FastMRT is a collaborative research project that aims to improve the speed and quality of MRI scans for thermometery. Our goal is to make MRI scans up faster while maintaining or even improving image quality.
FastMRT is a collaborative research project that aims to improve the speed and quality of MRI scans for thermometery.
Our team of researchers and engineers are working on developing new machine learning algorithms to achieve this goal. We are also working with clinical partners to validate our methods and ensure that they are safe and effective for patients.
If you are interested in learning more about our research or collaborating with us, please contact us at fastmr.thermometry@gmail.com.
We public the github codes at https://github.com/minipuding/FastMRT.
The FastMRT dataset is now available for download! Visit our Datasets page to learn more about phantom and ex vivo datasets.
FastMRT package is released on PyPI. You can view at here or install by command `pip install fastmrt`.
Our dataset was acquired through the application of a 128-element high-intensity focused ultrasound transducer (with a frequency of $1.1 MHz$, a focal length of $150 mm$, and a focal radius of $120 mm$), followed by imaging with a 3T MR system (Discovery MR750; GE Healthcare, Milwaukee, WI).
The images were obtained using the Fast Spoiled Gradient Echo (FSPGR) sequence, with 96 phase encoding steps, a TR/TE of $12/16 ms$, a flip angle of $30^\circ$, a slice thickness of $3mm$, a field of view (FOV) of $28\times 28 cm^2$, a Number of Excitation (NEX) of 1, and a bandwidth of $\pm62.5kHz$.
The dataset comprises of two distinct parts, namely in vivo heating data and ex vivo lean tissue heating data. For each part, there are 96 heating samples (consisting of 2186 slices) and 105 samples (consisting of 1623 slices), respectively, with each sample containing either 1 or 3 layers.
The temperature change at the focus was approximately 30 degrees Celsius, and the focus position was consistently located at the center of the image.
In order to enhance the speed of temperature measurement, we employed a smaller TR and fewer phase encoding steps, which led to lower signal-to-noise ratio and resolution of the acquired images. This underscores the significance of utilizing fast temperature measurement algorithms to compensate for the reduced image quality.
If you have any questions or comments about FastMRT, please contact us at fastmr.thermometry@gmail.com.
By entering my electronic signature here, I agree to the data use agreement above.
Please send your real name, affiliation, and email with the subject "FastMRT Data Set Request" to fastmr.thermometry@gmail.com. We will review your application within 24 hours and send the download link to the provided email.