MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202621061302 A) filed by Anish Sachin Katariya; Prajakta Pawar; Sana Munaf Bagban; Dr. Shagupta M. Mulla; Pramod Ashok Kharade; Shripurna Patil; Abhishek Padwal; Ribhav Nale; Rudra Mangate; Parth Patil; and Dr. Kuldeep Vayadande on May 14, 2026, for System And Method For Continuous Cardiovascular Monitoring Using An Iot-Enabled Multimodal Deep Learning Framework With Hybrid Cnn–transformer Architecture.
Inventors include Anish Sachin Katariya; Prajakta Pawar; Sana Munaf Bagban; Dr. Shagupta M. Mulla; Pramod Ashok Kharade; Shripurna Patil; Abhishek Padwal; Ribhav Nale; Rudra Mangate; Parth Patil; and Dr. Kuldeep Vayadande.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: Continuous cardiovascular monitoring is essential for the early detection and effective management of heart-related conditions, which often progress without noticeable symptoms. Existing monitoring systems are limited by their reliance on single-modal signals, lack of realtime processing capabilities, and dependence on cloud-based infrastructures, leading to issues such as latency, reduced reliability, and privacy concerns. The present invention proposes an IoT-enabled multimodal deep learning framework that integrates electrocardiography (ECG), photoplethysmography (PPG), and seismocardiography (SCG) signals to provide a comprehensive assessment of cardiovascular activity by capturing electrical, hemodynamic, and mechanical information simultaneously. The framework employs a hybrid CNN–Transformer architecture, where convolutional neural networks extract local temporal features and Transformer-based attention mechanisms model long-range dependencies, while an attentionbased multimodal fusion strategy dynamically enhances signal integration and robustness against noise. The system enables real-time estimation of systolic and diastolic blood pressure along with classification of cardiac conditions, and is specifically optimized for deployment on resource-constrained edge devices such as ESP32 microcontrollers using model compression and quantization techniques. By supporting on-device inference, the proposed approach reduces latency, minimizes dependency on external servers, and preserves user data privacy, while maintaining high predictive accuracy and efficiency. This makes the system highly suitable for wearable healthcare applications, continuous remote patient monitoring, and nextgeneration intelligent health monitoring solutions.
Disclaimer: Curated by HT Syndication.
