A framework to quantify and incorporate uncertainty for ethical application of AI-based quantitative imaging in clinical decision making

项目概述：定量成像（QI），从患者中计算数值/统计特征

Project Summary: Quantitative imaging (QI), where a numerical/statistical feature is computed from a patient image, is emerging as an important tool for diagnosis and therapy planning. Artificial intelligence (AI)-based QI tools are showing significant promise in this area. However, the measured quantitative value from these tools may also suffer from uncertainty due to various reasons such as limited training data, inaccurate ground truth, mismatch between training and test sets. For ethical application of AI-based QI tools, this uncertainty should be quantified and then incorporated in the clinical decision-making process. This is necessary for the ethical application of these tools, an inference that also emerged from a survey conducted by us across patient advocates (Birch et al, Nature Medicine 2022). Towards addressing this goal, in this proposal, we first propose to develop a novel no-gold-standard method to quantify uncertainty of AI-based QI tools using patient data. Existing uncertainty quantification techniques have mainly been developed for detection tasks, and typically require availability of gold standard. In contrast, the proposed technique will be developed for quantification tasks and not require any gold standard quantitative value. Next, to incorporate the uncertainty of the AI-based QI tool, we propose to propose to develop a questionnaire that will elicit the patient’s risk-value profiles towards treatments. For example, if an AI-based QI tool outputs a quantitative value that indicates aggressive therapy, but with high uncertainty, some patients may be risk averse and prefer to assign high weight to the uncertainty value, while other patients may value the benefits of the treatment and thus assign less weight to that uncertainty. To incorporate these patient preferences, we propose to develop a questionnaire that will elicit the patient’s risk- value profiles. This project will advance on the ongoing activities of our current R01 award on no-gold-standard evaluation of QI methods, extending that project in the context of uncertainty quantification, and thus enabling the use of our tools for not just evaluation, but generating personalized recommendations for each patient. The methods will be developed in the context of the highly significant clinical question of guiding therapy response in patients with stage III non-small cell lung cancer (NSCLC). Answering this question will help address a critical, urgent, and unmet need for strategies to personalize the treatment of NSCLC, a disease with high morbidity and mortality rates. A highly multi-disciplinary team consisting of imaging scientist with expertise in AI, AI ethicists, oncologist, and nuclear-medicine physician have been assembled for this study. This supplement is directly responsive to NOT-OD-22-065 in terms of developing a framework for ethical clinical use of AI. The project will also strengthen the impact of tools we are developing in the parent R01 by using them to guide clinical decision making. Impact will also be strengthened by collaboration with patient advocacy groups and industry partners. Overall, this project is poised to strongly impact the ethical clinical application of QI for treatment of NSCLC, as well as other cancers, cardiac and neurodegenerative diseases where QI has a role.