Skip links

Predictive models from image and video #computerVisionfrom signal and time series #timeSeriesForecastingfrom text #naturalLanguageProcessingbased on environment-agent interaction #reinforcementLearning

GIGAVISIÓN

GIGAVISIÓN

System for Marking Tumor Regions in Gigapixel Histological Images

The main objective is the appraisal of the results of the investigation of new artificial intelligence algorithms for the automatic analysis of histological images (WSI) applied to the diagnosis of different types of cancers, among them (although not limited to): prostate cancer, triple negative breast cancer (TNBC), and skin cancer.

The latest technological advances have led to a drastic change in the possibilities of health care, thus improving the conditions of medical care. But today’s pathology services still rely heavily on the presence of qualified pathologists to recognize characteristic findings in a tissue section under a microscope.

Digital pathology, and innovation in this area, solves multiple problems related to both the development of work, the quality of service, as well as the patient (diagnosis and safety).

Therefore, the main objective of this project is to create a web platform for visualization, annotation, and automatic evaluation of histological cases that supports the identification of different types of cancer. This tool will allow pathologists from all over the world to obtain online diagnostic help based on artificial intelligence techniques.

As a main novelty, said system will host predictive models generated from the most innovative techniques in the field of deep learning. Using new digitized histological samples from any hospital in the world, the predictive models hosted in the cloud will be retrained with these cases using innovative active learning techniques.

Agency

FEDER Fondo Europeo de Desarrollo Regional

Years

2021 – 2023

Share

Partners

PICASSO

PICASSO

A novel active learning-based digital pathology protocol annotation for histologic assessment in ulcerative colitis using PICASSO histologic remision index

Ulcerative colitis (UC) is an inflammatory bowel disease affecting the large intestine characterised by a relapse-remission cycle. Traditionally, the management of patients with ulcerative colitis has been based on the analysis of remission of inflammatory activity at the endoscopic level by evaluating videos recorded during endoscopic examinations. Recently, the treatment paradigm has shifted towards the study of biopsies for the analysis of histological remission at the microscopic level of tissues. One of the histological indicators of UC is the PICaSSO Histological Remission (PHRI) which consists of the assessment of inflammatory activity taking into account only neutrophils, key cells in inflammatory processes. PHRI is a simple assessment system which reduces the variability between subjects and facilitates the implementation of diagnostic support systems. The CVBLab group contributed to the validation of the application of the PHRI index in automated systems for the prediction of histological remission or the presence of ulcerative colitis-related activity. Close collaboration between CVBLab engineers and the group’s medical staff has enabled the translation of the problems they encounter in the clinic into solutions in the technical domain that contribute to the development of robust and effective deep learning models.

Texto proyecto

Años

2021 a 2025

Share

Socios

DISRUPT

DISRUPT

ON-CHIP TOMOGRAPHIC MICROSCOPY: A paradigm shift for revolutionizing LAB-ON-A-CHIP bioimaging technology

DISRUPT aims to revolutionise biomedical imaging by developing a pioneering technology: integrated tomographic microscopy. This innovative technique combines on-chip tomography and wireless photonics, together with microfluidics and artificial intelligence (AI).

The project aims to leverage CMOS compatibility to create more affordable and compact tomographic microscopes. It aims to demonstrate its potential in applications such as cancer detection and the identification of infected cells, leveraging key advantages in resolution, sensitivity, performance and energy efficiency.

The foundation of DISRUPT is based on CMOS compatibility, a paradigm shift that will enable the development of tomographic microscopes that are more accessible, lighter and more compact compared to current solutions. This innovative approach not only represents a technological improvement, but also has the potential to transform the accessibility and economic viability of advanced biomedical imaging.

DISRUPT’s proposed device not only aspires to be a technological innovation, but is also projected as a tangible solution for various biomedical applications. From early cancer diagnosis to cell characterisation, cancer and infectious disease research, immunocyte phenotyping, stem cell multipotency identification, tissue pathology, haematopathology and infected cell analysis, the range of possibilities is wide and promising.

The intrinsic characteristics of this technology, such as its mass production capability, compactness, low cost, mechanical robustness and energy efficiency, position it as a driver of innovation for future developments in multiple biomedical application fields. Furthermore, it is envisioned as a valuable alternative tool that addresses emerging needs for microscopic analysis and diagnosis in low-resource settings, telemedicine and point-of-care applications. This approach, with potentially enormous societal impact, has the potential to promote early diagnosis of cancer and other diseases and infections, transforming healthcare delivery and democratising access to advanced technologies in the biomedical field.

