logo and pictures


ADAM

Advanced Data Analysis in Medicine is a Research & Development Group for software solutions for Medicine, based in Lecce (Italy) in the "Laboratorio di Fisica Biomedica e Ambiente" of the University of Salento. It is associated with DReAM, Laboratory of interDisciplinary Research Applied to Medicine.

ADAM includes researchers  with competences in Physics, Informatics, Life Sciences, with a common experience in the field of software applications for Medicine.

Our concern is the advanced processing of bio-medical data, above all for diagnostics, but also for prognosis and therapy assessment, in various interest fields. Expert in the computerized processing of medical data with main interest on diagnostic images, our objective is to make some medical-diagnostic procedures automatic and operator-independent, in particular but not limited to the oncologic field.

Recently, our interest extended to bio-medical data of different nature, and we are developing software tools for EEG (electroencephalography) data processing, for the achievement of BCI (Brain-Computer Interface) systems.

CAD systems and Radiomics

Most of the software we build falls into the category of CAD or CADe (Computer-Assisted Detection) systems. They are pattern-recognition / machine-learning systems based on physical-mathematical models, on the calculation of discriminating features, on classification methods such as artificial neural networks. The purpose of these systems is the automatic detection of structures and pathologies in biomedical data (often, diagnostic images from Magnetic Resonance, Computed Tomography, and others). In some cases the term CADi is used, when the aim is not only detection, but also to support the physician in the diagnosis procedure, in turn related with prognosis and therapy.

An important keyword in this context is Radiomics, which is the automatic (and massive) extraction of measurable features from diagnostic images, and possibly their association with the patient genetic profile (Radiogenomics).

In recent times, machine learning has evolved into Deep Learning, an Artificial Intelligence approach which mimics the workings of the human brain in processing data e.g. when detecting objects. Deep learning, employes a hierarchical level of artificial neural networks to carry out the process of machine learning.

Some past projects

Automatic Segmentation of Brain Gliomas in Diffusion-Tensor Magnetic Resonance Images S.Raffaele

glioma In collaboration with the IRCCS San Raffaele Hospital and Università Vita-Salute San Raffaele, Milan (Neuroradiology Unit), we have developed a CAD system for the automatic or semi-automatic detection of brain gliomas in magnetic resonance imaging of different modality (T2W, FLAIR, DTI).

Chemiotherapic Follow-up of Brain Gliomas S.Raffaele

follow-upBrain gliomas invade surrounding tissues along white matter fibers, spreading beyond the pathological area highlighted by conventional MRI (Magnetic Resonance Imaging). The response to chemotherapy treatment is currently assessed on the basis of the variation in tumor volume, often carried out visually. The system we have developed, in collaboration with the IRCCS San Raffaele Hospital and Università Vita-Salute San Raffaele, Milan (Neuroradiology Unit),  performs a point-by-point assessment based on diffusion tensor maps.

Pleural Nodule Detection in Chest CTINFN

polmoneA CAD has been developed for the automatic detection of pleural nodules in pulmonary CT images. The project was part of the INFN experiments called MAGIC5 (Medical Applications on a Grid Infrastructure Connection) and M5L (MAGIC5 Lung).

MRI for Early Diagnosis of Alzheimer DiseaseINFN

alzheimerWe contributed to the development of an early detection system for neurodegenerative diseases, particularly Alzheimer's, in magnetic resonance imaging. The project was the responsibility of the INFN experiment called MAGIC5 (Medical Applications on a Grid Infrastructure Connection).

Current projects

Brain-Computer Interfaces  CRIL

bci
This project concerns the creation of a BCI based on EEG data. First results show that features calculated from EEG allow recognizing vowels in silent speech.      Collaboration with Mirko Grimaldi from CRIL, Centro di Ricerca Interdisciplinare sul Linguaggio, University of Salento.

Computational Fluid Dynamics in Stenotic Tracheas   San Camillo ForlaniniDISTEBA

trachea
This project concerns the CFD simulation of airflow in healthy and stenotic tracheas, with the purpose of correlating clinical parameters with CFD outcomes. Collaboration with Riccardo Buccolieri from DISTEBA (Dipartimento di Scienze e Tecnologie Biologiche e Ambientali, University of Salento), and the group of Gianni Galluccio MD from Azienda Ospedaliera San Camillo Forlanini in Rome.

Discrimination of Infiltrative vs In-Situ Breast Cancer in DCE-MRI  DiSumma_Perrino_Hospital_Brindisi

follow-up Breast cancer can be classified into two main groups: in situ and infiltrative, with the latter being the most common malignant. The prototype of a CAD system for the discrimination between in situ and infiltrating tumours is under development. Collaboration with the group of Maurizio Portaluri MD from the Di Summa - Perrino Hospital in Brindisi.

Reflectance Confocal Microscopy for Cutaneous Melanoma Detection  
UNIMOREUNIMORE

confocalRCM is  a well-established technique that has been shown to significantly increase the sensitivity and specificity of melanoma diagnosis when compared with the traditional naked-eye examination. A CAD system is under development to support the physician during diagnosis. Collaboration with Città di Lecce Hospital, with Massimo Federico MD and the group of Giovanni Pellacani MD from Università di Modena e Reggio.

Congenital Diaphragmatic HerniaPoliclinico Milano

ecmoA study is starting on the possibility of predicting the occurrence of pulmonary hypertension (PH) in the newborn from fetus radiomics features in MRI. As PH is related to the need of ECMO (ExtraCorporeal Membrane Oxygenation) for the newborn, early prenatal assessment may be vital for therapy. Collaboration with the group of G. Cavallaro MD PhD from Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Università degli Studi di Milano.

Members

Giorgio De Nunzio
Università del Salento
Scopus, ORCID, ResearchGate, Google Scholar, LinkedIn, unisalento web page, other web pages, e-mail Giorgio De Nunzio

Luana Conte
Università del Salento
ORCID, e-mail Luana Conte

Benedetta Tafuri
ex Università del Salento, now with Università di Bari
LinkedIn Benedetta Tafuri