Precision ALS highlights potential for collaborative approaches to Remote Monitoring, Machine Learning and Neurofilament studies at ALS/MND Conference

Precision ALS conference in Basel leads the way for huge potential in collaborative approaches to Remote Monitoring, Machine Learning and Neurofilament studies in Amyotrophic Lateral Sclerosis (ALS) at conference in Basel, Switzerland.

The Precision ALS project (P-ALS) featured at the 34th International Symposium on ALS/MND recently in Basel, Switzerland. The three-day conference was organised by the Motor Neuron Disease Association and brought together researchers from across the world to share new understandings of amyotrophic lateral sclerosis and motor neuron disease. P-ALS had one presentation and seven posters accepted into the conference representing eight P-ALS sites. The project also held a closed satellite meeting on Monday the 4th December to discuss the programme’s progress with academics and industry partners.

During the conference, Precision ALS CTO, Dr. Anthony Bolger, presented the current status of the Data Collection Tool and with his technical team consisting of Matthew Nicholson and Frances Gibbons. The team also gave demonstrations of the upgrades to the innovative application which included increased functionality around carryover and recurrence of data within encounters. Future works will include improvements in offline functionality, synchronising data and building more stability.

Director of the Precision ALS research programme and Professor of Neurology at Trinity College Dublin, Professor Orla Hardiman, summarised the extant dataset to attendees, describing it as “the largest dataset of its kind, capable of being interrogated”. She surmised that the vision for P-ALS was to harmonise data from across the sites to create a user-friendly platform where data could be migrated out creating an intelligible powerful repository capable of being resourced over a long time. The final platform will represent an automated system that maintains long term viability/sustainability to support clinical input into the system.

During the conference, three productive workshops were held where clinical, platform development and industry parties openly discussed the future of the project. Dr. Dara Meldrum led remote monitoring and digital outcome measures and other groups shared relevant experiences. Future works will identify barriers and limitations as well as the burden on both patients and caregivers with use of remote monitoring devices aiming to improve clinical care and oversight of patients in the home.

Critical input to conversations came from inciteful presentations from both Ruben van Eijk (Utrecht group) and Chris McDermott (Sheffield group). ADAPT academic Dr. Robert Ross (TU Dublin) led the Machine Learning (ML) workshop where he mediated key presentations from three core groups leading the way in ML models. Dr. Barbara di Camillo (University of Padova) presented their model based on Dynamic Bayesian networks for deep learning and the “Brainteaser” BRinging Artificial INTelligencE home for a better cAre of amyotrophic lateral sclerosis and multiple SclERosis project (https://brainteaser.health). Dr. Jonathan Cooper-Knock from Sheffield Institute for Translational Neuroscience (SITraN) presented their Pykox sublime Royston palmer model and their PredicTTE App for time to event prediction tasks by non-experts (www.predictte.org). Dr. Henk-Jan Westeneng from prediction model for survival that has been used in recent clinical trials. The model is currently being refined to include neurofilament levels.

Additionally, Dr. Robert Ross opened discussions to review the question of what the project aims to achieve. Some of the aims reviewed included: Clarifying use cases including documenting, describing and acknowledging issues to address explicitly; Establishing Best Models (ALS-FRS-R and Survival most relevant to Industry) to signify which model has the best suitability for different types of prediction; Objectives moving forward were to agree to share the datasets and present results and establish what different models can offer (Unified approach); Publishing the models and recommendations, potentially matrix/score card to assess/judicate the best fit model, and matching the model to the best directive to use.

Prof. Orla Hardiman mediated the final workshop on Neurofilaments and collated a body of work from expert led meetings on the lead up to the Conference. Philip Van Damme (Leuven), Dr. Lucie Bruijn (Novartis), Stephanie Fradette (Biogen) and Prof. Angela Genge (Quralis) were among the experts involved. Standardisation of technical equipment for measuring neurofilaments, automated testing and publishing a review of the current state of the art from clinical trials (state of the landscape) was identified as a key priority. Finally, proceedings concluded with a review of the potential for new studies including a structured collaborative study on neurofilaments and immune profiling identified.

About Precision ALS:

Precision ALS is a €10 million research programme involving researchers at the SFI Research Centres ADAPT and FutureNeuro along with the TRICALS Consortium, Europe’s largest ALS research initiative. Actively participating are national and international industry partners and charities including patient organisations. Further information on this research programme can be access via their website: https://www.precisionals.ie/