Feasibility of deep learning-reconstructed thin-slice single-breath-hold HASTE for detecting pancreatic lesions: A comparison with two conventional T2-weighted imaging sequences

Research in diagnostic and interventional imaging Pub Date : 2024-01-31 DOI:10.1016/j.redii.2023.100038

Kai Liu , Qing Li , Xingxing Wang , Caixia Fu , Haitao Sun , Caizhong Chen , Mengsu Zeng

{"title":"Feasibility of deep learning-reconstructed thin-slice single-breath-hold HASTE for detecting pancreatic lesions: A comparison with two conventional T2-weighted imaging sequences","authors":"Kai Liu , Qing Li , Xingxing Wang , Caixia Fu , Haitao Sun , Caizhong Chen , Mengsu Zeng","doi":"10.1016/j.redii.2023.100038","DOIUrl":null,"url":null,"abstract":"<div><h3>Objective</h3><p>The objective of this study was to evaluate the clinical feasibility of deep learning reconstruction-accelerated thin-slice single-breath-hold half-Fourier single-shot turbo spin echo imaging (HASTE<sub>DL</sub>) for detecting pancreatic lesions, in comparison with two conventional T2-weighted imaging sequences: compressed-sensing HASTE (HASTE<sub>CS</sub>) and BLADE.</p></div><div><h3>Methods</h3><p>From March 2022 to January 2023, a total of 63 patients with suspected pancreatic-related disease underwent the HASTE<sub>DL</sub>, HASTE<sub>CS</sub>, and BLADE sequences were enrolled in this retrospectively study. The acquisition time, the pancreatic lesion conspicuity (LC<sub>P</sub>), respiratory motion artifact (RMA), main pancreatic duct conspicuity (MPDC), overall image quality (OIQ), signal-to-noise ratio (SNR), and contrast-noise-ratio (CNR) of the pancreatic lesions were compared among the three sequences by two readers.</p></div><div><h3>Results</h3><p>The acquisition time of both HASTE<sub>DL</sub> and HASTE<sub>CS</sub> was 16 s, which was significantly shorter than that of 102 s for BLADE. In terms of qualitative parameters, Reader 1 and Reader 2 assigned significantly higher scores to the LC<sub>P</sub>, RMA, MPDC, and OIQ for HASTE<sub>DL</sub> compared to HASTE<sub>CS</sub> and BLADE sequences; As for the quantitative parameters, the SNR values of the pancreatic head, body, tail, and lesions, the CNR of the pancreatic lesion measured by the two readers were also significantly higher for HASTE<sub>DL</sub> than for HASTE<sub>CS</sub> and BLADE sequences.</p></div><div><h3>Conclusions</h3><p>Compared to conventional T2WI sequences (HASTE<sub>CS</sub> and BLADE), deep-learning reconstructed HASTE enables thin slice and single-breath-hold acquisition with clinical acceptable image quality for detection of pancreatic lesions.</p></div>","PeriodicalId":74676,"journal":{"name":"Research in diagnostic and interventional imaging","volume":"9 ","pages":"Article 100038"},"PeriodicalIF":0.0000,"publicationDate":"2024-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2772652523000170/pdfft?md5=721bb9a7136e8ca36e5bc3782b2f206e&pid=1-s2.0-S2772652523000170-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Research in diagnostic and interventional imaging","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2772652523000170","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

Abstract

Objective

The objective of this study was to evaluate the clinical feasibility of deep learning reconstruction-accelerated thin-slice single-breath-hold half-Fourier single-shot turbo spin echo imaging (HASTE_DL) for detecting pancreatic lesions, in comparison with two conventional T2-weighted imaging sequences: compressed-sensing HASTE (HASTE_CS) and BLADE.

Methods

From March 2022 to January 2023, a total of 63 patients with suspected pancreatic-related disease underwent the HASTE_DL, HASTE_CS, and BLADE sequences were enrolled in this retrospectively study. The acquisition time, the pancreatic lesion conspicuity (LC_P), respiratory motion artifact (RMA), main pancreatic duct conspicuity (MPDC), overall image quality (OIQ), signal-to-noise ratio (SNR), and contrast-noise-ratio (CNR) of the pancreatic lesions were compared among the three sequences by two readers.

Results

The acquisition time of both HASTE_DL and HASTE_CS was 16 s, which was significantly shorter than that of 102 s for BLADE. In terms of qualitative parameters, Reader 1 and Reader 2 assigned significantly higher scores to the LC_P, RMA, MPDC, and OIQ for HASTE_DL compared to HASTE_CS and BLADE sequences; As for the quantitative parameters, the SNR values of the pancreatic head, body, tail, and lesions, the CNR of the pancreatic lesion measured by the two readers were also significantly higher for HASTE_DL than for HASTE_CS and BLADE sequences.

Conclusions

Compared to conventional T2WI sequences (HASTE_CS and BLADE), deep-learning reconstructed HASTE enables thin slice and single-breath-hold acquisition with clinical acceptable image quality for detection of pancreatic lesions.

查看原文本刊更多论文

深度学习-重建薄片单次呼吸HASTE检测胰腺病变的可行性：与两种传统T2加权成像序列的比较

目的本研究旨在评估深度学习重建-加速薄层单呼吸半傅里叶单次涡轮自旋回波成像（HASTEDL）与两种常规T2加权成像序列：压缩传感HASTE（HASTECS）和BLADE相比，在检测胰腺病变方面的临床可行性。方法从2022年3月至2023年1月，共有63名疑似胰腺相关疾病患者接受了HASTEDL、HASTECS和BLADE序列的回顾性研究。两位读者比较了三种序列的采集时间、胰腺病变的清晰度（LCP）、呼吸运动伪影（RMA）、主胰管清晰度（MPDC）、总体图像质量（OIQ）、信噪比（SNR）和对比度-噪声比（CNR）。在定性参数方面，读者1和读者2对HASTEDL的LCP、RMA、MPDC和OIQ评分均明显高于HASTECS和BLADE序列；在定量参数方面，两位读者测量的胰头、胰体、胰尾和病灶的信噪比值以及胰腺病灶的CNR均明显高于HASTECS和BLADE序列。结论与传统的 T2WI 序列（HASTECS 和 BLADE）相比，深度学习重建的 HASTEDL 可进行薄切片和单次呼吸采集，其图像质量可为临床接受，可用于检测胰腺病变。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Research in diagnostic and interventional imaging Radiology and Imaging

自引率

0.00%

发文量