2018/2019

2018/2019

DM270

ING-INF/06 (ELECTRONIC AND INFORMATION BIOENGINEERING)

DEPARTMENT OF ELECTRICAL,COMPUTER AND BIOMEDICAL ENGINEERING

BIOENGINEERING

PERCORSO COMUNE

1°

2nd semester (06/03/2019 - 14/06/2019)

6

56 lesson hours

Italian

WRITTEN AND ORAL TEST

QUAGLINI SILVANA - 6 ECTS

Basic knowledge of probability theory is required. Knowledge of at least one high-level programming languages (Matlab, R). For the practical part, good familiarity is required with the use of the PC (Windows).

The aim of the course is to provide the methodologies for modeling complex medical problems, where it is required to make decisions in the presence of uncertainty and / or taking into account patient preferences and / or multi-attribute utility functions (e.g. when we have to balance costs and benefits). Diagnostic, therapeutic and monitoring problems can be treated. At the end of the course the student must be able to formalize a decision problem, identifying the variables of the domain and choosing the most appropriate formalisms, both for the acquisition of knowledge (interaction with the medical counterpart for the construction of the model and interaction with the patient for the elicitation of preferences), both for the purpose of solving the problem. As formalisms, probabilistic networks, influence diagrams and decision trees will be considered. Among the classes of decision-making problems, particular emphasis will be given to screening programs and economic assessments prior to the decision on whether or not a health program is started. The course foresees for several sessions for the practical use of IT tools for the resolution of decision models.

Teaching program

1. Introduction: uncertainty and preferences as fundamental topics of medical decisions
2. A brief review of the basic concepts of probability theory
2.1 Some probabilities of fundamental importance in medicine
2.2 Bayes theorem and its use in diagnostics (diagnostic tests)
2.3 Probabilistic networks
2.4 Use of a software tool for probabilistic networks
3. Decision theory
3.1 Outcomes (health status, survival, costs) and methods for quantification
3.2 Quality of life and methods for the quantification of utilities (standard gamble, time-trade-off, rating scale)
3.3 Expected utility of a decision
3.4 Probabilistic dominance of a strategy compared to the other possible ones
4. Decision trees
4.1 Methods for construction and resolution
4.2 use of a software for the management of decision trees
4.3 sensitivity and threshold analysis , univariate and multivariate
4.4 representation of Markov processes within a decision tree
5. Diagrams of influence
5.1 methodologies for construction and resolution
5.2 use of a software for influence diagrams
6. Economic evaluations of health programs
6.1 cost-effectiveness analysis, cost-benefit, cost-utility
6.2 Reference thresholds for cost / effectiveness ratio
6.3 critical reading of a literature article on the subject

The course consists of lectures and exercises with tools for probabilistic networks, decision trees, influence diagrams. A literature article will also be provided on the topic of the economic evaluation of health programs, which will be critically read in the classroom

M.C. Weinstein, H.V. Fineberg L’analisi della decisione in medicina clinica, F. Angeli Editore, 2008
R. Tarricone, Valutazioni economiche e management in sanità. Applicazioni ai programmi e tecnologie sanitarie, Milano, McGraw-Hill, 2004.

The course notes are available from the Kiro platform

The exam consists of a computer test, where you will have to formulate and solve a decision tree on a cost / effectiveness or cost / utility analysis, and an oral test. During the oral examination the student is invited to choose three questions, through the extraction of three tickets from a pool of tickets containing questions about the rest of the program.

In addition to the course notes, numerous tests are available on the Kiro platform