1
|
Awan KM, Goncalves Filho ALM, Tabari A, Applewhite BP, Lang M, Lo WC, Sellers R, Kollasch P, Clifford B, Nickel D, Husseni J, Rapalino O, Schaefer P, Cauley S, Huang SY, Conklin J. Diagnostic evaluation of deep learning accelerated lumbar spine MRI. Neuroradiol J 2024; 37:323-331. [PMID: 38195418 PMCID: PMC11138337 DOI: 10.1177/19714009231224428] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/11/2024] Open
Abstract
BACKGROUND AND PURPOSE Deep learning (DL) accelerated MR techniques have emerged as a promising approach to accelerate routine MR exams. While prior studies explored DL acceleration for specific lumbar MRI sequences, a gap remains in comprehending the impact of a fully DL-based MRI protocol on scan time and diagnostic quality for routine lumbar spine MRI. To address this, we assessed the image quality and diagnostic performance of a DL-accelerated lumbar spine MRI protocol in comparison to a conventional protocol. METHODS We prospectively evaluated 36 consecutive outpatients undergoing non-contrast enhanced lumbar spine MRIs. Both protocols included sagittal T1, T2, STIR, and axial T2-weighted images. Two blinded neuroradiologists independently reviewed images for foraminal stenosis, spinal canal stenosis, nerve root compression, and facet arthropathy. Grading comparison employed the Wilcoxon signed rank test. For the head-to-head comparison, a 5-point Likert scale to assess image quality, considering artifacts, signal-to-noise ratio (SNR), anatomical structure visualization, and overall diagnostic quality. We applied a 15% noninferiority margin to determine whether the DL-accelerated protocol was noninferior. RESULTS No significant differences existed between protocols when evaluating foraminal and spinal canal stenosis, nerve compression, or facet arthropathy (all p > .05). The DL-spine protocol was noninferior for overall diagnostic quality and visualization of the cord, CSF, intervertebral disc, and nerve roots. However, it exhibited reduced SNR and increased artifact perception. Interobserver reproducibility ranged from moderate to substantial (κ = 0.50-0.76). CONCLUSION Our study indicates that DL reconstruction in spine imaging effectively reduces acquisition times while maintaining comparable diagnostic quality to conventional MRI.
Collapse
Affiliation(s)
- Komal M Awan
- Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, USA
- Harvard Medical School, USA
| | | | - Azadeh Tabari
- Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, USA
- Harvard Medical School, USA
| | - Brooks P Applewhite
- Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, USA
- Harvard Medical School, USA
| | - Min Lang
- Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, USA
- Harvard Medical School, USA
| | | | | | | | | | | | - Jad Husseni
- Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, USA
- Harvard Medical School, USA
| | - Otto Rapalino
- Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, USA
- Harvard Medical School, USA
| | - Pamela Schaefer
- Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, USA
- Harvard Medical School, USA
| | | | - Susie Y Huang
- Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, USA
- Harvard Medical School, USA
- Harvard-MIT Health Sciences and Technology, Massachusetts Institute of Technology, USA
| | - John Conklin
- Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, USA
- Harvard Medical School, USA
| |
Collapse
|
2
|
Vossough A. Newer MRI Techniques in Pediatric Neuroimaging. Semin Roentgenol 2023; 58:131-144. [PMID: 36732007 DOI: 10.1053/j.ro.2022.10.002] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/17/2022] [Accepted: 10/27/2022] [Indexed: 11/23/2022]
Affiliation(s)
- Arastoo Vossough
- Children's Hospital of Philadelphia, University of Pennsylvania, Philadelphia, PA..
