1
|
Lin D, Liu J, Ke C, Chen H, Li J, Xie Y, Ma J, Lv X, Feng Y. Radiomics Analysis of Quantitative Maps from Synthetic MRI for Predicting Grades and Molecular Subtypes of Diffuse Gliomas. Clin Neuroradiol 2024:10.1007/s00062-024-01421-3. [PMID: 38858272 DOI: 10.1007/s00062-024-01421-3] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/17/2023] [Accepted: 05/03/2024] [Indexed: 06/12/2024]
Abstract
PURPOSE To investigate the feasibility of using radiomics analysis of quantitative maps from synthetic MRI to preoperatively predict diffuse glioma grades, isocitrate dehydrogenase (IDH) subtypes, and 1p/19q codeletion status. METHODS Data from 124 patients with diffuse glioma were used for analysis (n = 87 for training, n = 37 for testing). Quantitative T1, T2, and proton density (PD) maps were obtained using synthetic MRI. Enhancing tumour (ET), non-enhancing tumour and necrosis (NET), and peritumoral edema (PE) regions were segmented followed by manual fine-tuning. Features were extracted using PyRadiomics and then selected using Levene/T, BorutaShap and maximum relevance minimum redundancy algorithms. A support vector machine was adopted for classification. Receiver operating characteristic curve analysis and integrated discrimination improvement analysis were implemented to compare the performance of different radiomics models. RESULTS Radiomics models constructed using features from multiple tumour subregions (ET + NET + PE) in the combined maps (T1 + T2 + PD) achieved the highest AUC in all three prediction tasks, among which the AUC for differentiating lower-grade and high-grade diffuse gliomas, predicting IDH mutation status and predicting 1p/19q codeletion status were 0.92, 0.95 and 0.86 respectively. Compared with those constructed on individual T1, T2, and PD maps, the discriminant ability of radiomics models constructed on the combined maps separately increased by 11, 17 and 10% in predicting glioma grades, 35, 52 and 19% in predicting IDH mutation status, and 16, 15 and 14% in predicting 1p/19q codeletion status (p < 0.05). CONCLUSION Radiomics analysis of quantitative maps from synthetic MRI provides a new quantitative imaging tool for the preoperative prediction of grades and molecular subtypes in diffuse gliomas.
Collapse
Affiliation(s)
- Danlin Lin
- Department of Medical Imaging, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
| | - Jiehong Liu
- School of Biomedical Engineering, Southern Medical University, Guangzhou, China
| | - Chao Ke
- Department of Neurosurgery, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
| | - Haolin Chen
- Department of Radiation Oncology, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
| | - Jing Li
- Department of Medical Imaging, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
| | - Yuanyao Xie
- School of Biomedical Engineering, Southern Medical University, Guangzhou, China
| | - Jianhua Ma
- School of Biomedical Engineering, Southern Medical University, Guangzhou, China
- Guangdong Provincial Key Laboratory of Medical Image Processing & Guangdong Province Engineering Laboratory for Medical Imaging and Diagnostic Technology, Southern Medical University, Guangzhou, China
| | - Xiaofei Lv
- Department of Medical Imaging, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China.
| | - Yanqiu Feng
- School of Biomedical Engineering, Southern Medical University, Guangzhou, China.
- Guangdong Provincial Key Laboratory of Medical Image Processing & Guangdong Province Engineering Laboratory for Medical Imaging and Diagnostic Technology, Southern Medical University, Guangzhou, China.
- Guangdong-Hong Kong-Macao Greater Bay Area Centre for Brain Science and Brain-Inspired Intelligence & Key Laboratory of Mental Health of the Ministry of Education, Guangzhou, China.