Funding Entity

Comisión de las Comunidades Europeas

Years

2023-2025

Share

Partnerts

DICOMO

DICOMO

DICOMO

In this project we intend to develop computer vision algorithms for the evaluation of the progression of the treatment of Idiopathic Scoliosis in its different stages.

Idiopathic Scoliosis is a pathology of undefined cause that produces an abnormal development in the curvature of the spine. This pathology mainly affects young adolescents, with a prevalence of up to 5.2% of the population. Its main treatment is the use of a corrective corset. If this does not work, the patient is treated by surgery to fix the curvature of the spine.

Currently, the follow-up of corset and surgical treatment is done by torso X-rays and manual analysis of spinal markers by expert surgeons. This includes known inter-expert variability, and the markers used are not yet sufficiently robust to assess disease progression. Furthermore, the need for successive X-rays carries risks of excessive radiation in young patients. For this reason, DICOMO was born as a project whose main objective is to develop tools based on artificial intelligence capable of supporting expert surgeons in this task. Specifically, it seeks to reduce variability and workload by obtaining markers such as Cobb angles automatically, and tries to find novel early markers of success in treatment by corset and surgery.

 

Agency

PIDI-CV — I+D PIME-IVACE ( IMIDTA/2020/54)

Years

2021

Share

Partners

MINERVA

MINERVA

Mid- to Near Infrared Spectroscopy for Improved Medical Diagnostics

Taking advantage of several new breakthroughs in photonic technology to develop a new mid-IR technology platform and processes for early detection of cancer.

The final diagnosis of most types of cancers is performed by an expert clinician in anatomical pathology who examines suspicious tissue or cell samples extracted from the patient with a traditional optical microscope. Currently, this tedious and subjective assessment largely relies on the experience of the clinician and gives rise to suboptimal levels of discordance between different pathologists, especially in early stages of cancer development.

Recently, infrared (IR) spectroscopy has shown great potential to open a new chapter in the field of biomedical imaging. Mid-IR light is able to excite the vibrational modes of the most relevant biomolecules and the acquired signals inform about the chemical composition of the sample. Therefore, this emerging imaging technique has a high potential to perform objective pathological diagnoses and improve the detection and evaluation of the patient’s risk in screening and cancer surveillance.

To date, the lack of suitable sources, detectors and components in the mid-IR have restricted this technology to one of academic interest. The MINERVA project pursues several targets in parallel, from developing new IR instrumentation, such as fibre lasers, acousto-optic modulators, supercontinuum sources and detectors in the mid-IR range, to explore the performance and limitations of current IR spectroscopic technology in early cancer identification.

CVBLab group collaborates with other partners to explore and implement new mathematical and computational techniques from the interrelated fields of Digital Image Processing, Computer Vision, Machine Learning, Pattern Recognition, Multivariate Analysis and Chemometrics for the acquisition, processing and analysis of the bio-molecular spectral signatures of biological tissue and cells enclosed within IR hyperspectral images.

Agency

Comisión Europea bajo el Séptimo Programa Marco (FP7-317803)

Years

2012 to 2017

Share

Publications

Partners

SICAP

SICAP

Sistema de Interpretación de Imágenes Histopatológicas para la Detección del Cáncer de Prostata.

To develop a diagnostic aid system for prostate cancer by classifying the histopathological images from biopsies in different grades according to the Gleason scale.

Today, prostate cancer is one of the most common types of cancer in humans, along with lung cancer and breast cancer. To diagnose it, a physical examination and a PSA analysis are carried out. If there are indications that the patient may have cancer, a biopsy is performed to obtain prostate tissue samples. Afterwards, an expert doctor in pathological anatomy examines these samples and assigns them a score according to the Gleason classification system, which establishes that grades 1 and 2 correspond to a benign prostate tissue, while grades 3, 4 and 5 correspond to a malignant one.

Nowadays, the analysis to classify the samples is a very tedious and time-consuming manual task. In addition, it usually involves a considerable level of subjectivity between different specialists. For this reason, SICAP is born as a project whose main aim is to perform a diagnostic aid system that allows the automatic classification of biopsied samples, according to the Gleason scale. In this way, it would be possible to help the pathologists to improve in terms of time and effectiveness as well as to reduce the level of discordance that exists between them when they try to classify a certain sample.

