Thin-slice T<sub>2</sub>-weighted images and deep-learning-based super-resolution reconstruction: improved preoperative assessment of vascular invasion for pancreatic ductal adenocarcinoma.
Authors
Affiliations (4)
Affiliations (4)
- Department of Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, P.R. China.
- Department of Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, P.R. China. [email protected].
- Department of Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, P.R. China. [email protected].
- Department of Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, P.R. China. [email protected].
Abstract
To evaluate the efficacy of thin-slice T<sub>2</sub>-weighted imaging (T<sub>2</sub>WI) and super-resolution reconstruction (SRR) for preoperative assessment of vascular invasion in pancreatic ductal adenocarcinoma (PDAC). Ninety-five PDACs with preoperative MRI were retrospectively enrolled as a training set, with non-reconstructed T<sub>2</sub>WI (NRT<sub>2</sub>) in different slice thicknesses (NRT<sub>2</sub>-3, 3 mm; NRT<sub>2</sub>-5, ≥ 5 mm). A prospective test set was collected with NRT<sub>2</sub>-5 (n = 125) only. A deep-learning network was employed to generate reconstructed super-resolution T<sub>2</sub>WI (SRT<sub>2</sub>) in different slice thicknesses (SRT<sub>2</sub>-3, 3 mm; SRT<sub>2</sub>-5, ≥ 5 mm). Image quality was assessed, including the signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and signal-intensity ratio (SIR<sub>t/p</sub>, tumor/pancreas; SIR<sub>t/b</sub>, tumor/background). Diagnostic efficacy for vascular invasion was evaluated using the area under the curve (AUC) and compared across different slice thicknesses before and after reconstruction. SRT<sub>2</sub>-5 demonstrated higher SNR and SIR<sub>t/p</sub> compared to NRT<sub>2</sub>-5 (74.18 vs 72.46; 1.42 vs 1.30; p < 0.05). SRT<sub>2</sub>-3 showed increased SIR<sub>t/p</sub> and SIR<sub>t/b</sub> over NRT<sub>2</sub>-3 (1.35 vs 1.31; 2.73 vs 2.58; p < 0.05). SRT<sub>2</sub>-5 showed higher CNR, SIR<sub>t/p</sub> and SIR<sub>t/b</sub> than NRT<sub>2</sub>-3 (p < 0.05). NRT<sub>2</sub>-3 outperformed NRT<sub>2</sub>-5 in evaluating venous invasion (AUC: 0.732 vs 0.597, p = 0.021). SRR improved venous assessment (AUC: NRT<sub>2</sub>-3, 0.927 vs 0.732; NRT<sub>2</sub>-5, 0.823 vs 0.597; p < 0.05), and SRT<sub>2</sub>-5 exhibits comparable efficacy to NRT<sub>2</sub>-3 in venous assessment (AUC: 0.823 vs 0.732, p = 0.162). Thin-slice T<sub>2</sub>WI and SRR effectively improve the image quality and diagnostic efficacy for assessing venous invasion in PDAC. Thick-slice T<sub>2</sub>WI with SRR is a potential alternative to thin-slice T<sub>2</sub>WI. Both thin-slice T<sub>2</sub>-WI and SRR effectively improve image quality and diagnostic performance, providing valuable options for optimizing preoperative vascular assessment in PDAC. Non-invasive and accurate assessment of vascular invasion supports treatment planning and avoids futile surgery. Vascular invasion evaluation is critical for the surgical eligibility of PDAC. SRR improved image quality and vascular assessment in T<sub>2</sub>WI. Utilizing thin-slice T<sub>2</sub>WI and SRR aids in clinical decision making for PDAC.