Pre-courses

Pre-courses: focusing on topics such as artificial intelligence, big data & machine learning, experimental biomechanics and imaging & visualization, those interactive courses are mainly addressed to graduate students, post-docs, and young researchers

Date: Sunday 26th June, 2022
Place: Alfandega Congress Centre
Registration for pre-courses is needed.
Each pre-course will last 90-120 minutes (introduction, hands on experience)

Sunday 26th June, 13.30 – 15:30

Learning goals

Participants will:

  • Learn the basic principles of three common mechanical tests for soft biological tissue: uniaxial tensile testing, biaxial tensile testing and unconfined compression testing;
  • Get insights into the pitfalls and challenges of biological tissue testing;
  • Refresh three basic constitutive models for biological soft tissue: neohookean model, transversely isotropic ‘Gasser-Ogden-Holzapfel’ model, poro-elastic model;
  • Learn how to perform basic parameter fitting for each of these three models, using experimental data obtained from the described mechanical tests, and using a parameter fitting code in Python (which will be provided).
  • Get insights into the pitfalls and challenges of parameter fitting.

Participant requirements

Participants should:

  • Have a basic background in continuum mechanics (familiar with vectors & tensors, the concepts of stress, strain and deformation gradient tensor, Hooke’s law). A good introduction is e.g. provided in the first chapters of [1].
  • Bring their laptop (or pair up with someone who brings one), with python and a development environment for python installed. We recommend Spyder from Anaconda (version 3.7).

Nele Famaey

Nele Famaey is a professor at the Department of Mechanical Engineering, KU Leuven, heading the Soft Tissue Biomechanics research group. She obtained her PhD in Biomechanics at KU Leuven in 2012. Since 2017, she is the coordinator of FIBEr (KU Leuven core facility for biomechanical experimentation). She is currently coordinating the C4Bio initiative (c4bio.eu) as a co-chair of the VPHi Tissue Characterization task force.

Seyed Ali Elahi

Seyed Ali Elahi is a postdoctoral researcher at the Movement Sciences Department and Department of Mechanical Engineering, KU Leuven. He obtained his PhD in Mechanical Engineering at the University of Grenoble Alpes, France in 2018. In 2019, he successfully obtained a Marie Skłodowska-Curie Individual Fellowship on mechanobiological characterization and modelling of cartilage tissue.

Sunday 26th June, 16:00-18:00

Summary

The field of eXplainable Artificial Intelligence (XAI) has  rapidly  advanced  in  recent  years  as  Machine Learning and Deep Learning  models  are  becoming more and more popular in different application areas. As a result, a huge and increasing number of issues are being addressed, including the negative aspects of automated applications such as  biases and failures, which in turn have led to the development of new ethical guidelines  and  regulations such as the “Ethics Guidelines  for  Trustworthy  Artificial  Intelligence” presented in April 2019 by the  European Commission  High-Level  Expert  Group  on  AI . Consistent with the new goals,  different   XAI  algorithms  have  been  developed to  provide human-interpretable  explanations  for different predictive  models,  making  these  techniques  fully  adaptable  to  the clinical  context  and  personalized  medicine.
While these XAI approaches have proved useful in certain medical AI fields, their  application  to  the  clinical domains  is  still  in  its  infancy.  In  this  workshop, we will see the general taxonomy of the most popular XAI algorithms and we will focus on some applied clinical examples. The principal aim is to show how to realize fully interpretable AI models for supporting medical diagnoses.

Angela Lombardi

Angela Lombardi graduated with honors in Telecommunications Engineering at Technical University of Bari in 2014.
In January 2018 she got her PhD in Signal Processing and Computational Neuroscience cum laude with a thesis on multidimensional dynamic analysis of physiological brain signals at Department of Electrical and Information Engineering of Technical University of Bari.
She spent three years as a post-doc researcher in Medical Physics at National Institute of Nuclear Physics, Bari Section, involved in different national and international projects including Artificial Intelligence in Medicine (AiM).
In 2020 she was awarded the research grant  REFIN (REsearch For INnovation – POR PUGLIA FESR-FSE2014 / 2020), funded by Apulia Region to carry out the research project entitled “Biomarkers of brain connectivity from multimodal imaging for early diagnosis and personalized staging of neurodegenerative diseases with advanced methods of artificial intelligence in a distributed computing environment” (project code 928A7C98).

Currently, she is an Assistant Professor  at the  Physics Department at  University of Bari. Her research interests mainly concern nonlinear analysis of biological signals and Machine Learning techniques for classification and prediction tasks.
In particular, her work focuses on imaging processing techniques for characterizing, identifying and explaining complex patterns in medical imaging and electrophysiological recordings for different biological phenomena. Her activities mainly concern deep learning techniques for large datasets and both local and global Explainable Machine Learning Models (XAI) for interpretable and reproducible data analysis.