Analysis techniques based of Artificial Intelligence, including machine learning and deep-learning approaches, are widely used in medical diagnostics and therapy. Starting from sensor data processing for image reconstruction, specific solutions include a variety of data mining, image segmentation, annotation and analysis applications, to end with intelligent systems for computed-aided diagnosis and image-guided therapy.
Artificial Intelligence in Medicine (AIM)
The Artificial Intelligence in Medicine (AIM) project aims to exploit the expertise of INFN and associated researchers on medical data processing and enhancement, and turn it in the development of advanced and effective analysis instruments to be eventually clinically validated and translated into products.
The AIM Rationale In Brief
Big companies are quickly developing and placing on the market intelligent systems applications to assist clinicians in their daily tasks. Nevertheless, research institutions can provide relevant contributions in this still-open field of research. In particular, INFN can take advantage of its unique expertise in big data handling inherited for high-energy physics experiments and to the availability of extremely powerful computing centres mainly built to store and process those data. More importantly, a network of fruitful interactions between INFN Physicists and Clinicians of several Italian hospitals and clinical research centers has been built in the last two decades, thanks also to specific research initiatives funded by INFN-CSN5.
Image Modalities: mammography, RX, DBT, CT, MRI, fMRI, PET, SPECT, US
Software tools and packages: Matlab, Freesurfer, SPM, Python, PyRadiomics, Scikit-learn, Keras, TensorFlow
Methods: Image processing, data analysis with statistical methods, ML (ANN, RF, SVM, Network models) and DL (CNN, CAE, U-net)
Research Topics: Data Harmonization; Quantification; Predictive Models