Biomedical Engineering

FEBO CINCOTTI
FEBO CINCOTTI
EMMA COLAMARINO
EMMA COLAMARINO

General objectives
The course aims to introduce the principles, methodologies, and applications of the main engineering techniques used to study biomedical data pertaining the domains of statistics and machine learning.
Specific objectives
- Knowledge and understanding
Students will learn concepts of descriptive statistics, hypothesis testing, classification models and advanced biosignal processing
- Applying knowledge and understanding
Students will familiarize with basic tools to apply statistical tests and to train basic classification models
- Critical and judgment skills
Students will learn to choose the most suitable control methodology for a specific problem and to evaluate the complexity of the proposed solution.
- Communication skills
Students will learn to communicate in a multidisciplinary context the main issues of interfacing neurophysiological signals with artificial systems, and to convey possible design choices for this purpose.
- Learning ability:
Students will develop a mindset oriented to independent learning of advanced concepts not covered in the course.

1. Knowledge and understanding (DD1). At the end of the course, the student will be able to:
○ Know and understand the principles of descriptive and inferential statistics, including estimation procedures and hypothesis testing (e.g., parametric tests, ANOVA, chi2).
○ Understand the fundamental machine learning methodologies for classification (parametric approaches, data-driven approaches, artificial neural networks) and the relative metrics and procedures for performance evaluation.
○ Describe advanced biosignal processing techniques, with specific reference to the analysis of the electromyographic signal (e.g., MUAPs, muscle synergies) and the estimation of neuroelectric sources (e.g., forward/inverse problem, spatial filters, ERP, ERD/S).
2. Applying knowledge and understanding (DD2). The student will be able to:
○ Apply the correct statistical procedures to the analysis of real biomedical data, implementing dedicated scripts in the Matlab environment.
○ Apply and validate basic classification models for the analysis of biomedical data, in order to discriminate between different classes or conditions of interest.
○ Perform the analysis of electromyographic and neuroelectric signals using the learned techniques, in order to extract parameters of functional interest.
○ (These skills are verified mainly through the lab exam).
3. Making judgements (DD3). The student will be able to:
○ Critically evaluate which statistical test or classification model is most appropriate for a given biomedical dataset and analysis objective.
○ Interpret the results of an analysis (statistical or classification), evaluating the significance of the results and the limitations of the applied methodologies (e.g., risk of overfitting).
○ Choose the most suitable processing methodology for a specific biosignal based on the application.
○ (This skill is stimulated by the lab sessions and evaluated through the open-ended question and the practical test).
4. Communication skills (DD4). The student will acquire the ability to:
○ Describe the learned methodologies and the results obtained with clarity and technical rigor.
○ Argue in a multidisciplinary context (e.g., with engineers, doctors, biologists) the analytical questions that can be addressed using the acquired skills, critically evaluating their application limits.
○ Justify the design and methodological choices adopted (evaluated through the open-ended question and the lab test).
5. Learning skills (DD5)
○ The student will have developed the methodological skills and mindset necessary to autonomously deepen their understanding of advanced topics not explicitly covered in class (e.g., specialized statistical methods, machine learning algorithms, or emerging biosignal analysis techniques), by consulting scientific literature in the field.

Knowledge of the following is required:
● probability theory: single- and multi-dimensional random variables, probability functions and distributions, expected values, estimators;
● biosignal processing: physiology, instrumentation, and processing of electroencephalographic and electromyographic signals;
● programming notions and basic knowledge of the Matlab environment.

Section I: Statistics.
Introduction. Descriptive statistics. Probability distributions. Sampling distributions. Interval estimates. Inferential statistics. One-sample tests for means.
Independent-samples tests for means. Tests for more than two samples. Independent-samples analysis of variance (ANOVA).
Repeated-measures tests for means. Power of a test. Tests based on the Chi-square distribution.
Notes on non-parametric tests.
Section II: Introduction to classification.
Introduction. Metrics and performance evaluation procedures for a classifier. Parametric approaches. Data-driven approaches.
Non-recursive multilayer artificial neural networks. Overfitting.
Section III: Biosignal processing
Electromyographic signal processing. Review. Single-channel processing procedures.
Multi-channel acquisition from different muscle districts (muscle synergy, motor control), estimation of MUAPs and their functional interpretation, applications).
Neuroelectric source imaging. Review. Models for ERP estimation (classic and rephasing); Estimation of induced activity, ERD/S.
Multi-electrode recordings and spatial filters. Neuroelectric forward problem. Neuroelectric inverse problem.
Section IV: Computer lab sessions
Implementation of programs in the Matlab environment, applying the concepts learned in Sections I-III.

The teaching material provided by the instructors includes:
● lecture slides
● lecture recordings
● past exam papers and solutions
● additional self-administered exercises to gain more familiarity with the Matlab language.
The material is shared on the cloud with all students of the course, and access procedures are described in the Piazza class (https://piazza.com/uniroma1.it/spring2026/1044421).

Lane, DM. Online Statistics Education: A Multimedia Course of Study, https://onlinestatbook.com
Surface Electromyography : Physiology, Engineering, and Applications; Editor(s):Roberto Merletti, Dario Farina, 2016, The Institute of Electrical and Electronics Engineers, Inc.

Lectures (4 lessons/week) where program topics are explained and a selection of problems is solved.
Computer lab sessions (1/week) where students practice applying the learned concepts in the Matlab environment.

Attendance is not mandatory. However, participation in laboratory sessions is recommended.

The assessment of preparation will be carried out through written and computer lab tests.
All tests are held in the same session, separated by a short break.
The written test will consist of:
(i) multiple-choice questions in which knowledge and understanding of the topics covered during the course are evaluated
(ii) an open-ended question, in which the depth of study of the subject and communication skills are evaluated.
In the lab test, problems will be proposed to be solved by writing scripts in the Matlab environment, in order to evaluate the ability to apply knowledge and understanding.

All past exam papers and solutions are available in the teaching materials accessible to students attending the course.
As an example, an exam paper can be viewed at the following link:
https://drive.google.com/file/d/11N-FgN8omYS7sVfIk0EsYtITS1uiQ7Zh/view?usp=drive_link

• Introduction to statistics and descriptive statistics 


• Probability distributions and sampling distributions


• Interval estimates and inferential statistics


• Independent-samples tests for means 


• Repeated-measures tests for means


• Power of a test, Chi-square based tests, notes on non-parametric tests


• Introduction to classification, performance


• Linear parametric and data-driven classifiers


• Artificial neural networks, overfitting 


• Single-channel EEG


• Multi-channel EEG


• Single-channel EMG


• Multi-channel EMG