Emagister Eventos

Organizado por ETIC UPF

Biomedical Engineering: Bachelor's Thesis Presentation

  • Fechas:

    Del 09/07/18 al 11/07/18

  • Lugar:

    Sala Josep Marull, Auditori PRBB, Barcelona, España (mapa)

Web del evento

Descripción ↑ subir

During the last few months, UPF Biomedical Engineers in their last course have been developing their Bachelor Thesis. We are glad to invite you to the event that will take place in the Faculty of Health and Life Sciences (Campus del Mar) during the 9th, 10th and 11th of July, 2018. In it, each of the students will present the oral defence of their Bachelor Thesis, encompassing topics that go from purely biological to purely engineering.

These topics include minimally-invasive surgery, neuronal netwroks, heart modeling, cell membrane, biological signal processing, creation of a medical device, market analysis, machine learning, software developing, among others...

Come and enjoy this event by signing up now.

Ps: follow us on twitter to receive the latest news about the event @ThesisBE2018

Lugar ↑ subir

Programa ↑ subir

  Monday 9th Tuesday 10th Wednesday 11th
9:30 - 10:00 Video setup and welcome ceremony Video set up Video set up
TOPIC

Neuronal Network Modeling

Chair: Ernest Montbrió

Machine Learning

Chair: Antoni Ivorra

Entrepeneurship

Chair: Mamen Carmona

10:00 - 10:30 Miguel Ángel Casal Pablo Miki Martí Joana Gelabert
10:30 - 11:00 Santiago Galella Olga Galí Claudia Miralles
11:00 - 11:30 Roser Sánchez David Alonso Belén Cantos
12:30 - 12:00 BREAK
TOPIC

Biomedical Imaging

Chair: Verena Ruprecht

Prototyping or Machine Intelligence

Chair Bart Bijnens

Management

Chair: 12:00 - 12:30

Julia Pose Alejandro Antunes Esther Motjer
12:30 - 13:00 Savannah Piepadro Sergio Gómez Mar Moya
13:00 - 13:30 Sergio Moya Amaia Iribar Anna Manchón
13:30 - 14:00   Mikel Domingo Mireia Castella
14:00 - 14:30 BREAK                                                Closure
14:30 - 15:30
TOPIC

 Predictive Modelling

Chair: Óscar Cámara

 Implant characterization

Chair: Javier Macía

15:30 - 16:00 Guillem Garcia Marina Bujosa
16:00 - 16:30 Raquel Rodríguez Arnau Manasanch
16:30 - 17:00 Minerva Gándara Ainoha Aguado

 

Ponentes ↑ subir

Ainhoa Aguado
Optimization of LAAO implantation using virtual haemodynamic analysis.

 

The main objective of this work is to provide clinicians with a user-friendly interface in which they can virtually implant the desired LAA closure device. This interface includes different models and configurations of LAAO devices, and medical images of the patients under study. After the virtual implantation of the device using the interface, additional positions of the device are applied for the estimation of peri-device leak for comparison purposes using computer simulations.

 

 

Amaia Iribar
Improvement of the planning of minimally invasive interventions with a robotic arm in patients with epilepsy.

 

Optimisation of the planning and guidance of minimally invasive surgery with a robotic arm in patients with epilepsy. This project is carried out with an optical tracker that detects the exact position of the phantom and creates a virtual environment where the instructions for the intervention are given. The execution is then performed. Finally an error estimation needs to be done. 

 

 

 

 

Anna Manchón
Optimization of the Operating Room Theater Processes with the Implementation and Design of a new Software platform.

Analyzing the processes of the ORs in a given Hospital, the student will get to explain how a new Software Cloud Platform will improve the efficacy and the efficiency of the health care workflows in the Hospital.  

 

 

 

 

Miguel Ángel Casal

Consolidating memory in natural recurrent neuronal networks

Neuronal networks provide living organisms with the ability to process information. While artificial neural networks, widely used nowadays in machine learning, are mostly based on a feedforward architecture, natural neuronal networks are characterised by abundant recurrent connections. Through recurrence, neuronal signals that propagate through the network eventually come back and affect the neurones that emitted them, giving rise to strong feedback that dominates the dynamics of the system. Such feedback provides recurrent neuronal networks with fading memory, which enables them to process time-dependent information, but decays quickly. The goal of this project is to study ways in which long-term memory can be added to the recurrent computation paradigm, using the neuronal network of C. Elegans as a model system.

