Systematic Review Between Resting-State fMRI and Task fMRI in Planning for Brain Tumour Surgery

Figures & data

Figure 1 Diagram illustrating the search strategy used for the studies included in this systematic review.

Table 1 A Collection of Research Analyzing Brain Function Using Rs-fMRI

NO	Study location	Participants number and sex	Age	Handedness	Tumour Types	MRI Acquisition	Pre-Symptoms	Rs-fMRI Analysis
1	Gianvincenzo et al (2020)^Citation48	10 patients (7 male, 3 female)	25 to 67 years (range 51)	9 Right-handed 1 Left-handed	6 anaplastic astrocytoma 3 glioblastoma multiforme 1 anaplastic	3T MRI GE	3 Motor 6 Language 1 Visual	Seed-based
2	Lemée et al (2019)^Citation49	50 patients (34 male, 16 female)	49.6 ± 13.5 years (range 18–75)	44 Right-handed 6 Left-handed	42 glial tumors 2 metastases 6 nontumoral brain lesions	3T MRI Siemens	3 Motor 6 Language 1 Visual	Independent Component Analysis (ICA)
3	Niu et al (2021)^Citation43	50 patients (34 male, 16 female)	24–79 years (average: 54.4 ± 11.3)	109 Right-handed	58 Meningioma. 38 Glioma. 13 brain metastasis	3T MRI GE	43 motor deficits 66 no motor deficits but other symptoms	Independent Component Analysis (ICA)
4	Liouta et al (2019)^Citation44	69 patients (32 male, 37 female)	50 ± 15.5 years	69 Right-handed	33 glioblastomas 16 meningiomas 11 astrocytomas 5 oligodedrogliomas 2 hemangiomas 1 ependymoma 1 neurocytoma	1.5T MRI Siemens	8 Motor 7 Language 54 not display any symptoms	Independent Component Analysis (ICA)
5	Leuthardt et al (2018)^Citation54	191 patients (118 male, 73 female)	3 to 84 years (range 46)	Not reported	41 Low grade glioma 80 High grade glioma 32 Other tumors (DNET, Meningioma, Pilocytic Astrocytoma,	3T MRI Siemens	Not reported	Multilayer Perceptron (MLP)
6	Metwali et al (2020)^Citation55	34 patients (18 male, 16 female)	18–77 years (mean, 54)	Not reported	Ganglioglioma	3T MRI Siemens	Not reported	Region of Interest (ROI)
7	Cho et al (2021)^Citation45	29 patients (18 male, 11 female)	45 ± 14 years	29 Right-handed	5 astrocytoma 4 anaplastic 4 oligodendro 1 anaplastic 13 glioblastoma 2 glioma.	3T MRI GE	language deficits	Seed-based
8	Yordanova et al (2019)^Citation50	23 patients 15 male, 8 female)	23–65 years (mean= 38)	18 Right-handed 3 Left-handed 2 Ambidextrous	15 frontal lobes 2 temporo-occipital lobes 6 adjacent frontal and temporal regions	1.5T & 3T MRI Siemens	Not reported	Seed-based
9	Daniel et al (2021)^Citation51	57 patients (42 male, 15 female)	57.8 ± 13.9 years	22 Right-handed 29 Lift-handed 6 bilateral	57 glioblastoma	3T MRI Siemens	Not reported	Seed-based
10	Thakkar et al (2022)^Citation52	71 patients (34 male, 37 female)	18–75 years (mean= 45.29)	68 Right-handed 3 Left-handed	65 primary tumours 6 metastasis	3T MRI Philips	4 severe linguistic deficits. 17 moderate language disorders 49 mild language disorders	Seed-based
