Assistant Professor


Email: casiraghi (at) di (dot) unimi (dot) it

Research interests: Medical and Biomedical Applications, Computer Aided Diagnosis Systems, Digital Image and Signal Processing, Pattern Recognition, Arti cial Intelligence, Learning Theory and Application.

Research description: Elena Casiraghi's interest in the information technology research field date to the year 2000, when she worked in the Information Technology Department in VTT (Valtion Teknillinen Tutkimuskeskus, Helsinki, Finland), developing Virtual Reality applications and face recognition applications (project in cooperation with Nokia). Her research work in the Department of Informatics of Milan started with the deveopment of applications for face localization, identification, and recognition, by employing supervised and unsupervised learning algorithms.After she focused in the field of artificial intelligence, to develop automatic systems for medical and biomedical image processing and pattern recognition. Since then, she has been developing several collaborations. Specifically, she has worked on digital chest radiographs, abdominal computer tomography scans, magnetic resonance images of fetal brains, mouse images produced by molecular imaging, tissue images stained with different biological procedures and acquired by digital microscopes.She has also investigated novel learning algorithms for pattern recognition, manifold learning, and intrinsic dimensionality estimation, to develop novel theories and automatic algorithms dealing with high-dimensional datasets characterized by a small cardinality (Small Sample Size Problem). These researches led to the development of methods whose performance has been evaluated both by the comparison with state of the art techniques and by tests on synthetic and real datasets related to problems in the fields of signal processing, image analysis, and bioinformatics. In the past two years, she has been collaborating with the biological researchers of Consorzio M.I.A - Microscopic Image Analysis (University of Milan-Bicocca) to develop automatic systems for the analysis quantification and comparison of serialized microscopic images of arteriosclerotic plaques, with the aim of investigating the main factors behind carotid plaques' instability, the latest being the main cause of cerebral stroke. Due to its promising results the developed system (called MIAQuant) has been lately adapted and generalized through the usage of machine learning techniques in order to be able to process images depicting tissue sections belonging to different body structures. Precisely, the novel system (called MIAQuant\_Learn) extracts, quantifies and analyze the co-existence of markers characterized by any color and shape and being stained in contiguous sections extracted from any body tissue. The promising results obtained by the MIAQuant\_Learn motivate its extensive usage in the oncological field to quantify and analyze cancerous tissues images produced either by Ospedale San Raffaele (Milano) and by the Department of Experimental Oncology and Molecular Medicine (Fondazione IRCCS Istituto Nazionale dei Tumori).