News

Dr Sibendu Samanta, Assistant Professor in the Department of Electronics and Communication Engineering, and Ms Radha Abburi, a PhD Scholar, have made significant strides in the field of fetal health monitoring. Their paper, titled “Adopting Artificial Intelligence Algorithms for Remote Fetal Heart Rate Monitoring and Classification using Wearable Fetal Phonocardiography,” has been published in the prestigious Q1 Journal, Applied Soft Computing, which boasts an impressive impact factor of 7.2.

This pioneering study addresses the critical gaps in the analysis of Fetal Heart Rate (FHR) recordings by leveraging wearable Phonocardiography (PCG) signals and advanced AI algorithms. The primary goal of the research is to achieve accurate classification results through the remote monitoring of fetal heartbeats. Additionally, the study tackles complex issues related to data quantity and the inherent complexity of FHR analysis. Dr Samanta and Ms Abburi’s work represents a significant advancement in the field, promising to enhance the accuracy and reliability of fetal health monitoring, ultimately contributing to better prenatal care.

Abstract of the Research:

Fetal phonocardiography (FPCG) is a non-invasive Fetal Heart Rate (FHR) monitoring technique that can detect vibrations and murmurs in heart sounds. However, acquiring fetal heart sounds from a wearable FPCG device is challenging due to noise and artefacts. This research contributes a resilient solution to overcome the conventional issues by adopting Artificial Intelligence (AI) with FPCG for automated FHR monitoring in an end-to-end manner, named (AI-FHR). Four sequential methodologies were used to ensure reliable and accurate FHR monitoring. The proposed method removes low-frequency noises and high-frequency noises by using Chebyshev II high-pass filters and Enhanced Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ECEEMDAN) in combination with Phase Shifted Maximal Overlap Discrete Wavelet Transform (PS-MODWT) filters, respectively.

The denoised signals are segmented to reduce complexity, and the segmentation is performed using multi-agent deep Q-learning (MA-DQL). The segmented signal is provided to reduce the redundancies in cardiac cycles using the Artificial Hummingbird Optimization (AHBO) algorithm. The segmented and non-redundant signals are converted into 3D spectrograms using a machine learning algorithm called variational auto-encoder-general adversarial networks (VAE-GAN). The feature extraction and classification are carried out by adopting a hybrid of the bidirectional gated recurrent unit (BiGRU) and the multi-boosted capsule network (MBCapsNet). The proposed method was implemented and simulated using MATLAB R2020a and validated by adopting effective validation metrics.

The results demonstrate that the proposed method performed better than the current method with accuracy (81.34%), sensitivity (72%), F1-score (83%), Energy (0.808 J), and complexity index (13.34). Like other optimization methods, AHO needs precise parameter adjustment in order to function well. Its performance may be greatly impacted by the selection of parameters, including population size, exploration rate, and learning rate.

The title of the Research Paper in the Citation Format:
R. Abburi, I. Hatai, R. Jaros, R. Martinek, T. A. Babu, S. A. Babu, S. Samanta, “Adopting artificial intelligence algorithms for remote fetal heart rate monitoring and classification using wearable fetal phonocardiography”, Applied Soft Computing, vol. 165, pp. 112049, 2024, ISSN 1568-4946.

Practical Implementation or the Social Implications Associated with the Research

• Chebyshev filter and EC2EMDAN-PS-MODWT reduce low and high frequency noises.
• MA-DRL and optimization algorithms reduce complexity during classification.
• Machine learning spectrogram conversion to capture time, frequency, and spectral variations.
• Hybrid deep learning algorithms can be used to reduce positive rates.

Collaborations:

• Dr. Indranil Hatai (Signal Processing and FPGA, Mathworks, Bangalore, India)
• Dr. T. Arun Babu (HoD, Dept. of Pediatrics, All India Institute of Medical Sciences (AIIMS), Andhra Pradesh, India)
• Dr. Sharmila Arun Babu, MBBS, MS (HoD, Dept. of Obstetrics and Gynecology, All India Institute of Medical Sciences (AIIMS), Andhra Pradesh, India)
• Dr. Rene Jaros (Dept. of Cybernetics and Biomedical Engineering, Faculty of Electrical Engineering and Computer Science, VSB–Technical University of Ostrava, 708 00, Ostrava, Czechia)
• Prof. Radek Martinek (Dept. of Cybernetics and Biomedical Engineering, Faculty of Electrical Engineering and Computer Science, VSB–Technical University of Ostrava, 708 00, Ostrava, Czechia)

Future Research Plans:

• Design a low cost for continuous fetal heart rate (FHR) monitoring system
• Develop a proper deep learning algorithm to get a proper understanding of fetal’s abnormality.

Link to the article

The Department of Electronics and Communication Engineering is delighted to announce that Assistant Professor Dr Anirban Ghosh, PhD scholars Mr Naga Srinivasarao and Ms Manasa Santhi, and BTech student Mr Sk Abdul Hakeem have filed and published their patent, “A System and a Method for Low Transmission Delay and Energy Efficiency,” with Application No: 202441045389. The research cohort has demonstrated groundbreaking research on integrating Small-World Characteristics (SWC) into Low-Power Wide-Area Networks (LPWANs) through Reinforcement Learning.

Abstract

To support the rapid growth of Internet of Things (IoT) applications, networking technologies like Low-Power Wide-Area Networks (LPWANs) are evolving to provide extended network lifespan and broader coverage for Internet of Things Devices (IoDs). These technologies are highly effective when devices remain stationary under static conditions. However, practical IoT applications, ranging from smart cities to mobile health monitoring systems, involve heterogeneous IoDs that move dynamically, leading to changing network topologies. Typically, dynamic networks use multi-hop data transmission schemes for communication, but this method presents challenges such as increased data latency and energy imbalances. To address these issues, this patent introduces a novel approach that integrates recent advancements in social networks, specifically Small-World Characteristics (SWC), into LPWANs using Reinforcement Learning. Specifically, the SWCs are embedded into heterogeneous LPWANs through the Q-learning technique. The performance of the developed heterogeneous Small-World LPWANs is then evaluated in terms of energy efficiency (including the number of alive and dead IoDs, as well as network residual energy) and data transmission delay within the network.

Explanation of Research in Layperson’s Terms

The existing or the present technology moves around the applications that are either static or dynamic in nature, but the current invention considers a realistic IoT application that contains both static and dynamic nodes in the network. However, maintaining low data transmission delay and high network longevity over such a heterogeneous network is a challenge. By integrating SWCs over the developed heterogeneous networks using Q-learning technique helps in minimizing the data transmission delay and improves the network lifetime (energy efficient data transmission).

Practical Implementation of the Research

Applications that contain both static and dynamic nodes, such as smart health care systems, smart environmental monitoring systems, real-time traffic monitoring systems, and smart cities and homes, require less data transmission delay and high network longevity.

Collaborations

1. Dr Om Jee Pandey – Assistant Professor, Department of Electronics Engineering, Indian Institute of Technology (BHU) Varanasi
2. Dr Satish Kumar Tiwari – Assistant Professor, IIITDM Jabalpur, India

In the next phase of research, the reserach team will work towards investigating how the energy efficiency and other quality of service of smart devices in an IoT setting can be improved if they are completely mobile.