AI-Mind webinar on managing data available online
In an insightful online webinar held on 22nd November: “Navigating the Data Seas: Considerations for Managing Data in AI/ML Healthcare Projects,” experts from DNV shed light on the pivotal role of data in the domain of artificial intelligence (AI) and machine learning (ML) within healthcare. The event was organised by the AI-Mind consortium and featured insights from esteemed data scientists, Dr Abdillah Suyuthi and Dr Aleksander Babic.
“In any AI/ML development project, our primary goal is to deliver effective AI/ML solutions according to its use case and specification. This involves a series of key steps, including understanding the problem, collecting and comprehending data, preparing data, model training, evaluating outcomes, and deploying and monitoring solutions. Crucially, the success of AI/ML development hinges on meticulous management of the quality of processed data, transformed data, and trained models. It’s this attention to detail that propels projects towards success.“- says Dr Abdillah Suyuthi.
The webinar delved into the nuanced nature of data within AI/ML projects, emphasizing its extension beyond the conventional parameters of study outcomes. Data in this context encompasses aspects such as problem-framing details, featured data, and metadata related to machine learning models. Attendees gained comprehensive insights into each type of data and received guidance on handling it throughout the various stages of AI/ML projects, including data collection, preparation, modelling, evaluation, and deployment.
Participation in the webinar spanned diverse sectors, including biotechnology, research, health technology assessment, and artificial intelligence governance. Professionals working with an array of data types, ranging from open datasets found in scientific publications to more intricate magnetoencephalography (MEG), electroencephalography (EEG), and functional magnetic resonance imaging (fMRI) data, found the content particularly relevant. The webinar not only covered the diverse data types prevalent in ML projects but also addressed approaches for ensuring data quality when working with both raw and processed datasets.
“A thorough and very accessible presentation that gives a bird’s-eye view of data management of AI/ML projects.” This sentiment reflects the overall positive reception and accessibility of the information presented during the event.
The webinar has been made available online to facilitate wider access, encouraging those interested to watch and benefit from the wealth of knowledge shared.
It is noteworthy that the webinar builds upon the work conducted by DNV within the AI-Mind project, with detailed outcomes reported in the “AI-Mind Data Governance and Data Management Protection Framework,” which you can access at https://zenodo.org/records/7022612.
As we move forward, stay tuned for more updates from the AI-Mind project and the valuable knowledge shared by our project partners. The journey of exploration in the intersection of AI and healthcare continues, and we invite you to be a part of this ongoing dialogue.