computing, radar-based, and multimodal monitoring systems along with their datasets, methodologies, performance metrics, and identified limitations. Such AI-integrated monitoring systems possess significant potential to improve patient safety, strengthen continuous healthcare supervision, reduce emergency response delays, and support smart assisted living infrastructures.
Lisha M, L. Parthiban· 2026 7th International Confe...· 0 citations
Falls among older adults represent a critical public health challenge, with approximately 37.3 million fall-related incidents reported globally each year. Early and accurate prediction of falls is essential to enable timely, proactive interventions and to reduce associated injuries and fatalities. This work introduces a graph-based machine learning framework that leverages data from the cStick, a smart assistive Internet of Medical Things (IoMT) device. Bipartite graphs are constructed to model static correlations between multivariate sensor inputs— including heart rate variability (HRV), pressure, distance, SpO2, blood sugar levels, and accelerometer readings— and fall outcomes encoded as no fall, predicted fall, or definite fall. SHAP (SHapley Additive exPlanations) values are further integrated to enhance model interpretability and to identify the most influential sensor features through feature-only graph projections. Kernel Density Estimation (KDE) plots and pairplots are used to visualize feature distributions across fall categories. The proposed framework demonstrates that Pressure and Distance exhibit the strongest correlations with fall decisions (1.000 and −0.946, respectively), providing actionable insights for risk stratification. The integration of graph-based analysis with SHAP interpretability improves both predictive accuracy and transparency, facilitating proactive interventions and enhancing the safety, autonomy, and well-being of elderly individuals in real-world assistive care settings.
Krishna Kumari R., P. V· International Journal of Eng...· 0 citations
Falls are a significant health issue for older people, and can result in injury, disability, hospitalisation and loss of independence. Continuous monitoring and quick caregiver notification can be enabled using an AI-powered wearable IoT system. Objective: The purpose of this study was to design and test an edge-based wearable IoT framework for real-time fall detection based on the accelerometer and gyroscope signals. Methodology: A synthetic dataset was created by using the features of SisFall and MobiAct, including 270 falls and 630 activities of daily living from 30 simulated participants. Various temporal, statistical and frequency domain features were extracted from the sensor signals and these were filtered, normalized, segmented. Participant-wise Training and Testing were performed on the Random Forest and Support Vector Machine Models. The architecture of ESP32–IMU and the alert mechanism using MQTT was designed. On the synthetic test data, both models obtained a 100% accuracy, sensitivity, specificity, precision and F1-score, while the AUC is 1.000 for both models. Discussion: The results show technical feasibility, and probably represent good synthetic class separation, and should not be used as a measure of clinical performance. Discussion: The framework provides a good base for the private, rapid fall detection but still needs to be validated in the real world, tested with hardware, put through with a variety of elderly participants and evaluated for usability.
Chandrani Mukherjee· Journal of Intelligent Decis...· 0 citations
Falls among older adults result in over 3 million emergency department visits and 32,000 deaths annually in the United States, with medical costs exceeding $50 billion. Automated fall detection systems face a fundamental challenge: extreme class imbalance in continuous monitoring, where fall events constitute less than 2% of footage, causing high false-positive rates and alarm fatigue. We propose a cascade architecture that decomposes fall detection into two sequential stages with complementary optimization objectives. Stage 1 employs a lightweight 3D convolutional neural network trained via curriculum learning with adaptive focal loss to deliberately maximize recall on the minority class. Stage 2 applies a feature-based arbiter that analyzes CNN internal representations to validate candidates, reducing false positives while preserving sensitivity. We validated our approach on 190 videos with simulated falls (LE2I dataset) and 300 spontaneous falls from elderly residents in long-term care facilities (Robinovitch dataset). Our cascade architecture achieved 82.4% recall with 32.2% precision, representing a 38% relative precision improvement over single-stage CNN baselines (88.5% recall, 23.3% precision). The 40.5% false positive reduction substantially improves clinical viability while maintaining sensitivity above 80%, providing a foundation for real-world deployment in care settings facing severe data constraints.
Ethan S. Henley, N. E. Stark, Brianna A. Reilly et al.· Scientific Reports· 0 citations
Falls in geriatric patients continue to pose a significant source of harm, hospitalization, worsening of function, and impaired autonomy, prompting the development of smart technologies that no longer rely on reactive fall detection methods to move towards proactive predictions and prevention. A new application-oriented multimodal fall analytics framework that utilizes two supplementary sets of data, namely, IMU-based wearable data to detect binary falls and smart cStick IoMT data to distinguish between no fall, fall predicted, and definite fall is introduced in the paper. The suggested system incorporates motion-derived statistical measures like the maximum acceleration, the maximum gyroscopes, kurtosis, skew and after-impact with contextual and physiological data, including grip pressure, variations in heart rates, blood oxygen saturation levels, sugar level, and distance of proximity. This combination allows an event-based detection of risks and in-the-field anticipatory risk analysis in real time. Rather than selecting a single classifier, the framework introduces a heterogeneous stacked-generalization ensemble that combines Random Forest, Extremely Randomized Trees, and gradient boosting as base learners with a regularized logistic-regression meta-learner trained on out-of-fold predictions. The ensemble outputs are probability-calibrated (isotonic regression) so that a subject-specific, risk-aware adaptive decision threshold operates on trustworthy probabilities, and model behaviour is made transparent through SHapley Additive exPlanations (SHAP) for both global and local interpretability. A salient preprocessing, feature-engineering, and evaluation pipeline supports the comparison against conventional baselines including Logistic Regression, Naive Bayes, K-Nearest Neighbors, Support Vector Machine, and Decision Tree. Experimental analysis of IMU data shows that discrimination is very reliable and the proposed calibrated stacked ensemble attains a high value of 97.472% test accuracy, matching or marginally exceeding its strongest base learner, with good macro-F1 performance and good generalization. Multimodal characteristics are highly applicable in the practice on the cStick dataset since pressure, distance and HRV turned out to be the central predictors of instability and fall progression. The main innovation of the work is that it is risk-conscientious, multimodal and anticipatory constructed in such a manner that it can be applied in real-time to smart assistive devices, elderly care monitoring, home safety system, and remote healthcare applications. The developed framework provides a scalable base on the next-generation AI-empowered elderly fall prevention and intervention technologies
Abhishek Sharma, Ajay Khunteta, M. Sharma· Journal of Intelligent Decis...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
