Research Areas

My research topics of interest, past and present.

Topic List


Intrinsic Properties of Training Data and their Effect on Neural Network Generalization Ability

Research exploring how geometric and statistical properties of datasets, such as intrinsic dimension and label sharpness, influence neural network generalization across different imaging domains, particularly in how these differ betweeen natural and medical image models.

2024

  1. sciforDL24.jpg
    Pre-processing and Compression: Understanding Hidden Representation Refinement Across Imaging Domains via Intrinsic Dimension
    Nicholas Konz, and Maciej A. Mazurowski
    SciForDL @ NeurIPS, 2024
  2. iclr24.jpg
    The Effect of Intrinsic Dataset Properties on Generalization: Unraveling Learning Differences Between Natural and Medical Images
    Nicholas Konz, and Maciej A. Mazurowski
    ICLR, 2024

2022

  1. miccai22.jpg
    The Intrinsic Manifolds of Radiological Images and their Role in Deep Learning
    Nicholas Konz, Hanxue Gu, Haoyu Dong, and 1 more author
    MICCAI, 2022

Image Distribution Similarity Metrics and Generative Models

Development of novel perceptual metrics and generative models for medical imaging, including controllable image generation methods and specialized distance measures that better capture anatomical features than traditional computer vision metrics.

2025

  1. frd.jpg
    Fréchet Radiomic Distance (FRD): A Versatile Metric for Comparing Medical Imaging Datasets
    Nicholas Konz, Richard Osuala, Preeti Verma, and 16 more authors
    arXiv preprint, 2025

2024

  1. segdiff.jpg
    Anatomically-Controllable Medical Image Generation with Segmentation-Guided Diffusion Models
    Nicholas Konz, Yuwen Chen, Haoyu Dong, and 1 more author
    MICCAI, 2024
  2. midl24.jpg
    Rethinking Perceptual Metrics for Medical Image Translation
    Nicholas Konz, Yuwen Chen, Hanxue Gu, and 2 more authors
    MIDL, 2024
  3. contourdiff_pipeline.jpg
    ContourDiff: Unpaired Image-to-Image Translation with Structural Consistency for Medical Imaging
    Yuwen Chen, Nicholas Konz, Hanxue Gu, and 5 more authors
    arXiv preprint, 2024

Vision Foundation Models for Segmentation and Beyond

Investigation of the capabilities and limitations of large-scale foundation models for medical image analysis, including enhanced architectures and performance evaluation across diverse downstream tasks.

2025

  1. pipeline-for-breast-reg.jpg
    Are Vision Foundation Models Ready for Out-of-the-Box Medical Image Registration?
    Hanxue Gu*, Yaqian Chen*, Nicholas Konz, and 2 more authors
    Deep-Breath @ MICCAI, 2025
  2. segmentanymuscle.jpg
    SegmentAnyMuscle: A universal muscle segmentation model across different locations in MRI
    Roy Colglazier, Jisoo Lee, Haoyu Dong, and 8 more authors
    arXiv preprint, 2025
  3. mri-core.jpg
    MRI-CORE: A Foundation Model for Magnetic Resonance Imaging
    Haoyu Dong, Yuwen Chen, Hanxue Gu, and 4 more authors
    arXiv preprint, 2025
  4. slmsam-pipeline.jpg
    Accelerating Volumetric Medical Image Annotation via Short-Long Memory SAM 2
    Yuwen Chen, Zafer Yildiz, Qihang Li, and 5 more authors
    arXiv preprint, 2025

2024

  1. samfailure.jpg
    Quantifying the Limits of Segmentation Foundation Models: Modeling Challenges in Segmenting Tree-Like and Low-Contrast Objects
    Yixin Zhang*, Nicholas Konz*, Kevin Kramer, and 1 more author
    arXiv preprint, 2024

2023

  1. sam23.jpg
    Segment anything model for medical image analysis: an experimental study
    Maciej A. Mazurowski, Haoyu Dong, Hanxue Gu, and 3 more authors
    Medical Image Analysis, 2023

Neural Network Interpretability and Explainability

Methods for understanding what neural networks learn and how concepts are formed, including data attribution techniques and representation analysis to improve model transparency and trustworthiness, and others.