 

 

 

Santiago Galella

Biological validation of reservoir-based computation in a minimal nervous system

Reservoir computing has been recently proposed as a paradigm of how the brain processes time-dependent complex information. It relies on a recurrent core of neurons, known as the reservoir, that receives and encodes complex inputs in a high-dimensional phase space. Encoding takes place by combining the incoming input signal with the existing state of the network, which depends on past inputs. This provides this computational paradigm with its ability to process a temporally varying environment. While this concept was proposed a few years ago as a potential mechanism of information processing by the brain, and in spite of the overwhelming evidence of recurrent connectivity found in the brain, it has been difficult to validate this hypothesis given the extreme structural complexity of the mammalian brain. In this project we aim at validating the paradigm of reservoir computing in a much simpler model organism, namely the roundworm C. elegans, whose nervous system is fully mapped up to the level of single neurons. We will in particular examine whether the neuronal network of C. elegans can learn tasks within the framework of the reservoir computing paradigm, comparing if possible the experimentally observed electrophysiological behavior of the worm with that of our neuronal model.

 

 

 

Roser Sánchez

Algorithm to personalize brain network model based subject specific EEG data.

The objective of this project is to create an approach to personalize a virtual brain network model composed by a population of Neural Masses. The idea would be to combine (1) neurophysiological based computational models with (2) realistic structural data to generate a virtual brain. The vision of this is to generate personalize treatment with this models in the future.

 

 

 

Julia Pose

Identifying structural changes in bacterial cell membrane due to the administration of aminoglycosides using Hyperspectral Enhanced Dark Field Microscopy

The objective of this project is to create an approach to personalize a virtual brain network model composed by a population of Neural Masses. The idea would be to combine (1) neurophysiological based computational models with (2) realistic structural data to generate a virtual brain. The vision of this is to generate personalize treatment with this models in the future.

 

 

 

Savannah Piepadro

Combining T1-w and T2-w to map brain structural changes in Alzheimer’s disease

A new tool has been proposed, in order to study the brain structure, it consists on the ratio between standardized T1-weighted and T2-weighted MR images. First of all, the T1-w/T2-w ratio is compared with advanced MRI techniques (NODDI) in an animal model. Afterwards, it is computed the ratio of T1 and T2 images from ADNI database, with the aim of comparing Alzheimer’s disease and healthy control patients, performing both longitudinal and transversal studies.

 

 

 

Sergio Moya

Brain Structural and Connective differences between type 2 diabetes mellitus patients and healthy controls

The goal of this study the effect of DM2 in the brain, both structurally and functionally, mainly using voxel based morphometry and Resting-State fMRI analysis. Due to the fact that we work with a rich cohort with more than 300 clinical parameters, correlation analysis could provide interesting insights and further understanding of DM2.

 

 

 

Raquel Rodríguez

Predictive modelling of femur fracture from DXA scan using radiomics

In this study a radiomics approach is suggested as an image-based evaluation of the fracture risk at femur. The aim of the project is to create a fracture predictive algorithm using radiomics data extracted from DXA scans combined with non-skeletal data.

 

 

 

Minerva Gándara

Predictive modelling of cardiac motion from tag-MRI using radiomics

Abnormalities in heart wall motion and deformation may signify the presence of heart disease such as myocardial ischemia, consequence of coronary artery occlusion. To identify and localize these local alterations is therefore crucial for an early diagnosis and optimal assessment of CVDs. In this project we introduce the first learning approach to predict cardiac motion of the left ventricle from tagged MR images using radiomics.

 

 

 

Guillem Garcia

Heart rate variability measuring from facial videos to measure stress

This project addresses the non-contact estimate of Heart rate variability (HRV) as an indirect indicator of mental load and cognitive effort. HRV is a commonly used measure of autonomic nervous system (ANS) activity. The two branches of the ANS are the sympathetic nervous system (SNS) and parasympathetic nervous system (PNS) which dynamically control the beat-to-beat differences of the cardiac activity. While measures of sympathetic activity are pervasive, measures of parasympathetic activity have proven more elusive to obtain. However, it has been shown that while the lower frequency components of HRV are related to both sympathetic and parasympathetic activity, the higher frequency components reflect parasympathetic influence alone. Thus, spectral analysis of HRV is promising as an indirect measure of physiological aspects that are otherwise difficult to estimate, such as mental load or stress. Recently, it has been shown that it is possible to recover the blood volume pulse (BVP) from images of the human face, by amplifying the subtle color changes that occur on the skin due to the blood flow. Remote measurement of cognitive stress via HRV captured from digital cameras could allow for low-cost non-contact measurement of attention, concentration or engagement and be useful in the professional and educational settings.

 

 

 

 

Pablo Miki Martí

Characterization of HFpEF patients using unsupervised MKL

The main goal of this project is to unravel the complexity previously presented in the Heart Failure patient population. This heterogeneous group is challenging for both diagnosis and treatment, for this reason, if differences could be observed within this population, it may provide new insights on the syndrome and how to address it.

We present a framework to extract myocardial mechanic and hemodynamic features from echocardiographic images and their subsequent processing with Unsupervised Multiple Kernel Learning.

