Medical Data Sciences & Smart Services (MDSSS®) – Digitalization and automation of analysis and therapeutic decision in ENT

Introduction:

The evaluation at hand of a prototype for integrating AI-algorithms in clinical decision-making in a ENT consultation shows the effects the software system has on parameters such as process reliability and decision-making behavior.

Methods:

We evaluated the parameters process reliability (QI) and influence on the physician (DI) based on 500 patient data sets using Medical Data Science and Smart Services (MDSSS®). Using deep learning algorithms, MDSSS® allows algorithm-based evaluation of results and treatment recommendation.

For this purpose, relevant data is captured and digitalized in currently 17 chapters of averagely 26 parameters. Data out of internal quality assurance, guidelines, drug-information and scientific publications are integrated into the decision database.

Results:

During a time period of 12 months, an average process reliability of > 70% was achieved by the control group consisting of 27 physicians. The system's influence on the physician's decisions was rated with 81% (has an influence) and 19%(changes my recommendation).

Conclusions:

The study shows the possibility of AI-application in an ENT consultation and provides information about the learning curve and automation-effects. Here, the usual cost-benefit profiles for the application of automation and AI can be reproduced.

Patrik Lindenmaier

KOPFZENTRUM Gruppe,

Münzgasse 2, 04107

Leipzig

eMail: [e-mail]

Publikationsverlauf

Publikationsdatum:
23. April 2019 (online)