artificialintelligence – Progress in Research

TROPHY project kicks off

Posted on 28/09/202206/08/2024 by Erik Franco

Activities related to the TROPHY (ulTRafast hOlograPHic FT-IR microscopY) project have officially started. TROPHY is a research project, funded by the European Commission under the Horizon Europe programme, which aims to develop a novel label-free vibrational microscopy approach for cancer diagnosis.

Cancer diagnosis is traditionally done on intraoperative frozen tissue sections by post-surgical histopathologic analysis and, in selected cases, by elaborated and time-consuming molecular diagnosis. The analysis of the biopsy is performed through the staining of the tissue and the evaluation of the morphology of its cells under an optical microscope. This approach is neither fast nor quantitative, has an intrinsic variability in the interpretation depending on the experience of the histopathologist, and provides limited molecular information.

The microscope developed thanks to the TROPHY project will image molecular biomarkers with unprecedented speed and chemical selectivity for a rapid, precise, and non-biased tumor analysis. To this purpose, it will blend in a unique fashion elements of several microscopies developed in the past decades, namely photo-thermal infrared, Fourier transform infrared and Digital Holography Microscopy, bringing them to the unprecedented ultrafast timescale. It will also integrate Artificial Intelligence to produce fast results and assist in the tumour grading process even during surgery.

This microscope will be used to assist healthcare professionals during tumor biopsy diagnostics, provide an accurate diagnosis for curative oncosurgery, guarantee complete resection during intervention, determine the best therapeutic approach tailored to the patient, and identify resistant tumor clones under targeted therapy, paving the way for continual precision medicine in cancer.

The project is coordinated by the Politecnico di Milano with Prof. Marco Marangoni from the Department of Physics as scientific coordinator. The other project partners are Fundacio Institut de Ciences Fotoniques (ICFO, Spain), Consiglio Nazionale delle Ricerche (CNR, Italy), Lyncee Tec SA (LT, Switzerland), Universtaetsklinikum Jena (JUH, Germany), University of Exeter (UNEXE, UK), University of Cambridge (UCAM, UK).

AI-based personalized medical care for lung cancer patients

Posted on 30/05/202206/08/2024 by Erik Franco

I³LUNG is a new research initiative that aims to create a cutting-edge decision-making tool to aid both clinicians and patients in selecting the best lung cancer treatment plan, tailored to the specific needs and situation of each individual patient.

Lung cancer was the leading cause for cancer deaths in men and the second for women in 2020, with 370000 deaths in Europe alone. The consortium has thought out and developed this project to address the primary unmet clinical need in the field of non-small cell lung cancer (NSCLC), which is the lack of biomarkers predicting the response of affected patients to immunotherapy (IO)-based treatments.

The project will use artificial intelligence (AI), in particular deep and machine learning methodologies (DL and ML) to analyze a wide range of information such as baseline clinical features, radiomics, and available biological characteristics of the tumor.

I³LUNG and its partners will have a timeframe of 5 years and a €10M budget to turn their project’s hypothesis in a tangible tool and a new clinical reality. I3LUNG is the first platform enrolling such an important number of patients in both a retrospective (2,000) and prospective (200) manner including such a diversity of multiomic data, arising as an innovative and promising technology to provide an answer to the unmet clinical need of translational research data integration and AI use.

This project is envisioned to both generate novel therapeutic guidelines for clinical practice in lung cancer and support the growth of digital diagnostic tools. AI will push the standard of care towards a more personalized approach for each individual cancer patient. If successful, the approach presented in I³LUNG could in the near future be extended to other cancer types.

The team of the Politecnico di Milano, which combines computer science and biomedical engineering experts, aims to study together with medical partners if and how Artificial Intelligence can become an actor in the complex path of therapy selection. The goal is to develop artificial intelligent solutions that are not only capable of accuracy and precision, but also an understandable interlocutor, at the service of clinical experts, their knowledge and experience and worthy of trust for patients.
Prof. Alessandra Pedrocchi of the Department of Electronics, Information and Bioengineering, project coordinator for our university

The partners in the consortium I³LUNG are:

Fondazione IRCCS Istituto Nazionale dei Tumori(INT, Milano, Italia) with Dr Arsela Prelaj as coordinator of the Consortium
Politecnico di Milano(POLIMI, Milano, Italia)
Istituto di Ricerche Farmacologiche Mario Negri (IRFMN, Milano, Italia)
Istituto Europeo di Oncologia (IEO; Milano, Italia)
ML Cube(Milano, Italia)
LungenClinic Grosshansdorf GmbH (GHD, Grosshansdorf, Germania)
Universitaetsklinikum Hamburg-Eppendorf (UKE, Amburgo, Germania)
Vall d’Hebron Instituteof Oncology (VHIO, Barcellona, Spagna)
Medica Scientia Innovation Research (MEDSIR, Barcellona, Spagna & New Jersey, USA)
Metropolitan Hospital (MH, Pireo, Grecia)
Shaare Zedek Medical Center (SZMC, Gerusalemme, Israele)
Katholieke Universiteit Leuven (KUL, Leuven, Belgio)
Institutet for Halso-OCH Sjukvardsekonomi Aktiebolag (IHE, Lund, Svezia)
University of Chicago (UOC, Chicago, USA)
Aalborg Universitet (AAU, Aalborg, Danimarca)
Lung Cancer Europe (LUCE, Bern, Svizzera)

Photons to create an artificial quantum neuron

Posted on 13/04/202206/08/2024 by Erik Franco

A group of researchers of the Department of Physics at the Politecnico di Milano, the National Research Council (CNR) and the University of Vienna, have developed a device, called a quantum memristor, which could combine artificial intelligence and quantum computing.

This is an advance that can open up as yet unseen potential, allowing to employ the very high computational power guaranteed by quantum technologies in the fields of application of artificial intelligence, which already range from automatic speech interpretation to face recognition, from medical diagnostics to autonomous driving.

Artificial intelligence algorithms are based on mathematical models called neural networks, inspired by the biological structure of the human brain, which is made up of interconnected nodes (neurons). One of the fundamental components of neural networks is the memristor (or memory-resistor), a component that changes its electrical resistance based on a memory of the current that passed through it, in a way that is surprisingly similar to that of neural synapses, i.e. the connections between neurons in the brain.

The group of experimental physicists led by Roberto Osellame (CNR) has shown that it is possible to engineer an optical device with the same functional characteristics as the memristor, capable of operating on quantum states of light and thus encoding and transmitting quantum information: a quantum memristor.

We also simulated an entire optical network made up of quantum memristors, showing that it could be used to learn both classical and quantum tasks.
Andrea Crespi, professor of Experimental Physics at the Politecnico di Milano

This result seems to suggest that the quantum memristor may be the missing link between artificial intelligence and quantum computing, unleashing the potential of quantum resources within artificial intelligence applications.

The study received the cover of the April issue of Nature Photonics magazine.

The study on Nature Photonics

© photo: Equinox Graphics

The partners in the consortium I3LUNG are:

The partners in the consortium I³LUNG are: