Horizon2020 project BRAINTEASER has recently shared its artificial intelligence (AI) models for future application in its European pilot sites that include clinical tool testing amongst patients with multiple sclerosis (MS) and amyotrophic lateral sclerosis (ALS). This is the latest output from the data science project that aims to employ AI solutions to support patients with ALS and MS and their clinicians.
Currently, patients must manage to alternate between hospitalisations and at-home care. This results in both constant uncertainties regarding the timing of acute phases of the disease and a considerable psychological and economic burden for caregivers. At the same time, clinicians require tools capable of supporting them in all stages of patients’ treatments, including precision care and clear indications of urgently needed interventions.
As such, the BRAINTEASER project has developed the first set of AI models to achieve three primary aims: i) to describe disease mechanisms better; ii) to stratify patients according to their phenotype assessed throughout disease evolution; and iii) to predict disease progression in a probabilistic, time-dependent fashion to provide an integrated and broader breadth of information to physicians for improved prognostication and intervention timing in routine clinical practice.
In the coming future, the BRAINTEASER project will strengthen the robustness of these models further to include an assessment of their generalisability and explainability. The purpose of these activities is to achieve a better understanding amongst clinicians concerning which variables contribute to predictions. Professor Barbara Di Camillo from the University of Padua, who also serves as co-coordinator of the EU project, affirmed, “…This information must be provided in a readable and user-friendly format to increase the likelihood of acceptance by clinicians. We want physicians to feel comfortable and secure about the technology we are creating.” Furthermore, as it relates to disease progression predictions and patient stratifications, the models will be set to consider other aspects of disease evolution. This comprises taking into account technical, medical, psychological and societal needs.
Nevertheless, for the models to reach their full potential, the BRAINTEASER project must count on the support and involvement of stakeholders of all levels–from patients and caregivers to clinicians and social service workers and many more. Their input will help provide insights into relevant outcomes, unmet needs, and readability and ensure that the solution fits prospective real-world applications.
Beginning in winter 2023, BRAINTEASER will test these AI models using the project’s most recently developed clinical tool in pilot sites across Europe. These pilot sites will gather information and learnings via specially formulated questions from patients and clinicians to refine system functionalities and enhance overall efficacy.
