Current students


D'ARNESE ELEONORACycle: XXXV

Section: Computer Science and Engineering
Advisor: SANTAMBROGIO MARCO DOMENICO
Tutor: MARTINENGHI DAVIDE

Major Research topic:
CHiMEra: Custom Healthcare solutions for MEdical support diagnostic tool

Abstract:
In recent years, big data analysis is a concept that is becoming more and more prominent in healthcare, where the advent of new technologies is making it possible to capture a large amount of information over vast timeframes which, following current trends, are being digitalized under specific juridical requirements. Aside from the volume, medical data present extreme diversity of data types (e.g., imaging data, patient/physician-generated data and sensors data), and acquisition parameters within the same data type. Besides, a fast and reliable analysis is essential to provide the best care to patients. In fact, by discovering patterns, and relationships within the data, there is the potential of improving care, allowing early detection of diseases when they are more responsive to treatment. Nonetheless, the large amount of healthcare data makes unfeasible a pure human analysis in a reasonable time. Indeed, the employment of computer-aided tools offers many benefits, including speeding up the analysis up to 40%, which allows doctors to focus on patients and treatment decisions.
To provide these benefits different efforts have been put in exploring and developing tools that will help physicians in clinical day life. Even though physicians agree that standardization and fast analysis are needed, most of the data largely remain inadequately organized. They, indeed, differ in modality, dimension, and quality, which introduce new challenges, such as data integration and mining, especially if multiple datasets are involved. Understanding the dependencies among the data and designing efficient workflows demand new computer-aided tools able to manage data standardization and acceleration.
;
This research project aims at investigating the generation computation pipelines for medical data analysis, finding a common ground to the multiple subfields that currently the state of the art treats separately, leading to a modular yet integrated computer-aided solution spanning the most suitable computer architectures.