Back close

Towards Precision Dosing: AI – Enhanced Model Predictive Control Closed Loop Infusion Pump for ICUs 

Thematic Area: AI-Powered Closed loop Infusion Pump

Name of the Principal Investigator : Vidya S Nair (AM.EN.R4WNA22034-FT) , PhD Scholar

Team Members: Vidya S Nair, Dr. Rahul Krishnan PathinarupothiDr.Thushara Madathil (AIMS)

Name of the Indian CollaboratorsDr.Thushara Madathil (Assistant Professor), Department of Cardiac Anesthesiology, AIMS, Kochi

Towards Precision Dosing: AI – Enhanced Model Predictive Control Closed Loop Infusion Pump for ICUs 

Intravenous infusion pumps are used in hospitals to deliver medicines such as antibiotics, insulin, hormones, and certain nutrients. Infusion pumps are designed to deliver a preset infusion to the patient. Doctors or nurses usually set the infusion rate for one hour, and a constant amount of medicine is delivered during that period. However, a patient’s response to this preset rate may vary depending upon their demographics, diseases, and surgery which is subsequently reflected in patient vitals. Hence, we propose an AI-enhanced Model Predictive Control (MPC) optimization technique customized for a personalized closed-loop drug-vital feedback system. The system is integrated into a custom-made infusion pump, whose syringe plunger movements are controlled using stepper motor algorithms that are directed by the AI-MPC output. 

Publication Details 

  • V. S. Nair, R. K. Pathinarupothi, R. P and T. Madathil, “Personalized Algorithms and Techniques for Development of a Smart Infusion Pump for ICU,” 2023 IEEE 8th International Conference for Convergence in Technology (I2CT), Lonavla, India, 2023, pp. 1-7, doi: 10.1109/I2CT57861.2023.10126150. 

Proposal 

Inspire Proposal – Towards Precision Dosing: AI – Enhanced Model Predictive Control Closed Loop Infusion Pump for ICUs 

Proposed Future Work 

  • Drug-vital feedback system using complex digital twin model 
  • Drug-drug interaction study on critical sepsis patients supported with multiple infusion pumps 

Related Projects

Computational Modelling and Prediction of Cerebellar Input Layer function, Timing and Plasticity for Understanding Neurophysiological Disorders
Computational Modelling and Prediction of Cerebellar Input Layer function, Timing and Plasticity for Understanding Neurophysiological Disorders
Study of Energy Harvesting SWIPT Enabled Wireless Networks
Study of Energy Harvesting SWIPT Enabled Wireless Networks
Development of Source Code Methodologies – Plagiarism Detection Engine
Development of Source Code Methodologies – Plagiarism Detection Engine
3D Modelling from MRI Images
3D Modelling from MRI Images
Study of Socio-Economic Aspects of Sustainability and Influence of Human Behavior in Efficient Energy Usage Using Machine Learning
Study of Socio-Economic Aspects of Sustainability and Influence of Human Behavior in Efficient Energy Usage Using Machine Learning
Admissions Apply Now