Top Things to Know: Value Creation through Artificial Intelligence and CV Imaging
Published: January 09, 2024
Prepared by Katherine A. Sheehan, PhD.
- Although AI applications are rapidly being developed for CV imaging, their diversity in terms of imaging modality, type of patient, data features to be extracted and analyzed, and clinical application makes selecting the best one for a specific clinical situation challenging.
- This scientific statement establishes a framework for defining value in AI and CV imaging from an organizational perspective, to help identify activities where AI might create the greatest value for investment of resources.
- The statement seeks to define value for key health care stakeholders and introduce the value chain analysis approach, provide an overview of cardiac imaging AI applications, with focus on cardiac computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography, and explore how AI can enhance value creation within the CV Imaging value chain.
- Key stakeholders include clinicians, imaging professionals, hospitals, patients, AI developers, and payers. Integrating the needs of these groups is important in understanding how AI in CV imaging add value to patient care.
- Activities in the value creation analysis include decision making support, scheduling and use of imaging suites, image acquisition and analysis, reporting and communications and patient care. Supporting these primary activities are information technology, analytics and operations, market outreach, and finance.
- The four primary modalities for cardiac imaging which are ultrasound, CT, cardiac MRI (CMR), and radionuclide imaging, including single-photon emission tomography (SPECT) and positron emission tomography (PET). Each has unique needs and challenges which AI tools intend to address via improved image acquisition, reconstruction, or a tailored pixel-based analysis.
- Sustainability is a growing topic of interest among all stakeholders, given the high energy requirements for training and running AI models. Generation of the needed energy has associated greenhouse gas emissions which impact the environment, adding another factor in value consideration.
- Specific types of AI tools and their application are discussed, including natural language processing (NLP), image processing and generation (reconstruction), segmentation, classification, association and insight, decision analysis, and prediction of disease and outcomes.
- Pitfalls to the use of AI tools and potential solutions are addressed, such as bias in a data set that does not resemble the actual population being studied, and selection of appropriate statistical methods.
- AI has the potential to add value to cardiac imaging at every step of the patient’s journey. Integrating the perspectives of the various health care stakeholders is critical for creating value and ensuring the successful deployment of AI tools in the real-world clinical setting.
Citation
Hanneman K, Playford D, Dey D, van Assen M, Mastrodicasa D, Cook TS, Gichoya JW, Williamson EE, Rubin GD; on behalf of the American Heart Association Council on Cardiovascular Radiology and Intervention and Council on Lifelong Congenital Heart Disease and Heart Health in the Young. Value creation through artificial intelligence and cardiovascular imaging: a scientific statement from the American Heart Association. Circulation. Published online January 9, 2024. doi: 10.1161/CIR.0000000000001202