BrainMap

Mr. Adrian Groza

Professor - UNIVERSITATEA TEHNICA DIN CLUJ - NAPOCA

Researcher | Teaching staff | PhD supervisor

Curriculum Vitae (28/07/2026)

Expertise & keywords

New Optical Coherence Tomography Biomarkers Identified with Deep Learning for Risk Stratification of Patients with Age-related Macular Degeneration Call name: P 2 - SP 2.1 - Proiect experimental - demonstrativ
PN-III-P2-2.1-PED-2021-2709 2022 - 2024

Role in this project:

Coordinating institution: UNIVERSITATEA DE MEDICINA SI FARMACIE "IULIU HATIEGANU"

Project partners: UNIVERSITATEA DE MEDICINA SI FARMACIE "IULIU HATIEGANU" (RO); UNIVERSITATEA TEHNICA DIN CLUJ - NAPOCA (RO)

Affiliation:

Project website: http://old.medicina.umfcluj.ro/component/content/article/8-ro/703-delarmad?Itemid=216

Abstract:

Age-related Macular Degeneration (AMD) is one of the commonest causes of vision loss worldwide. In Europe, in 2040 the number of individuals with early AMD is expected to range between 14.9 and 21.5 million and with late AMD, between 3.9 and 4.8 million. Low vision severely impacts the quality of life among populations and poses an enormous global financial burden. Optical Coherence Tomography is a non-invasive imaging method that revolutionized ophthalmology and whose continuous improvement opened new perspectives on the management of retinal diseases. However, the high number of people with AMD requiring constant OCT monitoring results in a significant burden for the ophthalmologists. Simultaneously, the evolution of artificial intelligence (AI) has led to the development of methods capable to perform high-level analysis of images and other data. Consequently, AI-based healthcare services are expanding, with expectations of $6.6 bn by the end of 2021. The aim of this research is to develop Deep Learning (DL) instruments for an integrated solution supporting the clinical decision making for AMD patients. The medical novelty results from the search for better explanations of AMD pathogenesis based on DL techniques. From our knowledge, this is the first national project that applies DL techniques on OCT data gathered from the patients with AMD. All of them are in the records of the Ophthalmology Department, University of Medicine and Pharmacy from Cluj. Technologically, the novelty is issued by learning to predict the evolution of the disease based on sequence of images taken from successive visits, and by explaining to the ophthalmologist how the IA-tool has reached the conclusion. The project has a high potential to significantly contribute to the advancement of knowledge in the field and its results will increase the confidence in AI solutions within the scientific, medical and patient communities.

LELA - Collaborative Recommendation System in the Tourism Domain Using Semantic Web Technologies and Text Analysis in Romanian Language Call name: Cecuri de Inovare 2013
PN-II-IN-CI-2013-1-0120 2013 - 2014

Role in this project:

Coordinating institution: UNIVERSITATEA TEHNICA DIN CLUJ - NAPOCA

Project partners: RECOGNOS ROMANIA S.R.L. (RO); UNIVERSITATEA TEHNICA DIN CLUJ - NAPOCA (RO)

Affiliation: UNIVERSITATEA TEHNICA DIN CLUJ - NAPOCA (RO)

Project website:

Abstract:

Serviciul consta in dezvoltarea unei solutii arhitecturale si prototipizarea acesteia. Sistemul primeste ca intrare setul de texte in limba romana colectate de Recognos in domeniul turismului. Sistemul va furniza 4 capabilitati:

1. Identificarea automata a punctelor de interes, restaurante, hoteluri si alte obiective turistice conform unei taxonomii specificate de beneficiar, prin analiza textului in limba romana (acuratete 80%).

2. Extragerea si clasificarea opiniilor existente despre un anumit obiectiv, in 2 clase conform SentiWordNet (acuratete 80%), respectiv 6 clase conform WordNetAffect (acuratete 65%).

3. Furnizarea unei interfete de interogare a obiectivelor turistice in limbaj natural controlat cu suport de ghidare a interogarii conform ontologiei utilizate (grad de acoperire sabloane lingvistice cu vocabularul de baza al limbii de 80%).

4. Construirea de recomandari turistice pe baza preferintelor utilizatorilor si furnizarea la cerere de explicatii si argumente ale recomandarii generate.

Solutia propusa trebuie sa fie compatibila cu RDF AllegroGraph, utilizat cu licenta de Recognos. Dezvoltarea va fi iterativa, deciziile de proiectare si implementare vor fi argumentate de catre membrii responsabili pentru sarcinile specifice prin intalniri saptamanale cu beneficiarul.

Obiectivele tehnice sunt:

O1. Dezvoltarea ontologiei in domeniul turismului in limba romana (Dec 2013)

O2. Dezvoltarea metodei de populare automata a ontologiei (Mar 2014)

O3. Dezvoltarea modelului de clasificare a opiniilor (Apr 2014)

O4. Dezvoltarea sistemului suport de interogare in limbaj natural controlat (Mai 2014)

Necesitatea unui sistem suport turistic orientat pe limba romana este data de: i) profilul consumatorului de servicii turistice interne ii) doar 10% din calatoriile romanilor sunt externe, cu o medie de 2 nopti la nivel national, iii) slaba utilizare a resurselor (bloguri, opinii, descrieri) in generarea de recomandari personalizate si in promovarea obiectivelor interne.

FILE DESCRIPTION	DOCUMENT
List of research grants as project coordinator	Download (41.63 kb) 17/05/2016
List of research grants as partner team leader
List of research grants as project coordinator or partner team leader
Significant R&D projects for enterprises, as project manager	Download (39.62 kb) 17/05/2016
R&D activities in enterprises
Peer-review activity for international programs/projects