The contribution of Artificial Intelligence to the treatment of heart disease
A project that – thanks to the synergy of clinical activity and Artificial Intelligence – aims to define predictive models to characterize patients with heart disease in order to improve their treatment. The project is introduced by Dr. Cristina Panico, cardiologist of the Humanitas.
Aims of the research project
Cardiovascular diseases are the main cause of death in Italy. Both non-modifiable risk factors – such as age, sex, family history and genetic predisposition – and modifiable factors that are notably related to lifestyle – smoking, use of alcohol, poor diet, physical inactivity – play a role in their onset. A large part of medical prescriptions and information on crucial factors such as comorbidities, lifestyle and prescribed medicines are stored in a text format in medical records. Therefore, this type of data is not automatically accessible, as it is not collected in a structured form. The recovery of these data for research studies is currently manual, limiting the possibility of studies on large population samples and lengthening the timeframe for carrying out the studies.
Thanks to the use of Artificial Intelligence – and in particular the techniques of Natural Language Processing (NLP) – it is possible to extend the analysis to data points that are difficult to access using standard techniques, such as data recorded in a free text format. These processing techniques make it possible to read and understand the text and, subsequently, to extract useful information and make it available for any type of analysis. In order to train the neural networks that manage these processes, the guidance of the medical staff is crucial.
Together with the team of engineers and data scientists we create custom models from which the networks themselves will learn while the team supervises and evaluates their learning process. The first aims of the research are focused on improving the characterization of patients affected by cardiovascular diseases – such as heart failure, heart attack and valvulopathies – the stratification of patients according to the prognosis and the validation of new biomarkers – which may be useful for diagnosis – and new treatments and procedures.
Humanitas A.I. Health Center
Humanitas is building a cutting-edge Center focused on Clinical Research – Humanitas A.I. Health Center – that combines the network of doctors and researchers and a team of engineers specialized in Artificial Intelligence. Which scientists are involved in the project and how long will it last? Several key figures are involved in the project including cardiologists, basic researchers and data scientists.
How is medical knowledge and the expertise of data scientists integrated?
The project makes it possible to integrate translational research with artificial intelligence. Clinical questions come from the hospital while new biomarkers and treatments come from laboratories. Therefore, it is possible to define and implement research protocols with the support of the engineers.