Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Improving the Scan-rescan Precision of AI-based CMR Biomarker Estimation

Dewmini Hasara Wickremasinghe, Yiyang Xu, Esther Puyol-Antón, Paul Aljabar, Reza Razavi, Andrew P. King

TL;DR

The paper addresses the need for consistent, repeatable biomarker estimates from cine CMR across scan-rescan sessions. It introduces two interpolation-based strategies—image-based and segmentation-based—to improve through-plane resolution and, consequently, scan-rescan precision of biomarkers such as $LVEF$, $RVEF$, and $LVM$, while maintaining accuracy. Using 184 scan-rescan acquisitions from 92 subjects across three Siemens scanners, the study shows both interpolation methods reduce Bland-Altman limits of agreement and coefficient of variation compared with baseline, with image-based interpolation often delivering the strongest gains. The findings support the potential of interpolation-driven refinement to enable robust longitudinal analyses of cardiac function in DL-based CMR biomarker pipelines.

Abstract

Quantification of cardiac biomarkers from cine cardiovascular magnetic resonance (CMR) data using deep learning (DL) methods offers many advantages, such as increased accuracy and faster analysis. However, only a few studies have focused on the scan-rescan precision of the biomarker estimates, which is important for reproducibility and longitudinal analysis. Here, we propose a cardiac biomarker estimation pipeline that not only focuses on achieving high segmentation accuracy but also on improving the scan-rescan precision of the computed biomarkers, namely left and right ventricular ejection fraction, and left ventricular myocardial mass. We evaluate two approaches to improve the apical-basal resolution of the segmentations used for estimating the biomarkers: one based on image interpolation and one based on segmentation interpolation. Using a database comprising scan-rescan cine CMR data acquired from 92 subjects, we compare the performance of these two methods against ground truth (GT) segmentations and DL segmentations obtained before interpolation (baseline). The results demonstrate that both the image-based and segmentation-based interpolation methods were able to narrow Bland-Altman scan-rescan confidence intervals for all biomarkers compared to the GT and baseline performances. Our findings highlight the importance of focusing not only on segmentation accuracy but also on the consistency of biomarkers across repeated scans, which is crucial for longitudinal analysis of cardiac function.

Improving the Scan-rescan Precision of AI-based CMR Biomarker Estimation

TL;DR

The paper addresses the need for consistent, repeatable biomarker estimates from cine CMR across scan-rescan sessions. It introduces two interpolation-based strategies—image-based and segmentation-based—to improve through-plane resolution and, consequently, scan-rescan precision of biomarkers such as , , and , while maintaining accuracy. Using 184 scan-rescan acquisitions from 92 subjects across three Siemens scanners, the study shows both interpolation methods reduce Bland-Altman limits of agreement and coefficient of variation compared with baseline, with image-based interpolation often delivering the strongest gains. The findings support the potential of interpolation-driven refinement to enable robust longitudinal analyses of cardiac function in DL-based CMR biomarker pipelines.

Abstract

Quantification of cardiac biomarkers from cine cardiovascular magnetic resonance (CMR) data using deep learning (DL) methods offers many advantages, such as increased accuracy and faster analysis. However, only a few studies have focused on the scan-rescan precision of the biomarker estimates, which is important for reproducibility and longitudinal analysis. Here, we propose a cardiac biomarker estimation pipeline that not only focuses on achieving high segmentation accuracy but also on improving the scan-rescan precision of the computed biomarkers, namely left and right ventricular ejection fraction, and left ventricular myocardial mass. We evaluate two approaches to improve the apical-basal resolution of the segmentations used for estimating the biomarkers: one based on image interpolation and one based on segmentation interpolation. Using a database comprising scan-rescan cine CMR data acquired from 92 subjects, we compare the performance of these two methods against ground truth (GT) segmentations and DL segmentations obtained before interpolation (baseline). The results demonstrate that both the image-based and segmentation-based interpolation methods were able to narrow Bland-Altman scan-rescan confidence intervals for all biomarkers compared to the GT and baseline performances. Our findings highlight the importance of focusing not only on segmentation accuracy but also on the consistency of biomarkers across repeated scans, which is crucial for longitudinal analysis of cardiac function.
Paper Structure (16 sections, 3 figures, 2 tables)

This paper contains 16 sections, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Materials and Methods
  3. Scan-rescan Dataset
  4. Overview
  5. Segmentation Models
  6. Motion Correction
  7. Interpolation
  8. Image-based interpolation:
  9. Segmentation-based interpolation:
  10. Evaluation
  11. Experiments and Results
  12. Quantitative analysis
  13. Qualitative analysis
  14. Discussion and Conclusion
  15. Acknowledgments.
  16. ...and 1 more sections

Figures (3)

  • Figure 1: Overview of the motion correction and interpolation pipelines.
  • Figure 2: Bland-Altman plots for LVEF, RVEF, and LVM between scan and rescan for human ground truth (GT, red), baseline pre-interpolation (blue), image-based interpolation (green) and segmentation-based interpolation (orange).
  • Figure 3: Visual representations of the ground truth (GT), baseline (pre-interpolation), image-based interpolation, and segmentation-based interpolation segmentations.