Multi Parameter Adaptive Neuro Fuzzy Inference System Model For Blood Oxygen Saturation Level Maintenance Of Ventilated Patients

Document Type : Primary Research paper


1 Department Of Instrumentation, Cochin University Of Science And Technology, Kochi, Kerala, India

2 Anesthesia And Critical Care, Aster Medcity, Kochi Kerala, India


Sustaining The Blood Oxygen Level Of Patients In Mechanical Ventilators Is A
Serious Issue Faced By Physicians In Clinical Intensive Care Units. Even In Developed
Nations, Health Care Specialists Are Manually Adjusting Various Mechanical Ventilator
Parameters Periodically For Maintaining Oxygen Saturation. In Epidemic Conditions
Where The Patient Number Increases Violently, Automation Of Ventilator Reduces The
Risk And Effort Of Health Care Professionals. Real Time Maintenance Of Blood Oxygen
Saturation Level In Pandemic Situations Can Only Be Made Possible By The Application
Of Modern Machine Learning Assisted Artificial Intelligent Control Mechanisms. A
Blended Model Using Multiple Adaptive Neuro Fuzzy Inference System (ANFIS) Is Developed
Here To Predict The Inspired Oxygen Output Along With The Corresponding
Ventilator Mode For Maintaining The Blood Oxygen Saturation Level Within Desired
Limits Considering Real Time Patient Data And Ventilator Settings. Multiple ANFIS Output
Of Inspired Oxygen And Mode Prediction Is Compared With Physicians’ Decisions
And It Is Found That The Developed System Output Shows A Very Low Error Of Deviation
Of Less Than 5% From Physicians’ Prediction. Also, This Method Improves The
Speed Of Rule Development Compared With The General Fuzzy Inference Model.


Volume 12, Issue 3 - Serial Number 3
ICMMNT-2021 International Virtual Conference on Materials, Manufacturing and Nanotechnology, 30th June, 2021.
June 2021
Pages 1519-1528