11	Wongsripuemtet et al (2018)^Citation18	66 patients (38 male, 28 female)	40.8 ± 14.6 years	63 Right-handed 3 Left-handed	6 Diffuse astrocytoma 1 Fibrillary astrocytoma 9 Infiltrative astrocytoma 8 Anaplastic astrocytoma 8 Glioblastoma 26 Glioblastoma 3Anaplastic oligoastrocytoma 1 Myeloid sarcoma 1 Metastatic melanoma 1 Metastatic adenocarcinoma 2 No pathologic report	3T MRI Siemens	Not reported	Seed-based
12	Cai et al (2021)^Citation46	126 patients (72 male, 54 female)	42.21 ± 12.74 years	126 right-handed	30 astrocytoma 19 anaplastic 16 oligodendroglioma 1 anaplastic 22 oligoastrocytoma 8 anaplastic 30 glioblastoma	3T MRI Siemens	Not reported	Multilayer Perceptron (MLR)
13	Zacà, D et al (2018)^Citation47	6 patients (4 male, 2 female)	18–45 years	6 right-handed	1 glioma 1 oligodendroglioma 1 anaplastic astrocytoma 2 Cavernoma 1 astrocytoma.	1.5T MRI GE	Not reported	Independent Component Analysis (ICA)
14	Kumar et al (2020)^Citation53	49 patients (28 male, 21 female)	17–78 years (mean= 47.5)	45 Right-handed 2 Left-handed 2 Ambidextrous	10 astrocytoma 5 Anaplastic astrocytoma 22 Glioblastoma 6 Oligodendroglioma 3 Metastasis 1 Pleomorphic xanthoastrocytoma 2 Anaplastic pleomorphic xanthoastrocytoma.	3T MRI GE	12 poor performance 7 suffer from severe impairment (such as aphasia) 20 is weak in the language area. 8 no BOLD activation near tumour. 2 patient motion artifact	Seed-based
15	Anwar et al (2022)^Citation56	22 patients (15 male, 7 female)	2–18 years (mean= 8.6)	Not reported	4 Ganglioglioma 3 Xanthoastrocytoma 3 meningioma 3 dysembryoplastic neuroepithelial tumour 2 Anaplastic Ependymoma 1 Aneurysmal fibrous histiocytoma 1 Pilocytic astrocytoma 1 glioblastoma multiforme 1 Anaplastic astrocytoma 1 Low grade glioma 1 embryonal tumour with multilayered rosettes 1 Ewinga	3T MRI Philips	9 increased ICT or generalized seizures 2 motor seizures 10 neurological deficits 1 motor seizures	Seed-based & ICA
16	Luckett et al (2020)^Citation60	35 patients (23 male, 12 female)	23–71 years (mean= 44.8)	23 Right-handed 3 Left-handed 9 Not reported	10 Glioblastoma 1 Anaplastic glioma 2 Anaplastic mixed oligoastrocytoma 7 Mixed oligoastrocytoma 3 Anaplastic oligodendroglioma 5 Oligodendroglioma 1 Metastatic lung carcinoma 1 Low-grade diffuse glioma 1 Meningioma 1 Ependymoma 1 Low-grade glioneuronal tumour 1 Ganglioglioma 1 Anaplastic astrocytoma	3T MRI Siemens	35 Language	3D convolutional Neural Network (3D CNN)
17	Yahyavi et al (2017)^Citation15	26 patients (15 male, 11 female)	21–69 years (mean = 43.6)	Not reported	Not specific	3T MRI Siemens	26 Ventral Somatomotor Network	Independent Component Analysis (ICA)
18	Nandakumar et al (2021)^Citation61	62 patients (37 male, 25 female)	32–43 years (mean= 38.3)	Not reported	Not specific	3T MRI Siemens	62 Motor and Language deficits	CNN-MTL /Graph Neural Network (GNN)

