Scholarlink

Purpose <p>This study aimed to evaluate whether adding diffusion-weighted imaging (DWI) to dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) improves the differentiation between nonmalignant and malignant MRI-detected lesions.</p> Materials and methods <p>A total of 36 patients with 40 Breast Imaging-Reporting and Data System (BI-RADS) category 4 MRI-detected lesions were analyzed. Two readers independently evaluated lesion visibility on two <i>b</i>-value DWIs (<i>b</i> = 800&#xa0;s/mm<sup>2</sup>, DWI<sub>800</sub> and <i>b</i> = 1500&#xa0;s/mm<sup>2</sup>, DWI<sub>1500</sub>) using a five-point scale (from 0 = poor to 4 = excellent) and measured the apparent diffusion coefficient (ADC) values. DWI visibility scores and ADC values were compared between nonmalignant and malignant lesions. For the diagnostic performance analysis, lesions were classified as DWI-positive (scores 2–4, suggestive of malignancy) or DWI-negative (scores 0–1, suggestive of benignity).</p> Results <p>The inter-observer agreement of the DWI visibility score was substantial for both DWI<sub>800</sub> (κ = 0.923) and DWI<sub>1500</sub> (κ = 0.864). The lesion visibility score was higher on DWI<sub>1500</sub> than on DWI<sub>800</sub>. The score was significantly higher for malignant than for nonmalignant MRI-detected lesions on both DWI<sub>800</sub> (reader 1: 2.74 vs. 1.57, p = 0.012; reader 2: 2.95 vs. 1.81, p = 0.020) and DWI<sub>1500</sub> (reader 1: 3.11 vs 2.24, p = 0.048; reader 2: 3.00 vs. 1.95, p = 0.038). For mass lesions, both DWI<sub>800</sub> and DWI<sub>1500</sub> achieved 100% sensitivity and a negative predictive value for both readers, whereas all DWI-negative malignant lesions demonstrated non-mass enhancement. Overall, the mean and minimum ADC values did not differ significantly between nonmalignant and malignant lesions.</p> Conclusion <p>The present findings suggest that DWI visibility provides additional value for distinguishing between nonmalignant and malignant BI-RADS category 4 MRI-detected breast lesions. Particularly for mass lesions, DWI-negative MRI-detected lesions were associated with nonmalignant pathology, indicating a potential role for DWI in reducing unnecessary biopsies.</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

Additional diagnostic value of lesion visibility on diffusion-weighted imaging for differentiating BI-RADS category 4 MRI-detected breast lesions

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月10日

查看全文
  • Saeko Usui,
  • Mariko Goto,
  • Koji Sakai,
  • Toshiyuki Okamoto,
  • Yuka Onishi,
  • Kei Yamada

摘要

Purpose

This study aimed to evaluate whether adding diffusion-weighted imaging (DWI) to dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) improves the differentiation between nonmalignant and malignant MRI-detected lesions.

Materials and methods

A total of 36 patients with 40 Breast Imaging-Reporting and Data System (BI-RADS) category 4 MRI-detected lesions were analyzed. Two readers independently evaluated lesion visibility on two b-value DWIs (b = 800 s/mm2, DWI800 and b = 1500 s/mm2, DWI1500) using a five-point scale (from 0 = poor to 4 = excellent) and measured the apparent diffusion coefficient (ADC) values. DWI visibility scores and ADC values were compared between nonmalignant and malignant lesions. For the diagnostic performance analysis, lesions were classified as DWI-positive (scores 2–4, suggestive of malignancy) or DWI-negative (scores 0–1, suggestive of benignity).

Results

The inter-observer agreement of the DWI visibility score was substantial for both DWI800 (κ = 0.923) and DWI1500 (κ = 0.864). The lesion visibility score was higher on DWI1500 than on DWI800. The score was significantly higher for malignant than for nonmalignant MRI-detected lesions on both DWI800 (reader 1: 2.74 vs. 1.57, p = 0.012; reader 2: 2.95 vs. 1.81, p = 0.020) and DWI1500 (reader 1: 3.11 vs 2.24, p = 0.048; reader 2: 3.00 vs. 1.95, p = 0.038). For mass lesions, both DWI800 and DWI1500 achieved 100% sensitivity and a negative predictive value for both readers, whereas all DWI-negative malignant lesions demonstrated non-mass enhancement. Overall, the mean and minimum ADC values did not differ significantly between nonmalignant and malignant lesions.

Conclusion

The present findings suggest that DWI visibility provides additional value for distinguishing between nonmalignant and malignant BI-RADS category 4 MRI-detected breast lesions. Particularly for mass lesions, DWI-negative MRI-detected lesions were associated with nonmalignant pathology, indicating a potential role for DWI in reducing unnecessary biopsies.