Search    Print

Biomedical Signal and Data Processing
Laura Burattini

Seat Ingegneria
A.A. 2016/2017
Credits 9
Hours 72
Period I
Language ENG

Prerequisites
Basic knowledge of Matlab and signal processing.

Learning outcomes
KNOWLEDGE AND UNDERSTANDING:
Know and understand the main theoretical and practical tools for data acquisition and numerical processing of data and one-dimensional biomedical signals. The aim is to extract significant parameter for a clinical classification of the analyzed subjects. Case studies will cover the development of the electrocardiographic signal (ECG) and electromyogram (EMG).
CAPACITY TO APPLY KNOWLEDGE AND UNDERSTANDING:
This course is characterizing for the engineering sector (ING-INF / 06) and will allow to apply knowledge about the main theoretical and practical tools for biomedical data acquisition and processing to electrocardiographic (ECG) and electromyographic (EMG) signals. The students will also have to use such knowledge to develop a research project during the lab classes.
TRANSVERSAL SKILLS:
Abilities in synthesis, in group working, in orally reporting topics, in understanding clinical issues, in computer programming and in using specific instrumentation. Abilities to understand and critically analyze scientific papers an in contributing to the development of scientific projects, also thanks to basic knowledge of medicine, signal processing and advance analysis.

Program
Theory. Biomedical signals and classification according to their nature and characteristics. The four fundamental stages of biomedical signal processing (acquisition, transformation, selection of parameters and classification). Examples (electrocardiogram, electromyogram, EEG). Frequency analysis of biomedical signals: from the Fourier series to the continuous and discrete Fourier transforms. Analog/digital conversion and sampling theorem. FFT algorithm. Application examples (Tachogram). Z transform. Input-output relations: difference equations and transfer functions. Numeric filters (FIR and IIR) and their graphics solution. Random variables and their use in clinical decision. Bayes theorem. Clinical test, contingency table, correlation measures (sensitivity and specificity) and positive predictive value (prevalence). ROC curves and definition of a threshold. Coefficient of correlation, regression and least squares method. Stochastic process. Stationarity and ergodicity. Wiener-Kinchin for estimating spectral power. Tutorial. Exercises of practical application of theoretical knowledge. Development of a Matlab project.

Development of the examination
LEARNING EVALUATION METHODS
The level of student learning will be defined by evaluating: i) a written test, which is mandatory, 2.5-hour long , and consisting of 4 practical e exercises and a theoretical question. II) the developed Matlab project iii) a oral test, which is optional and consisting of 3 theoretical questions. The oral test can be accessed by the student only if he/she passed the written test (score of 18 or more). Written and oral tests must be done in the same appeal.

LEARNING EVALUATION CRITERIA
In order to pass the written exam, students must demonstrate that they have acquired a knowledge of the subject theoretical principles such that they learned how to use them in solving practical problems. The written test will be considered to be positive if and only if the exercise on the filter is at least partially solved (30%). Overall, the written test will be graded from 0 to 30. The Matlab project will be graded from 0 to 30. The oral test, if sustained, will be graded from 0 to 30.

LEARNING MEASUREMENT CRITERIA
Attribution of the final mark out of thirty

FINAL MARK ALLOCATION CRITERIA
The final mark (V) will be computed using the mark pf the written test (V1), the mark of the Matlab project (V2) and the mark of the oral test (V3), if sustained, as follows: -if no oral test sustained: V=0.7 V1+ 0.3 V2. -if oral test sustained : V=0.7 ((V1+V2)/2)+0.3 V2. The honors will be given to students who, having achieved the highest rating, have demonstrated complete mastery of the subject.

Recommended reading
1)Jackson LB. “Digital filters and signal processing“, Kluwer Academic Publishers, Boston, 1993. 2) Oppenheim A, Schafer R. “Discrete-Time signal processing“, Prentice Hall, Englewood Cliffs, NJ, 1989. 3) Peebles PZ. “Probability, random variables, and random signals principles“, McGraw-Hill Inc., Boston, 2001. 4) Akay M. “Biomedical signal processing“,Academic Press, San Diego, 1994. 5) Dispense.

Courses

• Biomedical Engineering (Corso di Laurea Magistrale (DM 270/04))