Biomedical and telecommunications engineers collaborate in the CVBLab group working on the implementation of computational techniques based on Machine Learning and Deep Learning applied to biomedical images, in order to find characteristics and patterns that allow to determine automatically not only if the patient has cancer, but also the severity of it.

Agency

Ministerio de Economía, Industria y Competitividad (DPI2016-77869-C2-1-R)

Years

2017 to 2020

Share

Socios

WIBEC

WIBEC

Wireless In-Body Environment Communications

WiBEC’s main objective is to provide high quality and innovative doctoral training in development of the technology for novel implantable devices that can contribute to improvement in quality, access, and efficacy of healthcare.

WIBEC (Wireless In-Body Environment Communications) is a Horizon 2020 Innovative Training Network (ITN) in which young researchers are recruited and trained in coordination by a group of Academia, Industry and Medical Centres. The training is aimed at the Social, Health and Technologies Challenges of the H2020: Wireless In-Body Devices.

Two devices are the focus of the individual researchers’ projects: cardiovascular implants and ingestible capsules to investigate gastro intestinal problems. These devices enable medical professionals to have timely clinical information at the point of care. The medical motivation is to increase survival rates and improvement of health outcomes with easy and fast diagnosis and treatment.The goal for homecare services is to improve quality of life and independence for patients by enabling ambient assisted living (AAL) at home.

Inter-sectorial and multi-discipline work is essential in this topic, as it requires cooperation between the medical field and engineering field, and between institutions and industry.WiBEC intends to offer the recruited early-stage researchers (ESRs) a joint programme in training and research,in addition to their individual research projects, with the aim of keeping the group of ESRs coordinated and promoting the future cooperation between them and between the participating institutions in what can be called a Network of Knowledge.

Innovative technology is rapidly becoming more important in surgery and medical assistance, and a large number of experts combining engineering and medical skills will be required in Europe to enable novel paradigms like AAL to be realised. The ESRs who are a part of this ITN will acquire the required diverse skills that will enable them to occupy privileged positions to join and promote EU leadership in ICT for Health.

https://www.ntnu.edu/wibec

Agency

European Commission, MARIE SKŁODOWSKA-CURIE ACTIONS, (H2020:MSCA:ITN, 675353)

Years

2016 to 2019

Share

Publications

Partners

SAMUEL

SAMUEL

Artificial Intelligence System for Molecular and Morphological Skin Cancer Characterization

To create a diagnostic aid system based on image analysis, epigenetic information and clinical data to detect skin cancer. In particular, the aim is to develop machine learning algorithms to differentiate between melanoma and nevus and assess the prognosis of cases of uncertain malignant potential.

The incidence of skin cancer in the Western world has followed an increasing trend in recent decades, with the European continent being the most affected by melanoma. The high incidence of skin cancer implies an increased demand for skin biopsies, which supposes a logistical challenge for pathology departments. As an added motivation, many of the lesions referred to pathology departments report very poor guidelines to substantiate the malignancy of the lesion. It leads to a significant workload for the experts, who have to spend much of their time manually analysing cases that ultimately turn out to be benign (approximately 80% of cases).

Among the different types of skin cancer, malignant melanoma is the most aggressive and dangerous, accounting for around 80% of deaths associated with skin cancer. In this respect, early detection and diagnosis of the disease at an early stage is essential in order to reduce, as far as possible, the associated complications. However, the characterisation and differentiation of these tumours from other benign melanocytic tumours or tumours of uncertain malignant potential are not straightforward, even for experienced pathologists. Moreover, there are different subtypes of malign melanoma with similar morphologies that pose a challenge for experts. For this reason, the SAMUEL project aims to develop diagnostic aids that provide pathologists with an automatic classification into different melanoma subtypes. To this end, SAMUEL proposes applying artificial intelligence algorithms to Whole-Slide Images (WSIs), which are high-resolution digitised biopsy samples. In addition to the histological images and the patient’s clinical data, molecular mechanisms, that allow for the dissemination of cutaneous melanoma, will be considered. Specifically, procedures such as DNA methylation and microRNAs will be carried out, as they play a determining role in the progression and development of melanocytic tumours.

