Promise and challenges of heart chamber segmentation from non-contrast CT scans using contrastive unpaired image translation: a feasibility study

2026-06-22

AI Summary

This study addresses the challenge of cardiac chamber segmentation in non-contrast CT scans, where the scarcity of annotated data hinders supervised learning. To overcome this limitation, the authors propose a self-supervised approach based on contrastive unpaired image translation. By enhancing the CUT framework with a disentangled contrastive learning loss, the method generates high-quality synthetic non-contrast CT images with accurate anatomical labels. These synthetic images are then used to train an nnU-Net architecture optimized with Hausdorff distance–based losses, enabling precise automatic segmentation of the left and right atria and ventricles without requiring real non-contrast annotations. Experimental results demonstrate strong performance: Dice scores of 0.91–0.94 and HD95 distances of 3.63–5.74 mm on synthetic data, and high correlation (Pearson’s r = 0.82–0.93) with low mean absolute percentage errors (9.22%–20.79%) in chamber volume estimation on real non-contrast CT scans, confirming the method’s efficacy and clinical potential.

Abstract

Purpose: To evaluate the feasibility and challenges of heart chamber segmentation from non-contrast CT scans using contrastive unpaired image translation and deep learning-based segmentation. Approach: We developed ChameleonNet, a framework utilizing the Contrastive Unpaired Translation (CUT) network with decoupled contrastive learning (DCL) loss to synthesize non-contrast CT from contrast CT scans. Using annotations of four heart chambers (left atrium (LA), left ventricle (LV), right atrium (RA), and right ventricle (RV)) from contrast scans, we trained a Hausdorff distance loss-enhanced nnU-Net on synthesized non-contrast images. The translation model was trained with 35,538 contrast-enhanced and 37,197 non-contrast CT slices. The segmentation model was trained with 292 synthesized non-contrast scans. Performance was evaluated using Dice similarity coefficient (DSC) and 95th Hausdorff distance (HD95) on 36 synthesized non-contrast scans, and volume agreement on 36 real non-contrast CT scans was assessed using Pearson correlation, mean absolute percentage error (MAPE), and mean percentage error (MPE). Results: The segmentation model achieved DSC of 0.94 (0.01), 0.91 (0.04), 0.92 (0.03), 0.93 (0.02), and HD95 of 3.63 (1.49), 5.74 (4.08), 5.18 (1.77), 5.51 (3.21) mm on synthesized non-contrast images for LA, LV, RA, and RV, respectively. On real non-contrast CT scans, Pearson correlations were 0.93, 0.82, 0.87, and 0.89 (all p<0.001), with MAPE ranging from 9.22% to 20.79%, and MPE ranging from -12.52% to 4.67%. Conclusions: ChameleonNet demonstrated feasibility for heart chamber segmentation from non-contrast CT without manual non-contrast annotations. However, volume errors, particularly for LV and RV, indicate that further refinement and validation are needed before clinical use.

Problem

• heart chamber segmentation
• non-contrast CT
• contrastive unpaired image translation
• deep learning
• medical image analysis

Innovation

• contrastive unpaired image translation
• decoupled contrastive learning
• heart chamber segmentation
• synthetic non-contrast CT
• nnU-Net with Hausdorff loss