Deep learning-based imaging reconstruction for MRI after neoadjuvant chemoradiotherapy for rectal cancer: effects on image quality and assessment of treatment response

• Published in: Abdominal imaging Vol. 48; no. 1; pp. 201 - 210
• Main Authors: Kim, Bona, Lee, Chul-min, Jang, Jong Keon, Kim, Jihun, Lim, Seok-Byung, Kim, Ah Young
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.01.2023; Springer Nature B.V
• Subjects: Aged; Cancer; Chemoradiotherapy; Chemotherapy; Colorectal cancer; Deep Learning; Female; Gastroenterology; Hepatology; Hollow Organ GI; Humans; Image processing; Image quality; Image reconstruction; Imaging; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; Medical imaging; Medicine; Medicine & Public Health; Middle Aged; Neoadjuvant Therapy; Noise reduction; Quality assessment; Radiation therapy; Radiology; Rectal Neoplasms - diagnostic imaging; Rectal Neoplasms - therapy; Rectum; Regression analysis; Sensitivity and Specificity; Signal to noise ratio; Tumors; Deep learning; High resolution; Rectal cancer; Magnetic resonance imaging; Chemoradiotherapy; Complete response
• ISSN: 2366-0058; 2366-004X; 2366-0058
• DOI: 10.1007/s00261-022-03701-3

Purpose To investigate the effects of deep learning-based imaging reconstruction (DLR) on the image quality of MRI of rectal cancer after chemoradiotherapy (CRT), and its accuracy in diagnosing pathological complete responses (pCR). Methods We included 39 patients (men: women, 21:18; mean age ± standard deviation, 59.1 ± 9.7 years) with mid-to-lower rectal cancer who underwent a long-course of CRT and high-resolution rectal MRIs between January 2020 and April 2021. Axial T2WI was reconstructed using the conventional method (MRI conv ) and DLR with two different noise reduction factors (MRI DLR30 and MRI DLR50 ). The signal-to-noise ratio (SNR) of the tumor was measured. Two experienced radiologists independently made a blind assessment of the complete response on MRI. The sensitivity and specificity for pCR were analyzed using a multivariable logistic regression analysis with generalized estimating equations. Results Thirty-four patients did not have a pCR whereas five (12.8%) had pCR. Compared with the SNR of MRI conv (mean ± SD, 7.94 ± 1.92), MRI DLR30 and MRI DLR50 showed higher SNR (9.44 ± 2.31 and 11.83 ± 3.07, respectively) ( p  < 0.001). Compared to MRI conv , MRI DLR30 and MRI DLR50 showed significantly higher specificity values ( p  < 0.036) while the sensitivity values were not significantly different ( p  > 0.301). The sensitivity and specificity for pCR were 48.9% and 80.8% for MRI conv ; 48.9% and 88.2% for MRI DLR30 ; and 38.8% and 86.7% for MRI DLR50 , respectively. Conclusion DLR produced MR images with higher resolution and SNR. The specificity of MRI for identification of pCR was significantly higher with DLR than with conventional MRI. Graphical abstract