Table 2 A Comparison of Rs-fMRI and T-fMRI Studies

Author(s)	Type of task	Coefficient of concordance and correlation between rs-fMRI and T-fMRI
Lemée et al (2019)^Citation49	Sentence generation (SG) Tone Listening (TL)	The use of Rs-fMRI for presurgical language mapping is straightforward, allowing for the identification of functional brain language networks with more sensitivity than T-fMRI
Niu et al (2021)^Citation43	Hand Movement	Using T-fMRI activation as a benchmark, the general linear model-based machine learning (GLM-ML) model demonstrated that it can accurately predict individual variations in task activation using low-resolution conventional Rs-fMRI data
Luckett et al (2020)^Citation60	Language task (word-stem completion)	State of rest fMRI can be utilised on all patients, and because it has a lower failure rate than T-fMRI, it is advantageous for patients who cannot participate in task-based tests.
Yahyavi et al (2017)^Citation15	Tongue Motor	At the single-subject level, there is variable concordance between Rs-fMRI and T-fMRI for the ventral somatomotor network
Kumar et al (2020)^Citation53	Letter Fluency Category Fluency Sentence Completion	This study suggests that in the absence of T-fMRI, Rs-fMRI may be a good alternative for clinical language mapping.
Liouta et al (2019)^Citation44	Assessment for Apparent (Paresis) Finger Tapping Assessment for Apparent (Paresis) Mild Language (Phonological Verbal Fluency)	The findings revealed that the Rs-fMRI BOLD signal of the motor and language networks was considerably altered by the tumours, indicating the method’s utility for measuring basic functioning in patients with brain tumours
Nandakumar et al (2021)^Citation61	Finger Tapping Tongue Moving Foot Tapping	The results clearly highlight the advantages of the study-specific graph convolution design combined with multitask learning and the potential for Rs-fMRI to be used as a preoperative mapping instrument

Table 3 Correlation and Validation of Preoperative Rs-fMRI by Intraoperative Cortical Mapping (DES and ECS)

No	Authors	Types of mapping cases	Correlation and validation of preoperative Rs-fMRI to intraoperative cortical stimulation
1	Zaca et al (2019)^Citation47	Motor, language, visual, and speech articulation	The use of ReStNeuMap as an Rs-FMRI processing pipeline showed an excellent spatial agreement between the predictions by preoperative Rs-fMRI and DES findings in almost all the mapping cases during the awake operation.
2	Lemée et al (2019)^Citation49	Language	Of the 32 brain language areas identified by intraoperative cortical mapping (ECS), Rs-fMRI data had 100% sensitivity in the eloquent brain language area identification and functional brain language areas were effectively identified in 4 patients by Rs-fMRI compared to none by T-fMRI. Similarly, Rs-fMRI successfully identified 15 brain language areas out of 18 subjects with negative cortical mapping.
3	Yordanova et al (2019)^Citation50	Mentalizing network	The combination of preoperative Rs-fMRI and intraoperative DES is validated as a viable approach to identify functional networks, particularly the mentalizing networks.

Figure 2 The findings of the content analysis of the studies included in this study.

Sparacia G, Parla G, Re VL, et al. Resting-state functional connectome in patients with brain Tumors before and after surgical resection. World Neurosurg. 2020;141:e182–e194. doi:10.1016/j.wneu.2020.05.054

 PubMed Web of Science ®Google Scholar

Lemée JM, Berro DH, Bernard F, et al. Resting-state functional magnetic resonance imaging versus task-based activity for language mapping and correlation with perioperative cortical mapping. Brain and Behavior. 2019;9(10):e01362. doi:10.1002/brb3.1362

 PubMed Web of Science ®Google Scholar

Niu C, Wang Y, Cohen AD, et al. Machine learning may predict individual hand motor activation from resting-state fMRI in patients with brain tumors in perirolandic cortex. Eur Radiol. 2021;31(7):5253–5262. doi:10.1007/s00330-021-07825-w

 PubMed Web of Science ®Google Scholar

Liouta E, Katsaros VK, Stranjalis G, Leks E, Klose U, Bisdas S. Motor and language deficits correlate with resting state functional magnetic resonance imaging networks in patients with brain tumors. J Neuroradiol. 2019;46(3):199–206. doi:10.1016/j.neurad.2018.08.002

 PubMed Web of Science ®Google Scholar

Leuthardt EC, Guzman G, Bandt SK, et al. Integration of resting state functional MRI into clinical practice-A large single institution experience. PLoS One. 2018;13(6):e0198349. doi:10.1371/journal.pone.0198349

