AFFILIAZIONE
anesthesiology, critical care and pain medicine division, department of medicine and surgery, university of parma, viale gramsci 14, 43126, parma, italy.
AUTORE PRINCIPALE
Dr. Panizzi Matteo
GRUPPO DI LAVORO
Dr. Allai Simone anesthesiology, critical care and pain medicine division, department of medicine and surgery, university of parma, viale gramsci 14, 43126, parma, italy.
Dr. Bezzi Francesca anesthesiology, critical care and pain medicine division, department of medicine and surgery, university of parma, viale gramsci 14, 43126, parma, italy.
Dr. Mion Matilde anesthesiology, critical care and pain medicine division, department of medicine and surgery, university of parma, viale gramsci 14, 43126, parma, italy.
Prof.ssa Bignami Elena Giovanna anesthesiology, critical care and pain medicine division, department of medicine and surgery, university of parma, viale gramsci 14, 43126, parma, italy.
AREA TEMATICA
ABSTRACT
The Model for Assessing Artificial Intelligence (MAS-AI) was proposed in 2022 with the aim to support the traditional Health Technology Assessment (HTA) in the evaluation of AI-based clinical models, in particular the imaging ones. The legal liability of these new technologies it’s of great importance in order to safely implement them into clinical practice. This framework was developed by a multidisciplinary team of expert in order to define the minimum standards necessary to integrate an AI system into clinical workflow.
This study investigates the MAS-AI adaptability to other types of AI technologies and its trasferability in the Italian context.
Methods
We applied MAS-AI, a framework for evaluating AI’s value in healthcare, to conduct a technology assessment of an AI model developed internally named BLOC-OP. BLOC-OP utilizes AI to enhance the operational efficiency of the surgical unit and enhance the operating list scheduling. It’s trained with the real time and automated collection of patient’s surgical block, operating room and recovery room lenght of staying and by integrating these data with patient’s features and type of surgery.
We evaluated BLOC-OP’s features outlined in the project documentation using the evaluation criteria provided by MAS-AI.
Results
We successfully adhered to the methodological framework of MAS-AI and conducted a comprehensive assessment of BLOC-OP across all dimensions. Detailed descriptions of each domain within the framework, along with a summary table, were compiled.
This case study suggests the potential adaptation of MAS-AI to the Italian context and its applicability in evaluating organizational AI models like BLOC-OP.
