Objective To assess whether a force-controlled robotic arm can improve image quality, consistency, and diagnostic accuracy of strain elastography (SE) and shear wave elastography (SWE) in differentiating benign from malignant thyroid nodules. Materials and methods In this prospective study, 131 thyroid nodules were examined by a junior physician, a senior physician, and a robotic arm with a PID-based force feedback system. Image quality was evaluated using the structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), mean squared error (MSE) and mean opinion score (MOS). Consistency was assessed with the Dice similarity coefficient (DSC) and the intraclass correlation coefficient (ICC) for SWE-derived Emax. Diagnostic performance was analyzed via ROC curves and AUC comparisons. Results The robotic arm achieved higher image quality (SSIM up to 0.94, PSNR 42.25 dB, MSE 3.06) and MOS scores (SE: 4.1 ± 0.3; SWE: 4.3 ± 0.2) than both human operators (all p < 0.001). It also showed better consistency (DSC up to 0.90; ICC up to 0.94) and diagnostic accuracy (AUC 0.90 for SE, 0.95 for SWE; p < 0.05). Conclusion The force-controlled robotic arm provides standardized, reproducible thyroid elastography with superior quality, consistency, and accuracy compared with manual scanning. Critical relevance statement This study explores robotic-assisted elastography to improve consistency in thyroid nodule assessment. Key Points <UnorderedList Mark="Bullet"> <ItemContent> The force-controlled robotic arm enhanced ultrasound elastography by generating clearer, more standardized images compared with manual acquisition. </ItemContent> <ItemContent> The force-controlled robotic arm scanning achieved superior intra-operator reproducibility. </ItemContent> <ItemContent> Diagnostic performance in differentiating benign from malignant thyroid nodules was significantly improved with robotic elastography, exceeding that of both junior and senior radiologists. </ItemContent> </UnorderedList> Graphical Abstract

Force-controlled robotic ultrasound elastography enhances diagnostic consistency for thyroid nodules

期刊论文
出版模式：

开放获取
发表日期：

2026年4月16日

查看全文

Tianhui Yan,
Xu Le,
Li Xie,
Xiangyu Qiu,
Delei Cheng,
Chong Pei,
Lei Hu

摘要

Objective

To assess whether a force-controlled robotic arm can improve image quality, consistency, and diagnostic accuracy of strain elastography (SE) and shear wave elastography (SWE) in differentiating benign from malignant thyroid nodules.

Materials and methods

In this prospective study, 131 thyroid nodules were examined by a junior physician, a senior physician, and a robotic arm with a PID-based force feedback system. Image quality was evaluated using the structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), mean squared error (MSE) and mean opinion score (MOS). Consistency was assessed with the Dice similarity coefficient (DSC) and the intraclass correlation coefficient (ICC) for SWE-derived Emax. Diagnostic performance was analyzed via ROC curves and AUC comparisons.

Results

The robotic arm achieved higher image quality (SSIM up to 0.94, PSNR 42.25 dB, MSE 3.06) and MOS scores (SE: 4.1 ± 0.3; SWE: 4.3 ± 0.2) than both human operators (all p < 0.001). It also showed better consistency (DSC up to 0.90; ICC up to 0.94) and diagnostic accuracy (AUC 0.90 for SE, 0.95 for SWE; p < 0.05).

Conclusion

The force-controlled robotic arm provides standardized, reproducible thyroid elastography with superior quality, consistency, and accuracy compared with manual scanning.

Critical relevance statement

This study explores robotic-assisted elastography to improve consistency in thyroid nodule assessment.

Key Points

The force-controlled robotic arm enhanced ultrasound elastography by generating clearer, more standardized images compared with manual acquisition.

The force-controlled robotic arm scanning achieved superior intra-operator reproducibility.

Diagnostic performance in differentiating benign from malignant thyroid nodules was significantly improved with robotic elastography, exceeding that of both junior and senior radiologists.

Graphical Abstract