 PubMed Web of Science ®Google Scholar

Metwali H, Raemaekers M, Ibrahim T, Samii A. Inter-network functional connectivity changes in patients with brain tumors: a resting-state functional magnetic resonance imaging study. World Neurosurg. 2020;138:e66–e71. doi:10.1016/j.wneu.2020.01.177

 PubMed Web of Science ®Google Scholar

Cho NS, Peck KK, Gene MN, Jenabi M, Holodny AI. Resting-state functional MRI language network connectivity differences in patients with brain tumors: exploration of the cerebellum and contralesional hemisphere. Brain Imag Behav. 2022;16(1):252–262. doi:10.1007/s11682-021-00498-5

 PubMed Web of Science ®Google Scholar

Yordanova YN, Cochereau J, Duffau H, Herbet G. Combining resting state functional MRI with intraoperative cortical stimulation to map the mentalizing network. Neuroimage. 2019;186:628–636. doi:10.1016/j.neuroimage.2018.11.046

 PubMed Web of Science ®Google Scholar

Daniel AG, Park KY, Roland JL, et al. Functional connectivity within glioblastoma impacts overall survival. Neuro-Oncology. 2021;23(3):412–421. doi:10.1093/neuonc/noaa189

 PubMed Web of Science ®Google Scholar

Thakkar I, Arraño-Carrasco L, Cortes-Rivera B, et al. Alterna- tive language paradigms for functional magnetic resonance imaging as presurgical tools for inducing crossed cerebro- cerebellar language activations in brain tumor patients. Eur Radiol. 2022;32(1):300–307. doi:10.1007/s00330-021-08137-9

 PubMed Web of Science ®Google Scholar

Wongsripuemtet J, Tyan A, Carass A, et al. Preoperative mapping of the supplementary motor area in patients with brain tumor using resting-state fMRI with seed-based analysis. Am J Neu-Roradiol. 2018;39(8):1493–1498.

 PubMed Web of Science ®Google Scholar

Cai S, Shi Z, Jiang C, et al. Hemisphere-specific functional remodeling and its relevance to tumor malignancy of cerebral glioma based on resting-state functional network analysis. Front Neurosci. 2021;14:611075. doi:10.3389/fnins.2020.611075

 PubMed Web of Science ®Google Scholar

Zacà D, Jovicich J, Corsini F, Rozzanigo U, Chioffi F, Sarubbo S. ReStNeuMap: a tool for automatic extraction of resting- state functional MRI networks in neurosurgical practice. J Neurosurg. 2018;131(3):764–771. doi:10.3171/2018.4.JNS18474

 PubMed Web of Science ®Google Scholar

Kumar VA, Heiba IM, Prabhu SS, et al. The role of resting-state functional MRI for clinical preoperative language mapping. Cancer Imaging. 2020;20(1):1–9. doi:10.1186/s40644-020-00327-w

 Web of Science ®Google Scholar

Anwar A, Radwan A, Zaky I, El Ayadi M, Youssef A. Resting state fMRI brain mapping in pediatric supratentorial brain tumors. Egypt J Radiol Nucl Med. 2022;53(1):1–10. doi:10.1186/s43055-022-00713-3

 Google Scholar

Luckett P, Lee JJ, Park KY, et al. Mapping of the language network with deep learning. Front Neurol. 2020;11:819. doi:10.3389/fneur.2020.00819

 PubMed Web of Science ®Google Scholar

Yahyavi-Firouz-Abadi N, Pillai J, Lindquist M, et al. Presurgical brain mapping of the ven- tral somatomotor network in patients with brain tumors using resting-state fMRI. Am J Neuroradiol. 2017;38(5):1006–1012. doi:10.3174/ajnr.A5132

 PubMed Web of Science ®Google Scholar

Nandakumar N, Manzoor K, Agarwal S, et al. Automated eloquent cortex localization in brain tumor patients using multi-task graph neural networks. Med Image Anal. 2021;74:102203. doi:10.1016/j.media.2021.102203

 PubMed Web of Science ®Google Scholar