- Department of Radiology, The First People's Hospital of Shunde, Southern Medical University, Foshan, China.
| |
Collapse
|
2
|
Ye Y, Xu J, Zhang Z, Zhang Y, Zhao Q, Xu J, Yuan H. Complex multi-dimensional integration for T 2* and R 2* mapping. Magn Reson Imaging 2024; 108:29-39. [PMID: 38301862 DOI: 10.1016/j.mri.2024.01.018] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/06/2023] [Revised: 01/21/2024] [Accepted: 01/29/2024] [Indexed: 02/03/2024]
Abstract
A dual Multi-Dimensional Integration (dMDI) method was proposed and demonstrated for T2* and R2* mapping. By constructing and jointly using both the original MDI term and an inversed MDI term, T2* and R2* mapping can be performed independently with intrinsic background noise suppression and spike elimination, allowing for high quantitative accuracy and robustness over a wide range of T2*. dMDI was compared to original MDI and curve fitting methods in terms of quantitative specificity, accuracy, reliability and computational efficiency. All methods were tested and compared via simulation and in vivo data. With high signal-to-noise-ratio (SNR), the proposed dMDI method yielded T2*and R2* values similar to curve fitting methods. For low SNR and background noise signals, the dMDI yielded low T2* and R2* values, thus effectively suppressing all background noise. Virtually zero spikes were observed in dMDI T2* and R2* maps in all simulation and imaging results. The dMDI method has the potential to provide improved and reliable T2* and R2* mapping results in routine and SNR-challenging scenarios.
Collapse
Affiliation(s)
- Yongquan Ye
- United Imaging Healthcare, Houston, TX, USA.
| | - Jian Xu
- United Imaging Healthcare, Houston, TX, USA
| | | | - Yan Zhang
- Beijing United Imaging Intelligent Imaging Technology Research Institute, Beijing, China
| | - Qiang Zhao
- Department of Radiology, Peking University Third Hospital, Beijing, China
| | - Jiajia Xu
- Department of Radiology, Peking University Third Hospital, Beijing, China
| | - Huishu Yuan
- Department of Radiology, Peking University Third Hospital, Beijing, China
| |
Collapse
|
3
|
De A, Grenier J, Wilman AH. Simultaneous time-of-flight MR angiography and quantitative susceptibility mapping with key time-of-flight features. NMR IN BIOMEDICINE 2024; 37:e5079. [PMID: 38054247 DOI: 10.1002/nbm.5079] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 01/15/2023] [Revised: 10/30/2023] [Accepted: 11/05/2023] [Indexed: 12/07/2023]
Abstract
A technique for combined time-of-flight (TOF) MR angiography (MRA) and quantitative susceptibility mapping (QSM) was developed with key features of standard three-dimensional (3D) TOF acquisitions, including multiple overlapping thin slab acquisition (MOTSA), ramped RF excitation, and venous saturation. The developed triple-echo 3D TOF-QSM sequence enabled TOF-MRA, susceptibility-weighted imaging (SWI), QSM, and R2* mapping. The effects of ramped RF, resolution, flip angle, venous saturation, and MOTSA were studied on QSM. Six volunteers were scanned at 3 T with the developed sequence, conventional TOF-MRA, and conventional SWI. Quantitative comparison of susceptibility values on QSM and normalized arterial and venous vessel-to-background contrasts on TOF and SWI were performed. The ramped RF excitation created an inherent phase variation in the raw phase. A generic correction factor was computed to remove the phase variation to obtain QSM without artifacts from the TOF-QSM sequence. No statistically significant difference was observed between the developed and standard QSM sequence for susceptibility values. However, maintaining standard TOF features led to compromises in signal-to-noise ratio for QSM and SWI, arising from the use of MOTSA rather than one large 3D slab, higher TOF spatial resolution, increased TOF background suppression due to larger flip angles, and reduced venous signal from venous saturation. In terms of vessel contrast, veins showed higher normalized contrast on SWI derived from TOF-QSM than the standard SWI sequence. While fast flowing arteries had reduced contrast compared with standard TOF-MRA, no statistical difference was observed for slow flowing arteries. Arterial contrast differences largely arise from the longer TR used in TOF-QSM over standard TOF-MRA to accommodate additional later echoes for SWI. In conclusion, although the sequence has a longer TR and slightly lower arterial contrast, provided an adequate correction is made for ramped RF excitation effects on phase, QSM may be performed from a multiecho sequence that includes all key TOF features, thus enabling simultaneous TOF-MRA, SWI, QSM, and R2* map computation.
