PT - JOURNAL ARTICLE AU - Fang, Qiang AU - Yang, Qing AU - Wang, Bao AU - Wen, Bing AU - Xu, Guangrun AU - He, Jingzhen TI - Enhancing Lesion Detection in Inflammatory Myelopathies: A Deep Learning–Reconstructed Double Inversion Recovery MRI Approach AID - 10.3174/ajnr.A8582 DP - 2025 May 01 TA - American Journal of Neuroradiology 4099 - http://www.ajnr.org/content/early/2025/04/30/ajnr.A8582.short 4100 - http://www.ajnr.org/content/early/2025/04/30/ajnr.A8582.full AB - BACKGROUND AND PURPOSE: The imaging of inflammatory myelopathies has advanced significantly across time, with MRI techniques playing a pivotal role in enhancing lesion detection. However, the impact of deep learning (DL)-based reconstruction on 3D double inversion recovery (DIR) imaging for inflammatory myelopathies remains unassessed. This study aimed to compare the acquisition time, image quality, diagnostic confidence, and lesion detection rates among sagittal T2WI, standard DIR, and DL-reconstructed DIR in patients with inflammatory myelopathies.MATERIALS AND METHODS: In this observational study, patients diagnosed with inflammatory myelopathies were recruited between June 2023 and March 2024. Each patient underwent sagittal conventional TSE sequences and standard 3D DIR (T2WI and standard 3D DIR were used as references for comparison), followed by an undersampled accelerated double inversion recovery deep learning (DIRDL) examination. Three neuroradiologists evaluated the images using a 4-point Likert scale (from 1 to 4) for overall image quality, perceived SNR, sharpness, artifacts, and diagnostic confidence. The acquisition times and lesion detection rates were also compared among the acquisition protocols.RESULTS: A total of 149 participants were evaluated (mean age, 40.6 [SD, 16.8] years; 71 women). The median acquisition time for DIRDL was significantly lower than for standard DIR (298 seconds [interquartile range, 288–301 seconds] versus 151 seconds [interquartile range, 148–155 seconds]; P < .001), showing a 49% time reduction. DIRDL images scored higher in overall quality, perceived SNR, and artifact noise reduction (all P < .001). There were no significant differences in sharpness (P = .07) or diagnostic confidence (P = .06) between the standard DIR and DIRDL protocols. Additionally, DIRDL detected 37% more lesions compared with T2WI (300 versus 219; P < .001).CONCLUSIONS: DIRDL significantly reduces acquisition time and improves image quality compared with standard DIR, without compromising diagnostic confidence. Additionally, DIRDL enhances lesion detection in patients with inflammatory myelopathies, making it a valuable tool in clinical practice. These findings underscore the potential for incorporating DIRDL into future imaging guidelines.AQP4+NMOSDAQP4-IgG positive neuromyelitis optica spectrum disordersDIRdouble inversion recoveryDLdeep learningICCintraclass correlation coefficientIQRinterquartile rangeMOGmyelin oligodendrocyte glycoproteinMOGADMOG antibody-associated diseasesNEXnumber of excitations