文献速递腹部技术与安全CT

深度学习图像重建的减剂量40 keV虚拟单能成像用于结直肠癌肝转移早期检出

Deep learning imaging reconstruction of reduced-dose 40 keV virtual monoenergetic imaging for early detection of colorectal cancer liver metastases.
PubMed 文献 · Eur J Radiol · 2023-09-29

前瞻性纳入35例病理确诊结直肠癌患者,行标准剂量与减剂量宝石能谱CT门静脉期扫描,比较40 keV虚拟单能图像经FBP、ASIR-V 60%及减剂量DLIR-M、DLIR-H重建后的图像质量与病灶检出。减剂量DLIR-M图像质量及病灶诊断信心高于标准剂量FBP,≤5 mm病灶检出率与诊断敏感度未见明显降低。

要点

  • 减剂量DLIR-M的40 keV虚拟单能图像可用于结直肠癌患者常规随访评估。
  • 减剂量DLIR-M对小病灶及早期肝转移的检出率与诊断敏感度未见明显降低。
  • DLIR-M主观图像质量及病灶诊断信心高于标准剂量FBP(均P<0.01)。

核心结论

  • 减剂量DLIR-M的40 keV虚拟单能图像可用于结直肠癌患者常规随访,优化评估并保证CT图像质量,小病灶及早期肝转移的检出率与诊断敏感度和特异度未见明显降低。

研究设计与样本

  • 前瞻性研究,2022年3月至8月纳入35例病理确诊结直肠癌患者,行标准剂量(CTDIvol=15.51 mGy)与减剂量(CTDIvol=7.95 mGy)宝石能谱CT门静脉期扫描,40 keV虚拟单能图像分别经FBP、ASIR-V 60%、减剂量DLIR-M及DLIR-H重建。

关键影像表现/征象

  • 共识别219个病灶,包括55个良性病灶和164个转移灶,平均大小7.37±4.14 mm。
  • 标准剂量FBP检出207个病灶,标准剂量AV60检出201个,DLIR-M检出199个,DLIR-H检出190个。
  • DLIR-M与DLIR-H的40 keV虚拟单能图像CNR、SNR及噪声优于标准剂量FBP(P<0.01),但与标准剂量AV60无显著差异(P>0.05)。
  • DLIR-M主观图像质量及病灶诊断信心高于标准剂量FBP(均P<0.01)。

诊断标准与关键数值

  • 病灶≤5 mm时,标准剂量FBP与DLIR-M检出率无统计学差异(χ2McNemar=1.00,P=0.32),标准剂量AV60与DLIR-M亦无统计学差异(χ2McNemar=0.33,P=0.56)。
  • 病灶≤5 mm时,标准剂量FBP、标准剂量AV60、DLIR-M与DLIR-H的总体诊断敏感度存在统计学差异(P<0.01)。
  • 标准剂量FBP、标准剂量AV60与DLIR-M之间病灶诊断敏感度无统计学差异(均P>0.05)。

临床应用与阅片要点

  • 减剂量DLIR-M的40 keV虚拟单能图像可用于结直肠癌患者常规随访评估。
  • 阅片时DLIR-M图像在主观图像质量及病灶诊断信心方面优于标准剂量FBP。

局限性

  • 减剂量DLIR-M对≤5 mm小病灶及早期肝转移的检出率与诊断敏感度和特异度未见明显降低。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

OBJECTIVE: To explore whether reduced-dose (RD) gemstone spectral imaging (GSI) and deep learning image reconstruction (DLIR) of 40 keV virtual monoenergetic image (VMI) enhanced the early detection and diagnosis of colorectal cancer liver metastases (CRLM).

METHODS: Thirty-five participants with pathologically confirmed colorectal cancer were prospectively enrolled from March to August 2022 after routine care abdominal computed tomography (CT). GSI mode was used for contrast-enhanced CT, and two portal venous phase CT images were obtained [standard-dose (SD) CT dose index (CTDIvol) = 15.51 mGy, RD CTDIvol = 7.95 mGy]. The 40 keV-VMI were reconstructed via filtered back projection (FBP) and iterative reconstruction (ASIR-V 60 %, AV60) of both SD and RD images. RD medium-strength deep learning image reconstruction (DLIR-M) and RD high-strength deep learning image reconstruction (DLIR-H) were used to reconstruct the 40 keV-VMI. The contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) of the liver and the lesions were objectively evaluated. The overall image quality, lesion conspicuity, and diagnostic confidence were subjectively evaluated, to compare the differences in evaluation results among the different images.

RESULTS: All 35 participants (mean age: 59.51 ± 11.01 years; 14 females) underwent SD and RD GSI portal venous-phase CT scans. The dose-length product of the RD GSI scan was reduced by 49-53 % lower than that of the SD GSI scan (420.22 ± 31.95) vs (817.58 ± 60.56). A total of 219 lesions were identified, including 55 benign lesions and 164 metastases, with an average size of 7.37 ± 4.14 mm. SD-FBP detected 207 lesions, SD-AV60 detected 201 lesions, and DLIR-M and DLIR-H detected 199 and 190 lesions, respectively. For lesions ≤ 5 mm, there was no statistical difference between SD-FBP vs DLIR-M (χ2McNemar = 1.00, P = 0.32) and SD-AV60 vs DLIR-M (χ2McNemar = 0.33, P = 0.56) in the detection rate. The CNR, SNR, and noise of DLIR-M and DLIR-H 40 keV-VMI images were better than those of SD-FBP images (P < 0.01) but did not differ significantly from those of SD-AV60 images (P > 0.05). When the lesions ≤ 5 mm, there were statistical differences in the overall diagnostic sensitivity of lesions compared with SD-FBP, SD-AV60, DLIR-M and DLIR-H (P<0.01). There were no statistical differences in the sensitivity of lesions diagnosis between SD-FBP, SD-AV60 and DLIR-M (both P>0.05). However, the DLIR-M subjective image quality and lesion diagnostic confidence were higher for SD-FBP (both P < 0.01).

CONCLUSION: Reduced dose DLIR-M of 40 keV-VMI can be used for routine follow-up care of colorectal cancer patients, to optimize evaluations and ensure CT image quality. Meanwhile, the detection rate and diagnostic sensitivity and specificity of small lesions, early liver metastases is not obviously reduced.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Li S, Yuan L, Lu T, et al. Deep learning imaging reconstruction of reduced-dose 40 keV virtual monoenergetic imaging for early detection of colorectal cancer liver metastases. Eur J Radiol. 2023. PMID: 37816301. doi:10.1016/j.ejrad.2023.111128

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。