4D upper gastrointestinal magnetic resonance imaging in healthy human subjects and gastroparetic patients

Study Purpose: The purpose of this study is to characterize other possible gastric dysfunctions in gastroparesis by using Cine magnetic resonance imaging (MRI) and elucidate the potential correlation between regional gastric motility parameters and symptom scores.

Data Collection: 3D and 4D MRI data was collected on both healthy volunteers and on gastroparetic patients before and after consuming a high contrast meal

Primary Conclusion: Multiple gastric motility parameters can be concurrently and quantitatively assessed by Cine MRI. Regional gastric abnormalities were found to be associated with patient symptoms. This study demonstrates that a fuller assessment of regional gastric functions in gastroparesis can improve the understanding of the relationship between gastric dysmotility and gastroparetic symptoms.

Curator's Notes

Experimental Design: This study aims to characterize gastric motility and emptying patterns in healthy human subjects and gastroparetic patients. Briefly, 4D gastric MRI scans were taken after subjects finished a caloric, nutritionally complete meal. Gastric volumes were segmented and reconstructed into 3D isosurfaces for the estimation of gastric motility. The frequency, amplitude, and coordination of peristaltic contraction waves could be computed from gastric wall movements. The data is taken from measurements on 27 healthy volunteers (11M/17F) and 19 gastroparetic patients (6M/13F).

Completeness: This dataset is complete.

Subjects & Samples: Male (n= 17) and female (n=30) human subjects were used in this study.

Primary vs derivative data: Primary data includes about 2500 files totaling about 670GB, comprising 3D and 4D NIFTI weighted gastric MRI data. The derivative folder contains segmented data and various measurements in Matlab format. In addition, an Excel spreadsheet of gastroparetic patient responses to a symptom assessment questionnaire is included.

Code Availability:The code folder contains the main source code (built under Windows 10 Matlab r2019a) related to Lu et al., 2021. Example datasets (human gastric MRI images) are provided along with the source code. The scripts are organized according to prefixes:

Global scripts: Path to root folder, parameters necessary for image processing, etc.

Amri scripts: Image processing scripts for main processing and sub-functions.

Tools scripts: functions for general purposes.