Abstract A54: Case-based causality: An application of artificial intelligence to environmental carcinogenesis

Abstract

Abstract Determining whether environmental exposures cause cancer is a complex and often controversial issue, involving epidemiologic and toxicologic studies and a family of methods discussed and debated for decades. These methods include the general scientific method, study design and statistical methods, and research synthesis methods—sometimes called “weight of evidence” methods—such as the systematic narrative review, meta-analysis, and criteria-based methods of causation. These methods have been applied to a wide range of topics involving environmental (and occupational and lifestyle) exposures as well as a wide variety of specific cancers. The purpose of this project is to bring a method from artificial intelligence (AI)—Case-Based Reasoning (CBR)—to bear on the issue of environmental causes of cancer. Case-Based Reasoning is an AI method that capitalizes upon past experience to solve current problems. In terms of the search for causes of disease, CBR (renamed Case-Based Causality) is a method that uses bodies of evidence from past examples of known causal relationships to inform the determination of causality for a body of evidence associated with an exposure-disease relationship under investigation. CBR is typically described at its conceptual level in terms of 5 “Rs”: Representation, Retrieval, Reuse, Revision, and Retention. Each “R” can be interpreted in terms of the causation issue in epidemiology and public health. “Representation,” for example, is the collection, description, summarization, and interpretation of a body of evidence regarding an exposure and a disease—what is basically provided by a high-quality systematic narrative review, which may include a meta-analysis. In this presentation, the 5 “Rs” of Case-Based Causality (CBC) will be described and applied to current issues in environmental carcinogenesis. In addition, the relationship of CBC to existing methods of causal inference will be noted as well as the links between CBC and the concept of reliability, i.e., the extent to which current methods of causal inference provide reproducible results. Fundamentally, Case-Based Causality is a method for examining whether a body of evidence under investigation can be considered causal by examining the extent to which its characteristics are similar to bodies of evidence from known (i.e., established) causal relationships. Citation Format: Douglas L. Weed. Case-based causality: An application of artificial intelligence to environmental carcinogenesis [abstract]. In: Proceedings of the AACR Special Conference on Environmental Carcinogenesis: Potential Pathway to Cancer Prevention; 2019 Jun 22-24; Charlotte, NC. Philadelphia (PA): AACR; Can Prev Res 2020;13(7 Suppl): Abstract nr A54.