Deep Learning-Based Reconstruction Improves the Image Quality of Low-Dose CT Colonography

Yanshan Chen; Zixuan Huang; Lijuan Feng; Wenbin Zou; Decan Kong; Dongyun Zhu; Guochao Dai; Weidong Zhao; Yuanke Zhang; Mingyue Luo

doi:10.1016/j.acra.2024.01.021

Deep Learning-Based Reconstruction Improves the Image Quality of Low-Dose CT Colonography

Acad Radiol. 2024 Jan 29:S1076-6332(24)00021-7. doi: 10.1016/j.acra.2024.01.021. Online ahead of print.

Authors

Yanshan Chen¹, Zixuan Huang², Lijuan Feng³, Wenbin Zou⁴, Decan Kong⁴, Dongyun Zhu⁴, Guochao Dai⁵, Weidong Zhao⁶, Yuanke Zhang⁷, Mingyue Luo⁸

Affiliations

¹ Department of Radiology, the Six Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.); Biomedical Innovation Center, the Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.); Department of Radiology, Jinling Hospital, the First School of Clinical Medicine, Southern Medical University, Nanjing, Jiangsu 210002, China (Y.C.).
² Department of Radiology, the Six Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.); Biomedical Innovation Center, the Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.); Department of Radiology, Guangdong Second Traditional Chinese Medicine Hospital, Guangzhou, Guangdong 510095, China (Z.H.).
³ Department of Radiology, the Six Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.); Biomedical Innovation Center, the Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.); Department of Radiology, the First Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi 530021, China (L.F.).
⁴ Department of Radiology, the Six Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.); Biomedical Innovation Center, the Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.).
⁵ Medical Imaging Center, the First People's Hospital of Kashi Area, Kashi, Xinjiang 844000, China (G.D.).
⁶ Department of Radiology, the Second Hospital of Shanxi Medical University, Taiyuan, Shanxi 030001, China (W.Z.).
⁷ School of Computer Science, Qufu Normal University, Rizhao, Shandong 276826, China (Y.Z.).
⁸ Department of Radiology, the Six Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.); Biomedical Innovation Center, the Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong 510655, China (Y.C., Z.H., L.F., W.Z., D.K., D.Z., M.L.). Electronic address: luomingy@mail.sysu.edu.cn.

PMID: 38290889
DOI: 10.1016/j.acra.2024.01.021

Abstract

Rationale and objectives: To evaluate the image quality of low-dose CT colonography (CTC) using deep learning-based reconstruction (DLR) compared to iterative reconstruction (IR).

Materials and methods: Adults included in the study were divided into four groups according to body mass index (BMI). Routine-dose (RD: 120 kVp) CTC images were reconstructed with IR (RD-IR); low-dose (LD: 100kVp) images were reconstructed with IR (LD-IR) and DLR (LD-DLR). The subjective image quality was rated on a 5-point scale by two radiologists independently. The parameters for objective image quality included noise, signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). The Friedman test was used to compare the image quality among RD-IR, LD-IR and LD-DLR. The KruskalWallis test was used to compare the results among different BMI groups.

Results: A total of 270 volunteers (mean age: 47.94 years ± 11.57; 115 men) were included. The effective dose of low-dose CTC was decreased by approximately 83.18% (5.18mSv ± 0.86 vs. 0.86mSv ± 0.05, P < 0.001). The subjective image quality score of LD-DLR was superior to that of LD-IR (3.61 ± 0.56 vs. 2.70 ± 0.51, P < 0.001) and on par with the RD- IR's (3.61 ± 0.56 vs. 3.74 ± 0.52, P = 0.486). LD-DLR exhibited the lowest noise, and the maximum SNR and CNR compared to RD-IR and LD-IR (all P < 0.001). No statistical difference was found in the noise of LD-DLR images between different BMI groups (all P > 0.05).

Conclusion: Compared to IR, DLR provided low-dose CTC with superior image quality at an average radiation dose of 0.86mSv, which may be promising in future colorectal cancer screening.

Keywords: Colonography; Computed Tomographic; Deep learning; Radiation protection.