Collapse
Affiliation(s)
- Ashmita De
- Department of Biomedical Engineering, University of Alberta, Edmonton, Canada
| | - Justin Grenier
- Department of Biomedical Engineering, University of Alberta, Edmonton, Canada
| | - Alan H Wilman
- Department of Biomedical Engineering, University of Alberta, Edmonton, Canada
- Department of Radiology and Diagnostic Imaging, University of Alberta, Edmonton, Canada
| |
Collapse
|
4
|
Duan M, Pan R, Gao Q, Wu X, Lin H, Yuan J, Zhang Y, Liu L, Tian Y, Fu T. A rapid multi-parametric quantitative MR imaging method to assess Parkinson's disease: a feasibility study. BMC Med Imaging 2024; 24:58. [PMID: 38443786 PMCID: PMC10916029 DOI: 10.1186/s12880-024-01229-0] [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] [Received: 10/19/2023] [Accepted: 02/15/2024] [Indexed: 03/07/2024] Open
Abstract
BACKGROUND MULTIPLEX is a single-scan three-dimensional multi-parametric MRI technique that provides 1 mm isotropic T1-, T2*-, proton density- and susceptibility-weighted images and the corresponding quantitative maps. This study aimed to investigate its feasibility of clinical application in Parkinson's disease (PD). METHODS 27 PD patients and 23 healthy control (HC) were recruited and underwent a MULTIPLEX scanning. All image reconstruction and processing were automatically performed with in-house C + + programs on the Automatic Differentiation using Expression Template platform. According to the HybraPD atlas consisting of 12 human brain subcortical nuclei, the region-of-interest (ROI) based analysis was conducted to extract quantitative parameters, then identify PD-related abnormalities from the T1, T2* and proton density maps and quantitative susceptibility mapping (QSM), by comparing patients and HCs. RESULTS The ROI-based analysis revealed significantly decreased mean T1 values in substantia nigra pars compacta and habenular nuclei, mean T2* value in subthalamic nucleus and increased mean QSM value in subthalamic nucleus in PD patients, compared to HCs (all p values < 0.05 after FDR correction). The receiver operating characteristic analysis showed all these four quantitative parameters significantly contributed to PD diagnosis (all p values < 0.01 after FDR correction). Furthermore, the two quantitative parameters in subthalamic nucleus showed hemicerebral differences in regard to the clinically dominant side among PD patients. CONCLUSIONS MULTIPLEX might be feasible for clinical application to assist in PD diagnosis and provide possible pathological information of PD patients' subcortical nucleus and dopaminergic midbrain regions.
Collapse
Affiliation(s)
- Min Duan
- Department of Radiology, Jiangsu Province Hospital of Chinese Medicine, Affiliated Hospital of Nanjing University of Chinese Medicine, Nanjing, China
| | - Rongrong Pan
- Department of Neurology, Nanjing First Hospital, Nanjing Medical University, No.68, Changle Road, 210006, Nanjing, Jiangsu Province, China
| | - Qing Gao
- Department of Neurology, Nanjing First Hospital, Nanjing Medical University, No.68, Changle Road, 210006, Nanjing, Jiangsu Province, China
| | - Xinying Wu
- Department of Radiology, Nanjing First Hospital, Nanjing Medical University, No.68, Changle Road, 210006, Nanjing, Jiangsu Province, China
| | - Hai Lin
- Central Research Institute, United Imaging Healthcare, Shanghai, China
| | - Jianmin Yuan
- Central Research Institute, United Imaging Healthcare, Shanghai, China
| | - Yamei Zhang
- Department of Radiology, Nanjing First Hospital, Nanjing Medical University, No.68, Changle Road, 210006, Nanjing, Jiangsu Province, China
| | - Lindong Liu
- Department of Radiology, Nanjing First Hospital, Nanjing Medical University, No.68, Changle Road, 210006, Nanjing, Jiangsu Province, China
| | - Youyong Tian
- Department of Neurology, Nanjing First Hospital, Nanjing Medical University, No.68, Changle Road, 210006, Nanjing, Jiangsu Province, China.
