Riccardo Miotto, PhD, is an Assistant Professor within the Department of Genetics and Genomic Sciences and a member of the Hasso Plattner Institute for Digital Health at Mount Sinai. Dr. Miotto is a computer scientist and data engineer with extensive experience in developing algorithms of information retrieval, data mining, and machine learning for real-world data collections. His primary expertise encompasses the study and design of novel frameworks to process health-care data for personalized medicine and medical search engines.
Dr. Miotto obtained his PhD in Information Engineering from the University of Padova, Italy, in 2011. During his graduate studies, Dr. Miotto worked on music information retrieval, in particular music auto-tagging and playlist recommendation. Several of his publications have shown that machine learning applied to the audio signals of the songs leads to effective search engines for music documents, such as playlist generation, cover identification, and music annotation.
Dr. Miotto’s current research at HPI•MS focuses on the secondary use of electronic health records (EHRs) and on the development and application of solutions to extract meaningful representations from patient data that can be used for clinical prediction and medical analysis. In this domain, he has been a pioneer in applying deep learning to EHRs through a “deep patient” methodology to predict the potential future clinical outcomes of patients from their clinical status. Dr. Miotto also investigates novel technologies for health-care monitoring (e.g., smart mirror), including applications to sports medicine.
In previous experiences Dr. Miotto also worked on search engines for clinical trials through processing of free-text eligibility criteria (based on natural language processing techniques) and similarity of patient EHRs and machine learning applied to music information retrieval, in particular semantic discovery and recommendation, automatic tagging, and cover identification. Dr. Miotto has also investigated potential improvements to clinical trial recruitment and his work showed that applying natural language processing to free-text eligibility criteria increases the performances of the ClinicalTrial.gov search engine. This research also showed that patient similarity in EHRs helps to identify potentially eligible patients for a trial from the currently enrolled participants.