 

 

 

 

Olga Galí

Predicting adverse perinatal outcomes in a low-income location

Pakistan is one of the countries where stillbirth rate and early neonatal mortality rates are among the highest in the world. The aim of the project is to obtain and analyze fetal haemodynamic information from Doppler and combine this with maternal demographic. A computational model, combined with machine learning based on the input data, will be able to predict adverse perinatal and neonatal outcomes.

 

 

 

David Alonso

Extension to Early Prediction of Alzheimer's Disease with Non-local Patch-Based Longitudinal Descriptors.

Extension of a previous research to predict Alzheimer development using longitudinal patch descriptors on magnetic resonance images. This work extends the results of the previous by increasing the Region of Interest of the brain, increasing the number of Follow-up images used per patient and applying different machine learning algorithms to train the model.

 

 

 

Alejandro Antunes
Thesis title

 

Thesis description pending to update

 

 

 

 

Sergio Gómez

Virtual/augmented reality tool to complement navigation system in epilepsy surgeries. (With UPF and Galgo Medical)

 

Thesis description pending to update

 

 

 

 

Mikel Domingo
Thesis title

This project consist on the design and development of a non-invasive electrical stimulator of the vagus nerve of mice. This stimulator will be later assesed for its ability to modify cognitive performance on mice and as a potential therapeutic approach to improve cognitive deficits in preclinical models of intellectual disability

 

 

 

Marina Bujosa

Design and Validation of Computational Model of the voltage Distribution created by the Auditory Midbrain Implant.

A fused medical image between MRI pre-surgery and CT post-surgery scans will be used to build a CAD model of the Inferior Colliculus and the two-shank Auditory Midbrain Implant (AMI) in AutoDesk Inventor. Backward telemetry will be applied to measure the impedances of the neural tissue in an AMI patient. The model will be imported in COMSOL and validated with real measurements data in order to analyse the current flow of the implant through the tissue.

 

 

 

Arnau Manasanch

Design of a medical integrative platform for seizure-onset zone identification in epileptic patients.

A medical integrative platform is created according to the preferences of the clinicians in order to adapt it to its major commodity. Basically explained, the clinicians are able to upload the signal files and the application will perform different signal processing methods in order to obtain two-dimensional and three-dimensional information of the patient's epilepsy type. These results, in combination with other information sources, e.g. MRI, can complement the final epileptic patient diagnosis in the identification of the seizure-onset zone. Once it is identified, the patient is proposed for epileptic resective surgery to become seizure-free. 

 

 

 

Joana Gelabert

Developing a software able to program and guide how to build a biomedical device combining different biosensors

Desarrollar una plataforma intuitiva para que personas no familiarizadas con el mundo de la programación (sobre todo orientado a alumnos de eso) puedan desarrollar un dispositivo biomédico combinando diferentes bioensores

 

 

 

Claudia Miralles

CEEASY: A guiding tool throughout the CE marking process

CEEASY is a project conceived to help entrepreneurs obtain the CE mark for medical devices according to Council Directive 93/42/EEC. It is a free on-line platform that simplifies the CE marking process with all the related information, resources, examples and samples of all the necessary documentation. As a case of study, an action plan and all the documents will be written to help a UPF Research team with their new medical device: MiWEndo.

 

 

 

 

Belén Cantos

Starting a Biomedical Engineering Business from Scratch. The Case Study of OASSYS: An Oncology Assistance System Startup

Cancer is one of the leading causes of morbidity and mortality worldwide, with approximately 24 million new cases expected by 2020. Many cancer patients suffer from frequent delays and even cancellations of their appointments. In addition, most of the patients do not know exactly what to expect during the treatment. And last but not least, once the patients complete their treatments they have little or none contact at all with their Radiation Therapy Oncologist. 

Oassys is the ongoing outcome of an exhaustive and thorough need analysis of how oncologist and cancer patients can benefit from mHealthOassys is a scalable and engaging platform that supports patients during radiation therapy treatment and aftercare. With Oassys, our objective is to empower patients by actively involving them in the management of their disease and improve current feedback strategies, by automatically detecting and alerting practitioners only when one of their patients has a suspicious symptom.

 

 

 

Esther Motjer

Research and testing of a technology for a quality control assessment in EVAR interventions.

Through the present study, we pretend to address the issue of quality control of the placement of stent graft components during an Endovascular Aneurysm Repair (EVAR) intervention. Several technological approaches have been considered and evaluated bearing in mind the objective of providing an assessment of the sealing zone of the stent at the proximal aorta. Following, Diffuse Reflectance Spectroscopy (DRS) has been tested as the candidate technology in an experimental set-up for the specific application.

 

 

 

Mireia Castella
Thesis title

 

Thesis description pending to update

 

 

Eventos relacionados