Establishing an acquisition and processing protocol for resting state networks with a 1.5 T scanner: A case series in a middle-income country

• Published in: Medicine (Baltimore) 2020-07, Vol.99 (28), p.e21125-e21125
• Main Authors: Moreno-Ayure, Michela, Páez, Cristian, López-Arias, María A., Mendez-Betancurt, Johan L., Ordóñez-Rubiano, Edgar G., Rudas, Jorge, Pulido, Cristian, Gómez, Francisco, Martínez, Darwin, Enciso-Olivera, Cesar O., Rivera-Triana, Diana P., Casanova-Libreros, Rosangela, Aguilera, Natalia, Marín-Muñoz, Jorge H.
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Brain - diagnostic imaging; Brain Mapping - methods; Clinical Case Report; Equipment Design; Female; Follow-Up Studies; Healthy Volunteers; Humans; Image Processing, Computer-Assisted; Magnetic Resonance Imaging - instrumentation; Male; Middle Aged; Nerve Net - diagnostic imaging; Reproducibility of Results; Rest - physiology; Retrospective Studies
• ISSN: 0025-7974; 1536-5964
• DOI: 10.1097/MD.0000000000021125

The aim of this study was to characterize the capability of detection of the resting state networks (RSNs) with functional magnetic resonance imaging (fMRI) in healthy subjects using a 1.5T scanner in a middle-income country. Ten subjects underwent a complete blood-oxygen-level dependent imaging (BOLD) acquisition on a 1.5T scanner. For the imaging analysis, we used the spatial independent component analysis (sICA). We designed a computer tool for 1.5 T (or above) scanners for imaging processing. We used it to separate and delineate the different components of the RSNs of the BOLD signal. The sICA was also used to differentiate the RSNs from noise artifact generated by breathing and cardiac cycles. For each subject, 20 independent components (IC) were computed from the sICA (a total of 200 ICs). From these ICs, a spatial pattern consistent with RSNs was identified in 161 (80.5%). From the 161, 131 (65.5%) were fit for study. The networks that were found in all subjects were: the default mode network, the right executive control network, the medial visual network, and the cerebellar network. In 90% of the subjects, the left executive control network and the sensory/motor network were observed. The occipital visual network was present in 80% of the subjects. In 39 (19.5%) of the images, no any neural network was identified. Reproduction and differentiation of the most representative RSNs was achieved using a 1.5T scanner acquisitions and sICA processing of BOLD imaging in healthy subjects.