In conclusion, SAMUEL project aims to develop a machine learning system to aid skin cancer diagnosis to differentiate between malignant and benign melanocytic tumours and evaluate the prognosis of cases of uncertain malignant potential. To this end, the project will consider the combination of clinical data with histopathological and epigenetic information.

Agency

Agencia Valenciana de la innovación

Years

2021-2023

Share

Partners

BRAIM

BRAIM

Advanced Techniques of Brain Medical Imaging

The objective of this project is the development and validation of tools for the early diagnosis of neurodegenerative diseases, such as Parkinson’s, Multiple Sclerosis or Alzheimer’s, as well as for the diagnosis and treatment of brain neoplasms.

BRAIM project develops and experimentally validates a set of techniques for processing different types of medical images that allow clinical professionals to make the best cooperative decisions for the diagnosis and treatment of brain diseases.

A set of tools is developed and validated. These tools, based on 3D resonance images of a patient, make possible to quickly and accurately calculate brain volumes and classify the patient according to a control database for the diagnosis of neurodegenerative diseases. In addition, this software allows, through the patient’s classification, to make an evolutionary prognosis of the patient under study. Although these tools can be useful in the diagnosis of multiple pathologies, the validation of the system in this project focuses on Parkinson’s and Multiple Sclerosis patients.

As for neuro-oncology, powerful tools are developed that allow clinical staff to determine, from anatomical images of 3D resonance, the total volume of a tumour lesion for the use of this variable in the next evolutionary controls. In addition, the information provided by these anatomical images can be merged with metabolic and functional images to, in a non-invasive way, provide greater knowledge about aspects related to the lesion (location, aggressiveness, extension, infiltrative pattern …). With all this, it is intended to help not only in the diagnosis of neoplasms and their follow-up, but also in making various crucial decisions such as therapeutic strategy.

Agency

Centro para el Desarrollo Tecnológico Industrial (IDI- 20130020)

Years

2012 to 2015

Share

Publications

Partners

HEPAPLAN

HEPAPLAN

Hepatic Planning: Tumor’s Monitoring and Analysis for Hepatic Surgery

To develop a software tool for prospective and retrospective analysis of liver lesions evidenced in anatomical studies that helps physicians in decision-making.

ONCOTIC project consists of 6 sub-projects where the common denominator is the use of ICT technologies to improve diagnosis, treatment or monitoring of oncological pathologies. HepaPlan is one of those sub-projects in which, specifically, it seeks to improve the diagnosis and monitoring of liver cancer.

Nowadays, doctors only have some 2D images to issue patient’s diagnosis. However, the goal of this project is to provide them with additional information that help them to decide on the most appropriate therapy or treatment. The software resulting from the project reconstructs a 3D model of the patient’s liver from studies of Magnetic Resonance (MR) and/or Computed Tomography (CT) along with its internal anatomical structures (veins, arteries and lesions, if any). Once the images are registered, volumetric measurements can be made of both the liver and the size of the tumor and their relative position, as well as the percentage of the affected liver, the ratio of healthy tissue/tissue with lesions and the distance between the tumor and the veins or nearest arteries. In addition, if previous studies of the patient are available, it is possible to monitor the lesion evolution from the beginning of treatment to the present. In order to plan a surgical intervention, the software also allows the specialists to simulate a possible resection and to evaluate in a virtual way the surgery effects on liver functional capacity.

HepaPlan was awarded the ‘Best Innovation in Technology 2013’ within the MIHealth Innovation Awards. These awards recognize the innovative and transformative spirit of health professionals and highlight the progress made in having a liver planner to address surgical interventions with a higher degree of reliability, which reduces risks for the patient.

Agency

Centro para el Desarrollo Tecnológico Industrial (CDTI) (IDI-20101153)

Years

2010 to 2013

Share

Publications

Partners

ACRIMA

ACRIMA

Fundus Image Processing for Automatic Screening of Ophthalmological Diseases

Development of an automatic screening system able of detecting the three most significant diseases related to visual impairment in the current society: glaucoma, diabetic retinopathy and age-related macular degeneration.

In this project, we propose new image processing algorithms not only focused on the detection of pathological patterns but also on the definition and detection of the “normal” retina. In addition, some advances will be carried out incorporating into automatic methods new markers that are currently used in clinical diagnosis but not in automatic detection. Moreover, these advances will be combined to define a new detection model that incorporates other clinical data to image features. For that reason, the final system resulting from the project will be more robust and with enough sensitivity and specificity facilitating its application in clinical practice.

