Embedding Artificial Intelligence in routine flow of diagnosis and care in the pediatric hematology unit of a university hospital: Challenges and solutions from the GenoMed4All project in Italy
Abstract
Sickle cell disease (SCD) is a genetic disorder characterized by chronic hemolytic anemia and vaso-occlusive crisis. Individuals present acute and chronic complications. Padua University Hospital is a public hospital, the regional reference center for pediatric hematology and member of a Pediatric Hematology Network (AIEOP); it is also center of excellence of the European Reference Network (ERN) EurobloodNet. Since 2021 we participate in “Genomic and Personalized Medicine for all Through Artificial Intelligence in Hematological disease” (Genomed4ALL-Horizon 2020), in the SCD use case, but faced many challenges as a public health institution to be involved with Artificial Intelligence (AI)
Challenges of health data standardization, harmonization and interoperability for rare hematological disorders in Europe in the GenoMed4all Project
Abstract
Poster presented by Mirco D’Agnolo (Università di Padova) at the AI4H Padova – Artificial Intelligence for Healthcare congress, under the title: “Challenges of health data standardization, harmonization and interoperability for rare hematological disorders in Europe in the GenoMed4all Project”
Benefits of Radiomics in hematology: improved Identification of Cerebral Silent Infarct across the lifespan in a real-world Sickle Cell Disease European Cohort
Abstract
Poster presented by Maria Paola Boaro (Università di Padova) at the AI4H Padova – Artificial Intelligence for Healthcare congress, under the title: “Benefits of Radiomics in hematology: improved Identification of Cerebral Silent Infarct across the lifespan in a real-world Sickle Cell Disease European Cohort”
Clinical and Genomic-Based Decision Support System to Define the Optimal Timing of Allogeneic Hematopoietic Stem Cell Transplantation in Patients with Myelodysplastic Syndromes
Abstract
Allogeneic hematopoietic stem cell transplantation (HSCT) is the only potentially curative treatment for patients with myelodysplastic syndromes (MDS). Several issues must be considered when evaluating the benefits and risks of HSCT for patients with MDS, with the timing of transplantation during the disease course being a crucial question. Recently the integration of genomic screening (by Molecular International Prognostic Scoring System, IPSS-M) into patient’s assessment has resulted into a significant improvement in predicting clinical outcomes with respect to the conventional prognostic score (Revised IPSS, IPSS-R), including better stratification of post-HSCT outcome.
Here, we aimed to develop and validate a Decision Support System to define the optimal timing of HSCT in MDS patients based on clinical and genomic information as provided by IPSS-M vs conventional IPSS-R.
Scalable and Portable Federated Learning Simulation Engine
Abstract
Federated learning (FL) is one of the most promising approaches to ensure privacy in the application of data-driven techniques to sensitive information. However, the implementation of such approaches in a production environment is still an important challenge. In this paper, we present a scalable, portable, hardware-independent, model-agnostic FL Simulation Engine (FLSE) with the aim of easing the job of researchers who want to train FL models to be deployed in production environments. The FLSE offers a tool that can be used both standalone or embedded within a larger architecture, it can be deployed seamlessly and allows concurrent, scalable, and highly available V&V assessment support for FL models. The tool allows researchers to understand the behaviour, in terms of metric performance, of their proposed models in production scenarios, allowing a boost in trustworthiness towards ethical AI.
Integrative diagnosis of Sickle Cell Disease patients for Personalized Medicine
Abstract
Sickle cell disease (SCD) is a chronic life threatening disorder, caused by the presence of structurally abnormal adult hemoglobin S (HbS). Under low oxygen saturation, HbS forms hemoglobin polymers that deform the red blood cell structure, referred to as ‘sickling’. Sickled erythrocytes result in hemolytic anemia and recurrent vaso-occlusive crisis, which lead to long-term morbidity and early death. The patient specific pO2 at which sickling starts (PoS) along with RBC deformability at normoxia (EImax) and upon deoxygenation (EImin) can be measured by oxygen gradient ektacytometry (Laser Optical Rotational Red Cell Analyzer (LoRRca)). In the GenoMed4ALL project, oxygen gradient ektacytometry data will be integrated with genomics, metabolomics and clinical data of 1000 SCD patients, allowing better characterization of SCD and development of Artificial Intelligence (AI) algorithms for personalized medicine.
Radiomics and Artificial Intelligence for identification and monitoring of silent cerebral infarcts in Sickle Cell Disease: First analysis from the GenoMed4All European Project
Abstract
The use of Artificial Intelligence (AI) for personalized medicine has recently guided improvements in the diagnostic pathway of many diseases. The EU Project GENOMED4ALL: “Genomics and personalized Medicine for All through Artificial Intelligence in Haematological Diseases” aims at using European data of patients affected by Sickle Cell Disease (SCD) to find correlation between -omics data – and phenotype, seizing the opportunity to improve diagnostics through AI.
2,3-diphosphoglycerate detection via direct infusion high-resolution mass spectrometry correlates with quantitative detection in blood patients with SCD
Abstract
Sickle cell disease (SCD) is a hereditary and chronic life-threatening disorder, characterized by haemolytic anaemia. Increased 2,3-diphosphoglycerate (2,3-DPG) concentrations, along with decreased oxygen affinity of hemoglobin, may be related to the variability of clinical outcomes in SCD. Furthermore, genomic health data holds promise to improve the prediction of disease severity in SCD. Based on the integration of genomics, metabolomics and clinical data from 1000 SCD patients, to be included in 2022, GenoMED4all aims to develop Artificial Intelligence (AI) based deep learning algorithms to improve the prediction of disease severity and phenotype in SCD.
Artificial Intelligence-based Deep Learning algorithms for patients with Sickle Cell Disease
Abstract
Sickle Cell Disease (SCD) is a hereditary red blood cell disorder characterized by hemolytic anemia, periodic painful ischemic vascular occlusion and long-term multiorgan failure. Pathophysiology of SCD is not completely understood and disease phenotypes vary
largely. The only curative treatment is hematopoietic stem cell transplantation. This is however limited in its availability and not without risks. ERN-EuroBloodNet now leads a SCD use case within the GenoMed4All initiative.