Abstract:Healthcare IoT, a fast-growing field, could revolutionize patient monitoring and intervention. This interconnection raises new security concerns, requiring real-time anomaly detection to protect patient data and device integrity. This study presents a novel security framework that uses combination of Hidden Markov Models (HMM) and Support Vector Machine (SVM) to detect anomalies in real-time healthcare IoT environments with high accuracy. The framework prioritizes real-time strength and efficiency. Sensor data from wearables, medical devices, and other IoT devices is carefully segmented into time intervals. Features are carefully derived from each segment, including statistical summaries, patterns, and frequency domain characteristics. Feature engineering is essential for accurate anomaly detection. Integrating HMM and SVM capabilities is the framework’s core. HMM accurately represent concealed states within divided data sequences and analyze patterns of change over time. After that, an independent-trained SVM examines each state. Using proximity to a decision hyperplane in feature space, this SVM can classify data points as normal or anomalous. This method augments HMM temporal capabilities with SVM classification efficiency, increasing sensitivity to anomalous patterns and reducing false positives. The framework’s exceptional performance is shown by extensive evaluations on the PhysioNet Challenge 2017 dataset, which includes diverse ECG recordings with labeled anomalies. HMM-SVM outperforms Naive Bayes, LSTM, and Random Forest with 98.66% accuracy. The framework also has high precision, recall, and F1-score, indicating a refined ability to detect real anomalies and reduce false alarms. The framework prioritizes real-time understanding and application alongside its remarkable precision. HMM-SVMs reveal hidden data state changes, revealing context and possible causes of anomalies. Modular design and efficient algorithms enable seamless integration of real-time functionality in low-resource IoT devices, enabling quick and effective security responses. To conclude, this study introduces an HMM-SVM framework for fast Healthcare IoT abnormality detection. The framework emphasizes comprehensibility and real-time applicability while achieving high accuracy. This framework can protect patient data, improve device security, and create a more reliable healthcare IoT ecosystem.

Detection of Anomaly Health Data by Specifying Latent Factors with SEM and Estimating Hidden States with HMM

Anomaly Detection Algorithm Based on Electric Equipment

Hidden Semi-Markov Model for Anomaly Detection

Hidden Semi-Markov Model-Based Methodology for Multi-Sensor Equipment Health Diagnosis and Prognosis

Research on Healthy Anomaly Detection Model Based on Deep Learning from Multiple Time-Series Physiological Signals

A Novel Method Using Adaptive Hidden Semi-Markov Model for Multi-Sensor Monitoring Equipment Health Prognosis

Bayesian Approximations to Hidden Semi-Markov Models for Telemetric Monitoring of Physical Activity

Real-time Estimation of Quality-Related Variable for Dynamic and Non-Gaussian Process Based on Semisupervised Bayesian HMM

Equipment health diagnosis and prognosis using hidden semi-Markov models

Anomaly Detection on Collective Moving Patterns: A Hidden Markov Model Based Solution.

Anomaly Detection Technique Based on Hidden Markov Model

Private anomaly detection of student health conditions based on wearable sensors in mobile cloud computing

Hierarchical Continuous Time Hidden Markov Model, with Application in Zero-Inflated Accelerometer Data

Incorporating Metabolic Information into LLMs for Anomaly Detection in Clinical Time-Series

Activity Detection from Wearable Electromyogram Sensors using Hidden Markov Model

Online Anomaly Detection for Smartphone-Based Multivariate Behavioral Time Series Data

A novel method for online health prognosis of equipment based on hidden semi-Markov model using sequential Monte Carlo methods

Realtime Anomaly Detection in Healthcare IoT: A Machine Learning-Driven Security Framework

Unsupervised Machine Learning Identifies Latent Ultradian States in Multi-Modal Wearable Sensor Signals

A Bayesian Hidden Semi-Markov Model with Covariate-Dependent State Duration Parameters for High-Frequency Data from Wearable Devices

Predicting user behavior using data profiling and hidden Markov model