The main motivation of this project is the high social and economic impact of the blindness in the current society in addition to the importance of an early diagnosis of the main diseases that cause blindness. Appropriate referral for treatment of these diseases is key because the severe visual impairment can be prevented in 90% cases.

Nowadays, a direct, regular, and complete ophthalmologic examination seems to be the best approach for risk population assessment. However, population growth, ageing, physical inactivity, and rising levels of obesity are contributing factors in increasing systemic diseases as the diabetes, which may be associated to ophthalmologic diseases, so that the number of ophthalmologists required for a direct examination of the risk population is high.

Our main hypothesis is that it is possible to develop an automatic system for screening of the risk population in an accurate, sensitive and specific way by making use of the findings in the retinal images along with other clinical data of the patient.

Entidad financiadora

Ministerio de Economía y Competitividad (TIN2013-46751-R)

Años

2014 a 2016

Share

Publications

Partners

AIDE

AIDE

Adaptive Multimodal Interfaces to Assist Disabled People in Daily Activities

The objective of this project is the development of a modular and adaptive multimodal interface customizable to the individual needs of people with disabilities by human-machine interface devices.

AIDE project has the ambition to strongly contributing to the improvement of the user-technology interface by developing and testing a revolutionary modular and adaptive multimodal interface customizable to the individual needs of people with disabilities. Furthermore, it focuses on the development of a totally new shared-control paradigm for assistive devices that integrates information from identification of residual abilities, behaviours, emotional state and intentions of the user on one hand and analysis of the environment and context factors on the other hand.

The AIDE project addresses different challenges. First, the development of a novel multimodal interface to detect behaviours and intentions of the user. Secondly, the implementation of a shared human-machine control system. Thirdly, the development of a modular multimodal perception system to provide information and support to the multimodal interface and the human-machine cooperative control. The system is composed of: brain machine interface (BMI) control based on EEG brain activity, wireless EMG surface interface, wearable physiological sensors to monitor physiological signals (such as, heart rate, skin conductance level, temperature, respiration rate, etc.), wearable ElectroOculographic (EOG) system, eye tracking module  to identify the gaze point, an RGB and depth camera (such as Kinect) to recognize and track the user  and objects and kinetic and dynamic information provided by upper-limb exoskeleton.

There exist different scenarios as possible targets for the AIDE system: drinking tasks, eating tasks, pressing a sensitive dual switch, making personal hygiene, touching another person, communication, control of home devices, entertainment and so on.

http://www.aideproject.eu/en/

Agency

European Union’s Horizon 2020 Research (H2020-645322)

Years

2015 to 2018

Share

Partners

IMFUTEC

IMFUTEC

Advanced Techniques in Medical Image Fusion

To develop tools for analysis of molecular images obtained through multi-voxel magnetic resonance spectroscopy that allow an exhaustive examination of cerebral metabolism.

ONCOTIC project is composed of 6 sub-projects where the common denominator is the use of ICT technologies to improve diagnosis, treatment or monitoring of oncological pathologies. IMFUTEC is one of those sub-projects in which, specifically, it seeks to improve the diagnosis and monitoring of brain cancer through a software tool that allows to obtain and analyze the molecular substances that comprise it. Its objective is to know in an objective and premature way the state of the brain tissue in the presence of a possible neurological disease.

The current treatment of patients with cancer implies a better knowledge of aspects such as the molecular alterations responsible for tumor phenotype appearance, the ability of early diagnoses, the monitoring of tumor mass evolution in response to treatment and the adequacy of therapeutic strategies to molecular alterations. In this context, IMFUTEC approaches the development of a software tool for the analysis of molecular images obtained through multi-voxel magnetic resonance spectroscopy. Thanks to the multi-voxel technique, clinicians can study not only the temporal response of brain metabolites, but also its spatial distribution, in order to know in a precise way the area where the tumor is located.

Making use of this tool, several signals obtained by proton spectroscopy can be analyzed and integrated in an interactive way to subsequently generate color maps belonging to the substances of the different anatomical areas under study. The results obtained from the spectroscopy can be fused with high resolution 3D anatomical images and, depending on the study in question, with Functional Magnetic Resonance (FMR) and/or PET images. This is an important milestone in the diagnosis and early detection of diseases such as Alzheimer’s or multiple sclerosis.

