For me, these challenges are worth the upside of being an academic. I know that it is not the case for many, and I respect that, but I love it!
With all this said, it's also important to celebrate the rare successes like this one. I'm looking forward to all the work with this great team!
All that work that went into preparing the proposal, and after the rejection, it felt like the world was as if (more or less) none of that work had happened. It can definitely generate a feeling of futility. But you have to take these failures as a part of the path that (hopefully) leads to success.
At this point, I cannot count the times I have had my grants rejected! And many times, especially earlier in my career, it was very discouraging.
Sadly, I hear that more folks are getting such a feeling when looking at social media. So ...
This is the third NIH R01 grant that I have received in my career (previous two as the PI), and I'm very grateful for that. But along with the successes, I also want to share the failures to avoid the fake picture that if you're getting rejected, you're not good at your job.
I will serve as the MPI (multiple PI) for the grant, along with my colleague Ben Wildman-Tobriner, a Radiologist here at Duke. The funding is provided by the National Cancer Institute.
Within this project, we will use deep learning to better diagnose and treat thyroid cancers. It's a multi-institutional collaboration between Duke, Stanford, UCSF, Penn, and UC Davis.
Happy to share that we got an NIH R01 grant!
Congrats to Nick Konz for defending his PhD dissertation!
Nick has done an amazing job developing a variety of machine learning algorithms in the context of breast imaging. Nick's next step will be a postdoc at UNC Chapel Hill, where he will continue working on machine learning.
Thank you to the many researchers who contributed to the creation of this dataset! Duke Spark: AI in Medical Imaging
Let us know what you think!
Modality: Magnetic Resonance Imaging (MRI)
Location: Cervical Spine
Number of Patients: 1,232
Annotations: segmentation masks of vertebral bodies and intervertebral discs
for 481 patients
Paper: www.nature.com/articles/s4...
Download data: data.midrc.org/discovery/H... (you have to be logged in)
NEW PUBLIC DATASET ALERT!
Just published in Nature Scientific Data.
We're happy to publicly release another medical imaging dataset: Duke University Cervical Spine MRI Segmentation Dataset (CSpineSeg). Here are some details:
Here is the paper: arxiv.org/pdf/2507.11569?
Congrats to Hanxue Gu, Yaqian Chen, and co-authors for receiving the best paper award at the MICCAI Deep Breath 2025 workshop!
The paper discusses the use of foundation models in the context of image registration.
Paper: raw.githubusercontent.com/mlresearch/...
Congrats to Hanxue Gu, who is the first author, and the interdisciplinary team of co-authors!
Our method:
- automatically segments radius and ulna bones
- uses a pose estimation network to assess rotational parameters of the bones
- automatically detects fracture locations
- combines all the information to infer the 3D fracture angles
The paper has been published at MIDL.
We propose a deep learning-based method that allows for measuring 3D angles from standard non-orthogonal planar X-rays, which allows for patient movement between the images are acquired.
Precise 3D measurement of fracture angles would be of enormous help in orthopedics, and yet it's very challenging from standard X-rays. We have a solution!
Our method:
- automatically segments radius and ulna bones
- uses a pose estimation network to assess rotational parameters of the bones
- automatically detects fracture locations
- combines all the information to infer the 3D fracture angles
The paper has been published at MIDL.
We propose a deep learning-based method that allows for measuring 3D angles from standard non-orthogonal planar X-rays, which allows for patient movement between the images are acquired.
We addressed this by using contours from the image to guide the diffusion model and showed quite a good performance of the model!
Congrats to Yuwen Chen, who is the first author, and the other team members!
The issue for such translation is that for a given body part, the CT and MRI images often have a different field of view, resulting in different structures being portrayed in the image.
Want to make a CT out of an MRI? It's possible thanks to generative models, but it has issues which we're addressing in our ContourDiff model (code available)!
Code: github.com/mazurowski-...
Paper:
openaccess.thecvf.com/content/CVP...
- we explored different ways of integrating adapted models
- we validated our method with 24 source domain-target domain splits for 3 medical imaging datasets
- our method outperforms SOTA by 2.9% on average in terms of Dice similarity coefficient
- published in a CVPR workshop
Segmentation models may perform poorly when test images belong to a different domain (e.g., a different medical center). We developed a method of adapting the models using a single unlabeled image from the test domain!
Congrats to Yuwen Chen, the lead author of the paper for this terrific work!
