1 code implementation • 24 Oct 2023 • Minh Nguyen, Alan Q. Wang, Heejong Kim, Mert R. Sabuncu
CoPA assumes that (1) generation mechanism is stable, i. e. label Y and confounding variable(s) Z generate X, and (2) the unstable conditional prevalence in each site E fully accounts for the unstable correlations between X and Y .
no code implementations • 21 Oct 2023 • Minh Nguyen, Gia H. Ngo, Mert R. Sabuncu
Given sufficient pairs of resting-state and task-evoked fMRI scans from subjects, it is possible to train ML models to predict subject-specific task-evoked activity using resting-state functional MRI (rsfMRI) scans.
1 code implementation • 5 Oct 2023 • Cagla Deniz Bahadir, Benjamin Liechty, David J. Pisapia, Mert R. Sabuncu
In this paper, our goal is to shed light on the inherent uncertainty of mitosis labels and characterize the mitotic figure classification task in a human interpretable manner.
no code implementations • 2 Oct 2023 • Alan Q. Wang, Batuhan K. Karaman, Heejong Kim, Jacob Rosenthal, Rachit Saluja, Sean I. Young, Mert R. Sabuncu
To answer these questions, we identify a need to formalize the goals and elements of interpretability in MLMI.
no code implementations • 6 Jul 2023 • Heejong Kim, Victor Ion Butoi, Adrian V. Dalca, Daniel J. A. Margolis, Mert R. Sabuncu
Most state-of-the-art techniques for medical image segmentation rely on deep-learning models.
no code implementations • 10 May 2023 • Suraj Rajendran, Weishen Pan, Mert R. Sabuncu, Yong Chen, Jiayu Zhou, Fei Wang
By offering a more comprehensive approach to healthcare data integration, patchwork learning has the potential to revolutionize the clinical applicability of ML models.
2 code implementations • 19 Apr 2023 • Alan Q. Wang, Evan M. Yu, Adrian V. Dalca, Mert R. Sabuncu
Our core insight which addresses these shortcomings is that corresponding keypoints between images can be used to obtain the optimal transformation via a differentiable closed-form expression.
no code implementations • 18 Apr 2023 • Zijin Gu, Keith Jamison, Mert R. Sabuncu, Amy Kuceyeski
Furthermore, aTLfaces and FBA1 had higher activation in response to maximal synthetic images compared to maximal natural images.
1 code implementation • ICCV 2023 • Victor Ion Butoi, Jose Javier Gonzalez Ortiz, Tianyu Ma, Mert R. Sabuncu, John Guttag, Adrian V. Dalca
We present UniverSeg, a method for solving unseen medical segmentation tasks without additional training.
1 code implementation • 5 Apr 2023 • Heejong Kim, Mert R. Sabuncu
For example, normalizing for nuisance variation might be hard in settings where there are a lot of idiosyncratic changes.
1 code implementation • 21 Mar 2023 • Xinzi He, Alan Wang, Mert R. Sabuncu
Head MRI pre-processing involves converting raw images to an intensity-normalized, skull-stripped brain in a standard coordinate space.
Ranked #1 on Reconstruction on PPMI
1 code implementation • 7 Dec 2022 • Alan Q. Wang, Mert R. Sabuncu
The first is the label weights, and the second is our novel concept of the ``support influence function,'' which is an easy-to-compute metric that quantifies the influence of a support element on the prediction for a given query.
1 code implementation • 1 Nov 2022 • Jinwei Zhang, Pascal Spincemaille, Hang Zhang, Thanh D. Nguyen, Chao Li, Jiahao Li, Ilhami Kovanlikaya, Mert R. Sabuncu, Yi Wang
In this paper, we present our new framework, called Learned Acquisition and Reconstruction Optimization (LARO), which aims to accelerate the multi-echo gradient echo (mGRE) pulse sequence for QSM.
no code implementations • 13 Oct 2022 • Minh Nguyen, Gia H. Ngo, Mert R. Sabuncu
We call our approach GLACIAL, which stands for "Granger and LeArning-based CausalIty Analysis for Longitudinal studies."
no code implementations • 24 Jul 2022 • Gia H. Ngo, Minh Nguyen, Nancy F. Chen, Mert R. Sabuncu
In this paper, we present Text2Brain, an easy to use tool for synthesizing brain activation maps from open-ended text queries.