 

Agency

Centro para el Desarrollo Tecnológico Industrial (CDTI) (IDI-20101153)

Years

2010 to 2013

Share

Publications

Partners

MIRACLE

MIRACLE

Intelligent CAD System for Dental Prostheses

This project aims to develop and validate an intelligent CAD system for the design, simulation and flexible fabrication of implant-supported dental prostheses.

MIRACLE is a project that aims to develop an advanced system for the design, simulation and flexible manufacture of implant-supported dental prostheses. Dental implants are screwed directly into the mandible or jaw, and it is to these that dental prostheses are anchored. The design and manufacture of dental prostheses is a very skilled process that involves high time costs and a method that lacks, in many cases, functional design specifications.
The aim is the evaluate the functional characteristics of a prosthesis and its biomechanical features so as to develop a just-in-time system that will enable the placement of the implants and prosthesis in a single intervention, such cases are known as immediate loading cases, and that avoids a healing period and removes the need for a second procedure.
The developed system in MIRACLE is a CAD/CAM system which allows to test the functional characteristics of dental prostheses considering mandible-maxilla interaction using virtual models, contrary to most commercial solutions where this test is performed using expensive anatomical replicas tested with mechanical articulators and evaluated with patients. Another objective of MIRACLE is to develop a parametric finite elements model (FEM) of the whole prosthesis in order to analyze the failure risk of dental implants and prostheses before its surgical implantation enabling a re-design process.

Agency

Ministerio de Educación y Ciencia (DPI2007-66782-C03-01-AR07)

Years

2007 to 2010

Share

Publications

Partners

NARALAP

NARALAP

Augmented Reality Navigation for Laparoscopic Surgery

Implementation of an augmented reality system for assistance in abdominal surgery and development of a liver biomechanical model that allows intra-operative navigation during laparoscopic surgery.

NaRALAp seeks to develop a navigation system with laparoscopy based on two pillars: virtual/augmented reality and a liver biomechanical model. Its objective is to advance in the creation of technology that facilitates safer laparoscopic interventions and in accordance with preoperative planning.

In minimally invasive abdominal surgeries, laparoscopy, it is of great importance the precise location of the incisions where the instruments will be introduced during the intervention (trocars) so that the surgery is effective. Sometimes, when the operation is complicated, determining the exact position where to make the incisions is a difficult task even for experienced surgeons. To determine this location, the surgeon is guided by palpation and 2D visualization of the anatomical and pathological structures of the patient obtained from preoperative CT images or Magnetic Resonance Imaging (MRI).

In this project, augmented reality (AR) is used to facilitate trocar placement. For this purpose, a 3D model of the abdominal organs of the patient extracted from preoperative anatomical studies is created. Subsequently, this model is projected onto the patient at the operating table, providing additional information to the surgeon to decide the most appropriate place to perform the incisions.

The system resulting from the project was validated in a real environment obtaining promising results and demonstrating that an augmented reality system can provide improvements to the surgeon in trocar placement for laparoscopic surgery.

Agency

Spanish Ministry of Industry, Tourism and Commerce through the AVANZA I+D sub-program (TSI-020100-2009-189)

Years

2009 to 2011

Share

Publications

Partners

EMBRIODEEP

EMBRIODEEP

Development of algorithms based on Deep Learning for the improvement of embryo selection using artificial vision technologies and its application in a prototype demonstrator for its exploitation in the healthcare field.

To develop a system of assistance for embryo selection using Deep Learning algorithms and the most robust artificial vision technology applied in a non-invasive clinical protocol for embryo selection, which automatically analyses: morphokinetic variables, variables related to genomics and proteomics, variables related to oxidative stress, other variables intrinsic to the assisted reproduction cycles, extrinsic variables resulting from the working environment and the historical health records of all the births cultivated in Time Lapse. 

Taking into account all the data from these variables, a predictive model of successful implantation will be obtained and patterns will be investigated to enable the development of an alert system to detect and simulate in silico situations in which an embryo is more likely to achieve successful implantation. These alerts will form part of the care support system for the selection of the best embryo and will facilitate decision-making in an assisted reproduction centre. 

