Scholarlink

Objective <p>Salivary gland tumors (SGTs) have complex anatomy and diverse pathological types in the head and neck. Our study evaluates the diagnostic efficacy of intravoxel incoherent motion (IVIM) and diffusion kurtosis imaging (DKI) models in differentiating benign from malignant salivary gland tumors (SGTs) and characterizing tumor subtypes.</p> Methods <p>Fifty histopathologically confirmed SGT patients undergoing multiparametric MRI (DKI and IVIM sequences) were stratified into four cohorts: malignant tumors (MTs, <i>n</i> = 18), pleomorphic adenomas (PAs, <i>n</i> = 15), Warthin tumors (WTs, <i>n</i> = 12), and other benign tumors (BTs, <i>n</i> = 5). Quantitative parameters including IVIM-derived true diffusion coefficient (D), pseudo-diffusion coefficient (D*), perfusion fraction (f), and DKI-derived mean kurtosis (MK) and mean diffusivity (MD) were systematically analyzed using the receiver operating characteristic (ROC) curves in SPSS 22.0/MedCalc13.0 (significance threshold <i>p</i> &lt; 0.05).</p> Results <p>Benign tumors exhibited significantly higher D (1.318 ± 0.467 × 10<sup>− 3</sup>mm<sup>2</sup>/s, <i>P</i> = 0.049) and MD (1.581 ± 0.577 × 10<sup>− 3</sup> mm²/s, <i>P</i> = 0.027), but lower MK (0.747 ± 0.302, <i>P</i> = 0.008) compared to malignant tumors. Among subtypes, pleomorphic adenomas (PAs) demonstrated the highest D (1.595 ± 0.427 × 10<sup>− 3</sup> mm²/s) and MD (1.923 ± 0.525 × 10<sup>− 3</sup> mm²/s), while Warthin tumors (WTs) showed the highest D* (59.134 ± 11.454 × 10<sup>− 3</sup> mm²/s, all <i>P</i> &lt; 0.001). Combining D and D* achieved an AUC of 0.967 for differentiating WTs from malignancies, whereas combining D, D*, and MD improved the AUC to 0.983. MK alone perfectly discriminated PAs from malignancies (AUC = 1.000).</p> Conclusion <p>IVIM and DKI synergistically enhance the preoperative differentiation of SGT subtypes by quantifying tumor microstructure and perfusion. Combined parameters outperform individual metrics, offering a non-invasive tool to guide clinical decision-making.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Intravoxel incoherent motion and diffusion kurtosis imaging for subtype differentiation in salivary gland tumors: a diagnostic performance study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月7日

查看全文
  • Jinru Yu,
  • Huimin Huang,
  • Jiawen Gao,
  • Han Zhang,
  • Shuo Shao,
  • Ning Zheng

摘要

Objective

Salivary gland tumors (SGTs) have complex anatomy and diverse pathological types in the head and neck. Our study evaluates the diagnostic efficacy of intravoxel incoherent motion (IVIM) and diffusion kurtosis imaging (DKI) models in differentiating benign from malignant salivary gland tumors (SGTs) and characterizing tumor subtypes.

Methods

Fifty histopathologically confirmed SGT patients undergoing multiparametric MRI (DKI and IVIM sequences) were stratified into four cohorts: malignant tumors (MTs, n = 18), pleomorphic adenomas (PAs, n = 15), Warthin tumors (WTs, n = 12), and other benign tumors (BTs, n = 5). Quantitative parameters including IVIM-derived true diffusion coefficient (D), pseudo-diffusion coefficient (D*), perfusion fraction (f), and DKI-derived mean kurtosis (MK) and mean diffusivity (MD) were systematically analyzed using the receiver operating characteristic (ROC) curves in SPSS 22.0/MedCalc13.0 (significance threshold p < 0.05).

Results

Benign tumors exhibited significantly higher D (1.318 ± 0.467 × 10− 3mm2/s, P = 0.049) and MD (1.581 ± 0.577 × 10− 3 mm²/s, P = 0.027), but lower MK (0.747 ± 0.302, P = 0.008) compared to malignant tumors. Among subtypes, pleomorphic adenomas (PAs) demonstrated the highest D (1.595 ± 0.427 × 10− 3 mm²/s) and MD (1.923 ± 0.525 × 10− 3 mm²/s), while Warthin tumors (WTs) showed the highest D* (59.134 ± 11.454 × 10− 3 mm²/s, all P < 0.001). Combining D and D* achieved an AUC of 0.967 for differentiating WTs from malignancies, whereas combining D, D*, and MD improved the AUC to 0.983. MK alone perfectly discriminated PAs from malignancies (AUC = 1.000).

Conclusion

IVIM and DKI synergistically enhance the preoperative differentiation of SGT subtypes by quantifying tumor microstructure and perfusion. Combined parameters outperform individual metrics, offering a non-invasive tool to guide clinical decision-making.