International Journal of Innovative Research in Computer and Communication Engineering

ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines

| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |


TITLE AI-Driven Design of Advanced Smart ICU Monitoring with Early Sepsis Risk Analysis
ABSTRACT This paper represents the design and development of a prototype real-time intensive care unit (ICU) and monitoring system that combines sensor-based with vital-sign acquisition, a machine-learning-based sepsis risk assessment module, the web dashboard and locally hosted language-model assistance. The ESP32 microcontroller interact with a MAX30102 pulse-oximetry sensor, the MLX90614 infrared temperature sensor and the MPU6050 motion sensor. The firmware samples and processes sensor readings, display status and measurements on an I2C LCD, and transmits structured JSON data to a Python application over a serial connection. Then Python backend uses Flask and Flask-SocketIO to receive the stream, derive the estimated blood-pressure value using a programmed heuristic, run the Random Forest classifier and broadcast the resulting measurements and risk category to the web interface. The model-training pipeline is configured to uses the PhysioNet 2019 Sepsis Challenge patient-record files and uses heart rate, systolic blood pressure, oxygen saturation and temperature as input features. The web application provides live monitoring, trend history, patient-contextual AI chat, medication-safety assessment, AI-generated summaries and structured shift-handover reports through a locally hosted Ollama model. The prototype also supports Firebase Firestore logging and a buzzer-based alert path for critical classifications. The uploaded implementation establishes the software and hardware workflow; however, no verified model-training output, measured predictive-performance results or clinical validation data were supplied for this manuscript. The system is therefore presented as a development prototype and not as a clinically validated diagnostic or treatment device.
AUTHOR DIWAKAR REDDY, DIVYAM SINGH, HARIPRIYA MP, SHIVANI RK, DR. M SELVI UG Students, Dept. of Computer Science and Design, Rajarajeshwari College of Engineering, Bengaluru, Karnataka, India HOD, Dept. of Computer Science and Design, Rajarajeshwari College of Engineering, Bengaluru, Karnataka, India
VOLUME 189
DOI DOI: 10.15680/IJIRCCE.2026.1410044
PDF pdf/44_AI-Driven Design of Advanced Smart ICU Monitoring with Early Sepsis Risk Analysis.pdf
KEYWORDS
References [1] PhysioNet, “2019 Sepsis Prediction Challenge,” dataset referenced by the project training and download scripts. Available: https://physionet.org/content/challenge-2019/1.0.0/
[2] Project implementation files: ESP32 firmware (sepsis_monitor.ino), Python backend (app.py), model-training pipeline (train_model.py), and frontend JavaScript modules. Internal project source code, supplied for preparation of this manuscript.
[3] M. A. Reyna, C. S. Josef, R. Jeter, S. P. Shashikumar, M. B. Westover, S. Nemati, G. D. Clifford, and A. Sharma, “Early Prediction of Sepsis From Clinical Data: The PhysioNet/Computing in Cardiology Challenge 2019,” Critical Care Medicine,
[4] L. Breiman, “Random Forests,” Machine Learning,
[5] Analog Devices, “MAX30102: High-Sensitivity Pulse Oximeter and Heart-Rate Sensor for Wearable Health,” Datasheet,
[6] Melexis, “Datasheet for MLX90614: Single and Dual Zone Infrared Thermometer,” Datasheet.
[7] InvenSense Inc., “MPU-6000 and MPU-6050 Product Specification,”
[8] Espressif Systems, “ESP32 Technical Reference Manual,” Espressif Systems.
[9] Pallets Projects, “Flask Documentation.”
[10] Flask-SocketIO Contributors, “Flask-SocketIO Documentation.”
[11] Google, “Cloud Firestore Documentation
[12] Chart.js Contributors, “Chart.js Documentation.”


Copyright © IJIRCCE 2020.All right reserved