Leukemia and Genomics: the Role of Artificial Intelligence

A project that, thanks to the synergy between the clinic and Artificial Intelligence, aims to trace the risk profile of the individual to develop leukemia, taking into account the presence of mutations within the genome of blood stem cells. There are several studies active in the field of onco-haematology and new technologies can be of great help in the field of research.

Artificial Intelligence (AI) has an important role in haematology. In the last decade, genomics – i.e. the study of the genome, the set of genes – has proved to be fundamental in clinical activity and several clinical decisions are made taking into account molecular information. This association between genomics and clinics is also demonstrated by the fact that in most cases blood diseases – acute and chronic leukemias – are characterized by the presence of specific molecular alterations.

The integration between genomics and clinic has several consequences: we have molecular and clinical data that can be analyzed with innovative technologies in order to improve the patients’ diagnosis, predict the individual prognosis and choose the most appropriate therapy for each patient, predicting its effectiveness.

In this kind of challenge, artificial intelligence is an innovative tool that could be able to integrate the data we have and translate them into useful information for the clinical management of the patient.

Humanitas A.I. Health Center Humanitas is a Center focused on Clinical Research which combines the network of doctors and researchers and a team of engineers specialized in Artificial Intelligence.

Aim of the project

We have always known that blood tumors are not known tumors and cannot be prevented at the primary level, but the growing knowledge of genomics is making us realize that haematological diseases run in the family in some cases. In other cases, it is possible to detect the healthy subjects who will develop acute or chronic leukaemia. Thanks to some very innovative studies, it has been possible to notice in the blood of some elderly patients the presence of some mutations – somatic ones, i.e. acquired and not congenitally present in the genetic heritage – related to a higher risk of developing acute and chronic leukaemias.

In the genome of haematopoietic stem cells, in fact, somatic mutations can occur, especially due to the cellular aging process. This alteration determines a genomic instability that brings with it an increased risk of acquiring further mutations that induce the appearance of blood diseases.

We hypothesize that the analysis of the data of about 5 thousand elderly patients – who were followed for twenty years in a study conducted by the Mario Negri Institute in Milan, a partner of this project – may help to define more specifically the relationship between specific mutational models in the hematopoietic system and the individual risk of developing leukemia.

The use of Artificial Intelligence will allow us to integrate all the data at our disposal and identify genetic associations and key information on the different subgroups to interpret the global survival of our population. Finally, we will combine genomic and clinical information in predicting the individual prognosis by creating a personalized risk profile of the individual.

The project will last 3 years and 8 researchers are involved (clinical researchers and data scientists). The integration of medical and technological competences represents the added value of this project. Clinical knowledge in haematology is relevant to accurately define the relevant questions to be answered by data analysis, while technological skills will be used to define methods of analysis of complex data allowing a global view of the clinical and molecular profile of the individual.


Humanitas Research Hospital – Via Manzoni 56, Rozzano (Milano), Italia – P. IVA 10125410158 Autorizzazioni: Decreto Presidente Regione Lombardia n°1906 del 18.04.1996 Direttore Sanitario: dott. Michele Lagioia.