Ariel Hannum: Towards Optimized and Vendor Agnostic Cardiac Diffusion Tensor Imaging MRI
Ariel Hannum: Towards Optimized and Vendor Agnostic Cardiac Diffusion Tensor Imaging MRI
This summer, I worked as a guest Scientist in the Division of Medical Physics at the University Medical Center Freiburg, Germany, within the Department of Diagnostic and Interventional Radiology. I was based in the MR Technologies Group under the guidance of Principal Investigator Professor Maxim Zaitsev and Group Leader Sebastian Littin. My time in Freiburg, Germany allowed me to advance my PhD research, build new collaborations, and generate innovative ideas.
In front of the ultra-high-performance MRI system where I conducted most of my research this summer.
At Stanford University, my work focuses on advancing Cardiac Diffusion Tensor Imaging (cDTI), a non-invasive MRI technique that maps the microstructure of heart tissue and shows promise as a biomarker for cardiovascular disease monitoring and treatment planning. Recently, I explored optimized gradient waveform methods (GrOpt) developed at Stanford University to improve diffusion-weighting in cDTI images. These techniques enhance signal quality by reducing diffusion encoding time, achieving high performance while maintaining safety standards. I began implementing this approach using Pulseq, an open-source tool for designing MRI sequences. Since Pulseq was developed by the MR Technologies Group in Freiburg, this summer presented a unique opportunity to work directly with the creators of these tools, furthering my research and expanding the impact of my research.
Example images acquired using the sequence. A volunteer’s brain was scanned during sequence development to validate the protocol on the MRI system.
In Freiburg, I collected data using an ultra-high-performance MRI system (Cima.X, Siemens), one of the few such systems globally. Ultra-high performance MRI systems offer great potential for improving cDTI signal quality. However, these systems are often limited by suboptimal MRI sequence implementations. By using gradient optimization, I was able to fully leverage the hardware’s capabilities, and daily data acquisition allowed me to iterate quickly on my sequence development.
Building connections with fellow lab mates in Freiburg gave me the opportunity to explore new ideas and concepts. This exchange of ideas included correcting for “ghost” artifacts that appear in the image, online reconstruction of sequences with Pulseq, and modifications for magnetic field imperfections. I partnered with other students collecting cardiac MRI data allowing us to collectively test our protocols. Additionally, I participated in my colleagues ‘studies, volunteering for different MRI scans to assist others with their sequence development. Also, I had the opportunity to observe the decommissioning of an MRI system, which does not occur often. The clinic “quenched” the magnet of an older MRI system, in which all the helium is released from the MRI machine, eliminating the magnetic field. The department observed the large cloud of helium released.
The helium cloud from the controlled quench of an older MRI system at the hospital. The entire department gathered for the event. The helium cloud from releasing helium is shown in this image.
In early September, I shared my research with the broader MRI community at the International Society of Magnetic Resonance in Medicine (DACH-ISMRM) conference in Tübingen, Germany by giving an oral presentation at the meeting. It was wonderful to connect with researchers and students attending the conference from other German-speaking cities in Europe. It was a great way to share how utilizing both GrOpt and Pulseq together can leverage ultra-high-performance systems.
Presenting my research at the German-Speaking Chapter of the International Society of Magnetic Resonance in Medicine (DACH-ISMRM) conference in Tübingen, Germany. Here, researchers in Germany, Austria, and Switzerland present their research in English.
My summer in Freiburg provided me with an incredible opportunity to grow as an independent researcher and individual. As this was the longest time I spent living outside of the US, I explored Freiburg and immersed myself in the German culture. Every day, I enjoyed traditional German meals served at the Casino (hospital cafeteria) with the MR Technologies Group. I attended local events in Freiburg, including weekly markets, live music, and the Freiburg Wine Festival. I particularly enjoyed a department German barbecue with fellow trainees, celebrating the beautiful summer weather in the local Seepark. I lived in a shared apartment with a student studying physical therapy at the University, which allowed me to learn new recipes and explore areas of Germany I might not have discovered on my own, including my roommate’s hometown near Überlingen, on Lake Constance. In addition, I visited a former student from the lab at Stanford University in Nuremburg and Erlangen, where we tandem biked through villages and traveled to Munich with a colleague from Freiburg.
Münstermarkt takes place weekly in the center of Freiburg. Fresh fruits, vegetables, dairy, baked goods, and meats are sold there. A recommended treat that I tried was Stefans Käsekuchen, a delicious cheesecake.
I am continuing this research at Stanford University and look forward to collaborating with the Freiburg team to submit our work to the upcoming ISMRM annual meeting and co-author a manuscript. We are discussing my likely return to Freiburg early next year to collect additional data.
I am deeply grateful for this experience and excited to continue the collaboration that has been established at Freiburg.