Conference Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 551 - Statistical Machine Learning and Artificial Intelligence in Multi-Parametric Quantitative Imaging
Type: Topic Contributed
Date/Time: Thursday, August 11, 2022 : 10:30 AM to 12:20 PM
Sponsor: Section on Statistical Computing
Abstract #323484
Title: Phenotype Classification with Multiparametric Quantitative Imaging Biomarkers
Author(s): Jana Delfino*
Companies: US Food and Drug Administration, CDRH
Keywords: phenotype classification; multiparametric classification; multi-class classification; QIBA; multi-parametric quantitative imaging biomarkers (mp-QIBs)

Multiparameter quantitative imaging offers the possibility to combine anatomical, functional, and/or behavioral biomarkers to characterize tissue, detect disease, identify phenotypes, define longitudinal change, or predict outcome with greater clinical utility than single quantitative imaging biomarkers (QIBs). RSNA’s Quantitative Imaging Biomarker Alliance (QIBA) is exploring a series of use cases for multiparameter quantitative imaging, including a multi-dimensional descriptor, phenotype classification, and risk prediction. Here we outline approach and statistical methodology for the development of a phenotype classification task from a set of multiparametric QIBs. We summarize the diagnostic accuracy and interchangeability claims that are supportable for multiclass phenotype classification. We discuss various approaches to model development and present options for statistical methodology that can be used in validation of the developed model. This work is part of the broader QIBA effort to describe common technical performance characteristics and their metrics, providing a structure for the development, estimation, and testing of multiparameter quantitative imaging.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2022 program