Utilizes Real-Time Data Processing to Continually Monitor for Patient Motion and Imaging Artifacts That Could Render fMRI Data Unusable
This software provides quantitative real-time quality assessments of functional magnetic resonance imaging (fMRI) data. Functional MRI (fMRI) measures brain activity by detecting changes in magnetization between oxygen-rich and oxygen-poor blood; fMRI is being adapted widely to better understand how the human brain functions both in a normal and pathologic capacity. Generally, patient motion is a well-known problem associated with MRI scans. Minuscule movements, which are imperceptible to the human eye, can result in corrupted or non-satisfactory data and scans. Unfortunately, such data corruption is typically discovered once the scan is complete and the patient is no longer available. Attempts have been made to develop MRI sequences that monitor for patient motion and can realign the scanner with the head of the patient, but the time-dependence of the data acquisition prohibits reacquiring a previous corrupted data point. Researchers at the University of Florida have created a better alternative, a quantitative real-time assessment of data quality measures in fMRI. Based upon multiple measures of data quality, the system using machine-learning systems to provide simple and intuitive feedback to the user regrading data quality (GREEN/YELLOW/RED notifications) This software will alert users in real-time, as the scan is being acquired, if a data quality issue arises. Users can then end the scan to prevent continued waste of valuable scanner time and attempt to better immobilize or counsel the patient on the importance of remaining still.
Functional MRI software that provides quantitative real-time assessments of data quality measures
- Assesses data quality in real-time, increasing the likelihood of acquiring usable fMRI data
- Alerts users using color-coded signals, making it easy to identify failing, satisfactory, and high quality results
- Provides basic evaluation of head displacement and patient motion, reducing cost and time wasted due to MRI repetition
- Software may be added to existing fMRI systems, increasing functionality and decreasing costs
The software utilizes real-time data processing to continually monitor for patient motion and other common imaging artifacts that could render data unusable. Quality is determined based on a classifier system analyzing a unique set of parameters identified by the inventors that are critical to data quality. The parameters include signal to noise ratio, signal to noise fluctuation ratio, signal to ghost ratio, head displacement, global intensity average and variance for a respective four-dimensional data set. The user will receive simple color-coded feedback (red, yellow, green) on the overall data quality of the data, providing clinical decision support to help physicians, medical personnel, and patients obtain high quality MRI scans.