• algorithmic implementation
• artificial intelligence
• monitoring
• PDTA
• predictive scoring
• robot
• server

The object of study of this thesis is predictive monitoring by clinical severity scores through robot-patient interaction of bedridden patients in case of adverse events such as multi-organ dysfunction (MOF) including heart failure, sepsis infection and acute kidney injury by analyzing PDTA (diagnostic therapeutic care pathway). The predictive severity scores that I used and programmed algorithmically through php and excel programming are as follows:
- MEWS: Modified Early Warning Score to monitor cardiovascular problems;
- SOFA: Sequential Organ Failure Assessment for sepsis or possible septic shock;
- KDIGO: Kidney Disease for acute kidney damage.
Each Clinical Score is associated with a certain clinical protocol that allows the nursing staff to change the level of attention on the patient.
The objective of my thesis was to perform computer programming, analysis and testing on these predictive scores and on an innovative score I created for the prediction of multi-organ dysfunction so as to identify its various stages and related alert states to be proposed to the clinician. Thus, my experimental work fits within the PON project "Sì Robotics" which, through hospital server, robotic platform and Artificial Intelligence algorithms, aims to provide assistance to healthcare personnel in monitoring bedridden patients by assessing their vital, laboratory parameters and borderline states of consciousness.