Periodic Reporting for period 1 - CoMPaSS-NMD (Computational Models for new Patients Stratification Strategies of Neuromuscular Disorders)
Reporting period: 2023-05-01 to 2024-10-31
CoMPaSS-NMD project, proposed by a heterogeneous consortium of European organizations with clinical, biotechnological, ICT, AI, ethical and legal, communication and exploitation competences, aims to create an innovative and universal AI based tool for the accurate diagnostic stratification of Hereditary NeuroMuscular Diseases (HNMD) patients. The final objective is to facilitate personalized treatments and generate evidence-based guidelines for patient management to support health care professionals. CoMPaSS-NMD by leveraging AI-based method will be able to produce robust data sets (genetic, Nuclear Magnetic Resonance, histopathologic) to develop computational tools delivering, as the main project outcome, the public AI-based CoMPaSS-NMD Atlas Platform able to provide precise clinical characterization, supporting patients, caregivers, researchers and health professionals.
Initially, the consortium is conducting a retrospective observational study on previously collected genetic, imaging/radiomic, and histopathological data from clinical centers in Great Britain (UNEW), France (CERBM), Finland (SFF), and Italy (FSM). This study will inform the design of algorithms to stratify patients and identify HNMDs. Subsequently, a prospective study will be conducted on new clinical, genetic, imaging/radiomic, and histopathological data from 500 HNMD patients in Italy (FSM, UNIMORE) and Germany (LMUM) to validate the algorithms.
Ultimately, the project aims to deliver the ATLAS Platform, a comprehensive web repository that integrates multiple sources of HNMD metadata, revolutionizing the understanding and clinical treatment of HNMDs. This will impact creating a new patient stratification tool integrating multi-omics data that will generate a 30% increment of the HNMD diagnostic rate.
Initially, the consortium is conducting a retrospective observational study on previously collected genetic, imaging/radiomic, and histopathological data from clinical centers in Great Britain (UNEW), France (CERBM), Finland (SFF), and Italy (FSM). This study will inform the design of algorithms to stratify patients and identify HNMDs. Subsequently, a prospective study will be conducted on new clinical, genetic, imaging/radiomic, and histopathological data from 500 HNMD patients in Italy (FSM, UNIMORE) and Germany (LMUM) to validate the algorithms.
Ultimately, the project aims to deliver the ATLAS Platform, a comprehensive web repository that integrates multiple sources of HNMD metadata, revolutionizing the understanding and clinical treatment of HNMDs. This will impact creating a new patient stratification tool integrating multi-omics data that will generate a 30% increment of the HNMD diagnostic rate.
In order to create a new generation of methods for “precision diagnosis” that allow researchers and healthcare professionals to successfully classify patients affected by Hereditary Neuromuscular Diseases (HNMDs) the consortium have worked on retrospective data collection and analysis through the application of methods based on the use of Artificial Intelligence (AI) and Machine Learning (ML) Developing procedures for efficient noise filtration, data normalisation, and feature extraction; tuning of the feature selection methods and implementing and testing of the clustering algorithm for high-dimensional single omics data with complex internal structure. Moreover, also through co-design activities aimed to the requirements analysis and feedback from the SAB, it was implemented the first ATLAS operating platform prototype, while, in parallel the consortium is still working on SOPs (standard operating procedures) to collect – through the platform - clinical, genetic, Nuclear Magnetic Resonance, histopathologic data from 500 new patients which are under recruitment in Italy (UNIMORE, FSM) and Germany (LMUM). Also, partners on the data clusters and superclusters, carried out the designing and development of the framework for local modelling being compliant with privacy regulation and establishing privacy metrics in the algorithm execution. Finally, the consortium, with reference to the prospective data, developed the research protocol and informed consents for the use of health data and a proper data management plan. Finally, partners are carrying out engagement activities, among which the first delivered training and summer school, which were able to reach more than 50 people around the world. Consortium, targeting stakeholders as well as citizens and patients, is also promoting the project through events, social media, networking activities with the aim to build up a CoMPaSS-NMD community able to promote a change on how to approach and treat HNDMs innovating the diagnostic stratification.
Till now, CoMPaSS-NMD was able to analyse retrospective data (genetic, MRI, histological) received from clinical partners (UNEW, LMUM/KUM, CERBM, SFF) with the aim of identifying and describing patient subpopulations based on their genetic, MR imaging and histopathological data. Additionally, the algorithm verification was performed using genomic and histopathological data from the FSM (not foreseen in the grant agreement). Using the available datasets, the construction of the data filtering and clustering pipeline for genetic data was completed. Moreover, the first working prototype of the ATLAS platform architecture was delivered as a demo test to collect feedback from partners. The consortium is committed to deliver high quality data and results. Moreover, project is faster attracting stakeholders and both patients/caregivers and clinicians/researchers are demanding AI tool to deliver more appropriate diagnosis. Thus, CoMPaSS-NMD invite interesting people and patients to participate in the clinical research in order to collect clinical data, from muscle MRI (on clinical indication), muscle biopsy (on clinical indication) and DNA analysis from patients suffering from hereditary neuromuscular diseases https://compass-nmd.eu/details-of-the-compass-nmd-clinical-research/(opens in new window) enlarging the data set to validate the model creating a new generation of methods for “precision diagnosis”.