This project arises from the concern of the IVI Group’s parent company for the need to implement personalised clinical protocols that help to increase the success of each assisted reproduction cycle and also produce a reduction in the number of attempts made until reproductive success is achieved. The aim is to incorporate digitalisation technologies” through the development of services and models of advanced analytics and data visualisation that help in decision making, specifically with regard to competent embryos and critical points in the process. 

 Currently, there is very limited work using Artificial Intelligence (AI) and computer vision techniques to assess which embryos are most likely to have a successful final implantation. This project will provide a more robust computer vision technology framed in a non-invasive clinical protocol that allows in silico monitoring and evaluation of embryo growth and implantation success with Deep Learning; optimising processes and their performance through computer vision technologies and the development of algorithms based on Deep Learning.

Therefore, the establishment of a new, more robust machine vision technology framed in a non-invasive clinical protocol capable of automatically selecting the embryo most likely to implant from a cohort while alerting of risk situations, would mean a reduction in the time spent by embryologists on data evaluation during laboratory procedures. It would also reduce inter-observer variability.

The expected results are important because of the different family projects (heteroparental, homoparental, single-parent families), the delay in childbearing, the increase in obesity, and many other factors have increased the demand for more personalised assisted reproduction services and led to a change in the value chain approach of assisted reproduction clinics. 

A better, more successful and personalised service is expected to be an attraction for potential patients and a very important competitive advantage that will result in an increase in the number of patients and an increase in turnover.

The classical method based on spot observation by experts has a number of limitations:

– It is a subjective assessment.

– The assessment is made at discrete times, based on point-in-time observation.

– There are negative effects of manipulation of the culture environment.

– The classification of embryos is divided into 4 categories defined by ASEBIR (Association for the Study of Reproductive Biology).

Faced with these limitations, artificial vision techniques and Deep Learning algorithms can mitigate or minimise them, leading to an assessment method with the following characteristics:

– Obtaining objective quantitative parameters.

– Faster evaluation of embryos.

– Alarm system that allows early identification during embryo culture. 

– Provides automated information on a wide variety of parameters (morphokinetic, proteomic, genomic, oxidative stress, etc.) to facilitate decision-making by embryologists.  

Therefore, the aim of this project is to solve the limitations of the aforementioned research by developing a new system based on Deep Learning algorithms and artificial vision technologies that automatically analyse a wide range of parameters that affect the success of implantation. With all this, it will be possible to develop a new diagnostic method that will facilitate decision-making by embryologists, increasing the success of embryo implantation and, consequently, increasing the success of the assisted reproduction treatments offered by the entity.  In addition, the aim is to develop an alert system that allows the early identification of anomalous or suboptimal values in previously identified parameters of interest with respect to the success of implantation. The satisfactory achievement of this project would mean the clinical application of a new protocol for embryo identification and selection, as well as an associated alert system, which will substantially improve the effectiveness of the technique and, consequently, the reproductive success of the treatments offered by the entity.

In this sense, since the incorporation of the technology targeted by this project will enable greater control and better monitoring and follow-up of the process of other centres, and will allow us to identify the relationship between specific problems or inefficiencies, it will be an important attraction for our suppliers or other companies in the pharmaceutical sector or in the development of healthcare equipment with a view to carrying out controlled clinical trials or for the validation of new products or technologies.

Funding Entity

AVI-Generalitat valenciana

Years

2020 - 2021

Share

Partners

ADVICE

ADVICE

Train/Track Dynamic Interaction for the Optimization of Railway Track Maintenance

The objective of this project is the development of a train/track interaction method based on axlebox accelerations while a train runs on a track.

Railway track maintenance is becoming a challenge for Railway Engineers due to the need of meeting increasingly high quality requirements by means of cost-effective procedures. This can be only achieved by implementing some technological developments from other fields into the railway sector, such as Digital Signal Processing. Indeed, the present project delves into data acquisition and processing techniques in order to enhance track surveying processes. In particular, the project is based on train/track dynamic interaction through the gathering and subsequent analysis of axlebox accelerations. For this purpose, the accelerations produced while trains run are gathered and analysed in different ways: varying sampling and filtering frequencies, the location of accelerometers along the train and the different parameters that define some time-frequency representations.

The obtained results show the optimal values for train/track dynamic interaction as well as the best location for the accelerometers. In addition, it is demonstrated that, through spectral analysis and time-frequency diagrams, it is possible to identify and classify diverse track defects, track singularities and vibration modes laying the foundations for the application of digital image processing techniques to railway track maintenance.