| | - Tong Fu
- Department of Radiology, Nanjing First Hospital, Nanjing Medical University, No.68, Changle Road, 210006, Nanjing, Jiangsu Province, China.
| |
Collapse
|
5
|
Gu Y, Pan Y, Fang Z, Ma L, Zhu Y, Androjna C, Zhong K, Yu X, Shen D. Deep learning-assisted preclinical MR fingerprinting for sub-millimeter T 1 and T 2 mapping of entire macaque brain. Magn Reson Med 2024; 91:1149-1164. [PMID: 37929695 DOI: 10.1002/mrm.29905] [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] [Received: 06/29/2023] [Revised: 09/10/2023] [Accepted: 10/10/2023] [Indexed: 11/07/2023]
Abstract
PURPOSE Preclinical MR fingerprinting (MRF) suffers from long acquisition time for organ-level coverage due to demanding image resolution and limited undersampling capacity. This study aims to develop a deep learning-assisted fast MRF framework for sub-millimeter T1 and T2 mapping of entire macaque brain on a preclinical 9.4 T MR system. METHODS Three dimensional MRF images were reconstructed by singular value decomposition (SVD) compressed reconstruction. T1 and T2 mapping for each axial slice exploited a self-attention assisted residual U-Net to suppress aliasing-induced quantification errors, and the transmit-field (B1 + ) measurements for robustness against B1 + inhomogeneity. Supervised network training used MRF images simulated via virtual parametric maps and a desired undersampling scheme. This strategy bypassed the difficulties of acquiring fully sampled preclinical MRF data to guide network training. The proposed fast MRF framework was tested on experimental data acquired from ex vivo and in vivo macaque brains. RESULTS The trained network showed reasonable adaptability to experimental MRF images, enabling robust delineation of various T1 and T2 distributions in the brain tissues. Further, the proposed MRF framework outperformed several existing fast MRF methods in handling the aliasing artifacts and capturing detailed cerebral structures in the mapping results. Parametric mapping of entire macaque brain at nominal resolution of 0.35× $$ \times $$ 0.35× $$ \times $$ 1 mm3 can be realized via a 20-min 3D MRF scan, which was sixfold faster than the baseline protocol. CONCLUSION Introducing deep learning to MRF framework paves the way for efficient organ-level high-resolution quantitative MRI in preclinical applications.
Collapse
Affiliation(s)
- Yuning Gu
- School of Biomedical Engineering & State Key Laboratory of Advanced Medical Materials and Devices, ShanghaiTech University, Shanghai, China
| | - Yongsheng Pan
- School of Biomedical Engineering & State Key Laboratory of Advanced Medical Materials and Devices, ShanghaiTech University, Shanghai, China
| | - Zhenghan Fang
- Department of Biomedical Engineering, Johns Hopkins University, Baltimore, Maryland, USA
| | - Lei Ma
- School of Biomedical Engineering & State Key Laboratory of Advanced Medical Materials and Devices, ShanghaiTech University, Shanghai, China
| | - Yuran Zhu
- Department of Biomedical Engineering, Case Western Reserve University, Cleveland, Ohio, USA
| | - Charlie Androjna
- Cleveland Clinic Pre-Clinical Magnetic Resonance Imaging Center, Cleveland Clinic Foundation, Cleveland, Ohio, USA
| | - Kai Zhong
- High Magnetic Field Laboratory, Chinese Academy of Sciences, Hefei, China
- Anhui Province Key Laboratory of High Field Magnetic Resonance Imaging, Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei, China
- Biomedical Engineering Department, Peking University, Beijing, China
| | - Xin Yu
- Department of Biomedical Engineering, Case Western Reserve University, Cleveland, Ohio, USA
- Department of Radiology, Case Western Reserve University, Cleveland, Ohio, USA
| | - Dinggang Shen
- School of Biomedical Engineering & State Key Laboratory of Advanced Medical Materials and Devices, ShanghaiTech University, Shanghai, China
- Shanghai United Imaging Intelligence, Shanghai, China
- Shanghai Clinical Research and Trial Center, ShanghaiTech University, Shanghai, China
| |
Collapse
|