2023

  1. attrib.jpg
    Attributing Learned Concepts in Neural Networks to Training Data
    Nicholas Konz, Charles Godfrey, Madelyn Shapiro, and 3 more authors
    ATTRIB @ NeurIPS (Oral), 2023
  2. emnlp23.jpg
    Understanding the Inner-workings of Language Models Through Representation Dissimilarity
    Davis Brown, Charles Godfrey, Nicholas Konz, and 2 more authors
    EMNLP, 2023
  3. imbalance_illustrate.jpg
    A systematic study of the foreground-background imbalance problem in deep learning for object detection
    Hanxue Gu, Haoyu Dong, Nicholas Konz, and 1 more author
    arXiv preprint, 2023

Domain Adaptation and Analysis

Techniques for adapting models across different imaging domains, scanners, and acquisition parameters, addressing the challenge of domain shift that commonly affects medical AI systems in practice, as well as understanding the nature of the domain shift problem itself.

2024

  1. intent.jpg
    Medical Image Segmentation with InTEnt: Integrated Entropy Weighting for Single Image Test-Time Adaptation
    Haoyu Dong, Nicholas Konz, Hanxue Gu, and 1 more author
    DEF-AI-MIA @ CVPR (Oral), 2024
  2. performancesurvey.jpg
    The impact of scanner domain shift on deep learning performance in medical imaging: an experimental study
    Brian Guo, Darui Lu, Gregory Szumel, and 4 more authors
    arXiv preprint, 2024

2023

  1. midl23.jpg
    Reverse engineering breast mris: Predicting acquisition parameters directly from images
    Nicholas Konz, and Maciej A. Mazurowski
    MIDL, 2023
  2. star_style_more_loss.jpg
    Deep learning for breast mri style transfer with limited training data
    Shixing Cao, Nicholas Konz, James Duncan, and 1 more author
    Journal of Digital imaging, 2023

Anomaly Detection and Localization

Unsupervised and self-supervised approaches for detecting abnormalities in medical images, with applications ranging from breast cancer screening to general outlier detection in high-resolution imaging data.

2023

  1. picard.jpg
    Unsupervised anomaly localization in high-resolution breast scans using deep pluralistic image completion
    Nicholas Konz, Haoyu Dong, and Maciej A. Mazurowski
    Medical Image Analysis, 2023
  2. swssl.jpg
    SWSSL: Sliding window-based self-supervised learning for anomaly detection in high-resolution images
    Haoyu Dong, Yifan Zhang, Hanxue Gu, and 3 more authors
    IEEE Transactions on Medical Imaging, 2023

2021

  1. swiecicki.jpg
    A generative adversarial network-based abnormality detection using only normal images for model training with application to digital breast tomosynthesis
    Albert Swiecicki, Nicholas Konz, Mateusz Buda, and 1 more author
    Scientific reports, 2021

2018

  1. rcr.jpg
    Robust chauvenet outlier rejection
    MP Maples, DE Reichart, Nicholas Konz, and 7 more authors
    The Astrophysical Journal Supplement Series, 2018

Misc. Breast Imaging Analysis

Specialized methods for breast imaging applications, including lesion detection algorithms, registration techniques, and style transfer methods tailored to the unique challenges of breast MRI and tomosynthesis.

2025

  1. guidedmorph.jpg
    GuidedMorph: Two-Stage Deformable Registration for Breast MRI
    Yaqian Chen, Hanxue Gu, Haoyu Dong, and 5 more authors
    arXiv preprint, 2025

2023

  1. dbtex.jpg
    A competition, benchmark, code, and data for using artificial intelligence to detect lesions in digital breast tomosynthesis
    Nicholas Konz, Mateusz Buda, Hanxue Gu, and 8 more authors
    JAMA Network Open, 2023

2022

  1. dbtcaption.jpg
    Lightweight transformer backbone for medical object detection
    Yifan Zhang, Haoyu Dong, Nicholas Konz, and 2 more authors
    In MICCAI Workshop on Cancer Prevention through Early Detection, 2022