Once the different track aspects have been detected and classified, it is also possible to monitor the evolution of the maintenance conditions track geometry and its components.

Agency

Spanish Ministry of Economy and Competitiveness (IDI-20110461)

Years

2011 to 2014

Share

Publications

Partners

FUTURWALL

Futurwall

Immersive study of human behaviour in commercial spaces and product placement

Generation of a modular system for human behavior study in a commercial establishment.

Nowadays, with the aim of carrying out a marketing study, companies hire specific workers to gather information about the type of people who visit their shops, their characteristics, their itineraries and their purchasing trends. This task is arduous and laborious at the same time that it depends largely on the person who registers that data, directly influencing the precision and the quantity of them.

To automate this process, we propose a system with different interconnected modules that perform different but synchronized tasks so that the data obtained is consistent. One of the modules is devoted to obtaining the trajectory of the people who circulate in the trade. For this purpose, a network of cameras with a zenith layout is created, generating a scalable image and processing it in such a way that the trajectories and various metrics that parameterize the user’s behavior are obtained. Another module detects people who pass in front of a linear and the product they take from it. These modules use different types of cameras for image acquisition, both of visible and infrared light and with different types of lenses. To provide a total package, a last module detects the gender and age range of the people who visit the establishment. In this way, an individualized and generic study can be done to redistribute the products and generate offers according to the users.

From the data collected automatically by the system, a report is generated from which the company can make objective decisions, adapt the layout of its products, or customize them according to the characteristics of its clientele.

Agency

Internal financing

Años

2014 to 2017

Share

Partners

VCF EMOCIONAL

VCF EMOCIONAL

Brain Activity Quantification for Sport Audiovisual Content Visualization Using EEG

The overall goal of this project is to understand the feelings and behaviours of football fans when they are watching a match or they are enjoying a title won by their football club.

This project consists in the pre-processing and analysis of electroencephalography (EEG) signals acquired during the observation of audio-visual contents about the Valencia C.F. football team history.

The main objective is to provide answers to different experimental questions about the supporter’s feelings:

  • Are there differences in cerebral activity during the observation of positive and negative emotional videos?
  • May the cerebral activity be objectively quantified and may this metrics and descriptors be used to automatically classify emotional videos?
  • Is it possible to automatically determine the video frames that produce a significant increase or decrease in cerebral activity?

In order to scientifically answer the aforementioned questions the first step is to pre-process the EEG signal. In this phase, the cerebral activity is isolated from physiological artefacts (related to human movements, skin-electrode interface, interferences from other physiological signals, etc.) and external artefacts (related to noise from acquisition devices, 50 Hz interference, etc.). After the pre-processing stage, different features are extracted from the biosignals applying own metrics and descriptors in order to extract the relevant information. Using this information is possible to obtain some conclusions after a classification or statistical analysis stages.

Agency

Internal financing

Years

2013

Share

Publications

Partners

SUPERBOWL

SUPERBOWL

Commercial Content Evaluation from Physiological Signal Analysis

The main goal of this project is to develop an automatic system for commercial content classification. This classification is performed according to the quality of the ad and taking into account physiological information. In addition, some conclusions about the cognitive cerebral processes and the human behaviour during content exposure are extracted.

From electroencephalography (EEG), electrocardiography (ECG), galvanic skin response (GSR) and respiration signal is possible to extract relevant features to quantify attention, memorization and pleasantness levels. The main objective of this project is to develop a system able to classify audio-visual contents using feature extraction from the physiological registers. These registers are acquired while participants are watching a documentary in which different interleaved commercials appears.

After the signal pre-processing (performed to obtain reliable results) different metrics related to the temporal and frequency domains are computed to quantify the user’s feelings during the visualization of each commercial. Using this information automatic predictive models are learned using advanced Machine Learning algorithms.

Another objective of this project is to provide answer to different experimental questions about the human behaviours when commercial contents are exposed to the participants:

  • Are there differences in cerebral activity during the observation of commercials between the population that remember and forget the ad?
  • May the cerebral activity be objectively quantified and may be used to identify potential successful advertisements?
  • Is possible to automatically determine the video frames that produce a significant increase or decrease in the cerebral activity and correlate this cerebral activity with other physiological measures?

Agency

Internal financing

Years

2013 to 2015

Share

Publications

Partners

This website uses cookies to improve your web experience.
Explore
Drag