| |
Collapse
|
3
|
Yun SY, Heo YJ. Accelerated Nonenhanced 3D T1-MPRAGE Using Wave-Controlled Aliasing in Parallel Imaging for Infant Brain Imaging. AJNR Am J Neuroradiol 2022; 43:1797-1801. [PMID: 36265893 DOI: 10.3174/ajnr.a7680] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/16/2022] [Accepted: 09/19/2022] [Indexed: 02/01/2023]
Abstract
BACKGROUND AND PURPOSE MPRAGE is the most commonly used sequence for high-resolution 3D T1-weighted imaging in pediatric patients. However, its longer scan time is a major drawback because pediatric patients are prone to motion and frequently require sedation. This study compared nonenhanced accelerated MPRAGE using wave-controlled aliasing in parallel imaging (wave-T1-MPRAGE) with standard MPRAGE in infants. MATERIALS AND METHODS We retrospectively evaluated 68 infants (mean age, 1.78 [SD. 1.70] months) who underwent nonenhanced standard and wave-T1-MPRAGE. Two neuroradiologists independently assessed each image for image quality, artifacts, myelination degree, and anatomic delineation using the 4-point Likert scale. For diagnostic performance, both observers determined whether nonenhancing lesions were present in the brain parenchyma in 2 types of nonenhanced MPRAGE sequences. RESULTS Wave-T1-MPRAGE showed a significantly lower mean score and lower interobserver agreement for overall image quality and anatomic delineation than standard MPRAGE (P< .001 for each). However, there were no significant differences between the 2 types of MPRAGE sequences for motion artifacts (P = .90 for observer 1, P = .14 for observer 2) and degree of myelination (P = .16 for observer 1, P = .32 for observer 2). Among the nonenhancing pathologic lesions observed on standard MPRAGE by both observers, only 2 were missed on wave-T1-MPRAGE, and they were very tiny, faint, nonhemorrhagic WM injuries. CONCLUSIONS Although wave-T1-MPRAGE showed lower overall image quality than standard MPRAGE, the diagnostic performance for nonenhancing parenchymal lesions was comparable. Wave-T1-MPRAGE could be an alternative for diagnosing intracranial lesions in infants, with marked scan time reduction.
Collapse
Affiliation(s)
- S Y Yun
- From the Department of Radiology, Inje University Busan Paik Hospital, Busan, South Korea
| | - Y J Heo
- From the Department of Radiology, Inje University Busan Paik Hospital, Busan, South Korea
| |
Collapse
|
4
|
Gallo-Bernal S, Bedoya MA, Gee MS, Jaimes C. Pediatric magnetic resonance imaging: faster is better. Pediatr Radiol 2022:10.1007/s00247-022-05529-x. [PMID: 36261512 DOI: 10.1007/s00247-022-05529-x] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 06/23/2022] [Revised: 08/29/2022] [Accepted: 10/03/2022] [Indexed: 10/24/2022]
Abstract
Magnetic resonance imaging (MRI) has emerged as the preferred imaging modality for evaluating a wide range of pediatric medical conditions. Nevertheless, the long acquisition times associated with this technique can limit its widespread use in young children, resulting in motion-degraded or non-diagnostic studies. As a result, sedation or general anesthesia is often necessary to obtain diagnostic images, which has implications for the safety profile of MRI, the cost of the exam and the radiology department's clinical workflow. Over the last decade, several techniques have been developed to increase the speed of MRI, including parallel imaging, single-shot acquisition, controlled aliasing techniques, compressed sensing and artificial-intelligence-based reconstructions. These are advantageous because shorter examinations decrease the need for sedation and the severity of motion artifacts, increase scanner throughput, and improve system efficiency. In this review we discuss a framework for image acceleration in children that includes the synergistic use of state-of-the-art MRI hardware and optimized pulse sequences. The discussion is framed within the context of pediatric radiology and incorporates the authors' experience in deploying these techniques in routine clinical practice.
Collapse
Affiliation(s)
- Sebastian Gallo-Bernal
- Department of Radiology, Massachusetts General Hospital, Boston, MA, USA.,Department of Radiology, Harvard Medical School, Boston, MA, USA
| | - M Alejandra Bedoya
- Department of Radiology, Harvard Medical School, Boston, MA, USA.,Department of Radiology, Boston Children's Hospital, 300 Longwood Ave., 2nd floor, Main Building, Boston, MA, 02115, USA
| | - Michael S Gee
- Department of Radiology, Massachusetts General Hospital, Boston, MA, USA.,Department of Radiology, Harvard Medical School, Boston, MA, USA
| | - Camilo Jaimes
- Department of Radiology, Harvard Medical School, Boston, MA, USA. .,Department of Radiology, Boston Children's Hospital, 300 Longwood Ave., 2nd floor, Main Building, Boston, MA, 02115, USA.
| |
Collapse
|