2 code implementations • 24 May 2022 • Richa Rastogi, Yair Schiff, Alon Hacohen, Zhaozhi Li, Ian Lee, Yuntian Deng, Mert R. Sabuncu, Volodymyr Kuleshov
We introduce semi-parametric inducing point networks (SPIN), a general-purpose architecture that can query the training set at inference time in a compute-efficient manner.
1 code implementation • 17 Mar 2022 • Tianyu Ma, Benjamin C. Lee, Mert R. Sabuncu
For segmentation tasks with multiple classes, the standard approach is to use a network that computes a multi-channel probabilistic segmentation map, with each channel representing one class.
2 code implementations • 22 Feb 2022 • Alan Q. Wang, Adrian V. Dalca, Mert R. Sabuncu
The typical approach is to train the model for a hyperparameter setting determined with some empirical or theoretical justification.
1 code implementation • 6 Feb 2022 • Tianyu Ma, Alan Q. Wang, Adrian V. Dalca, Mert R. Sabuncu
The key building block of a CNN is the convolutional kernel that aggregates information from the pixel neighborhood and shares weights across all pixels.
2 code implementations • 28 Sep 2021 • Gia H. Ngo, Minh Nguyen, Nancy F. Chen, Mert R. Sabuncu
In this work, we propose Text2Brain, a neural network approach for coordinate-based meta-analysis of neuroimaging studies to synthesize brain activation maps from open-ended text queries.
no code implementations • 9 Jun 2021 • Zhilu Zhang, Vianne R. Gao, Mert R. Sabuncu
We show that the proposed subnetwork ensembling method can perform as well as standard deep ensembles in both accuracy and uncertainty estimates, yet with a computational efficiency similar to MC dropout.
1 code implementation • IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2022 • Tianyu Ma, Adrian V. Dalca, Mert R. Sabuncu
In this paper, we propose a powerful novel building block, the hyper-convolution, which implicitly represents the convolution kernel as a function of kernel coordinates.
1 code implementation • 17 May 2021 • Alan Q. Wang, Aaron K. LaViolette, Leo Moon, Chris Xu, Mert R. Sabuncu
Compressed sensing fluorescence microscopy (CS-FM) proposes a scheme whereby less measurements are collected during sensing and reconstruction is performed to recover the image.
2 code implementations • 15 May 2021 • Zijin Gu, Keith W. Jamison, Meenakshi Khosla, Emily J. Allen, Yihan Wu, Thomas Naselaris, Kendrick Kay, Mert R. Sabuncu, Amy Kuceyeski
NeuroGen combines an fMRI-trained neural encoding model of human vision with a deep generative network to synthesize images predicted to achieve a target pattern of macro-scale brain activation.
2 code implementations • 6 Jan 2021 • Alan Q. Wang, Adrian V. Dalca, Mert R. Sabuncu
In this paper, we explore a novel strategy of using a hypernetwork to generate the parameters of a separate reconstruction network as a function of the regularization weight(s), resulting in a regularization-agnostic reconstruction model.
1 code implementation • 9 Oct 2020 • Adrian V. Dalca, Ramesh Sridharan, Mert R. Sabuncu, Polina Golland
We demonstrate prediction of follow-up anatomical scans in the ADNI cohort, and illustrate a novel analysis approach that compares a patient's scans to the predicted subject-specific healthy anatomical trajectory.
1 code implementation • NeurIPS 2020 • Meenakshi Khosla, Gia H. Ngo, Keith Jamison, Amy Kuceyeski, Mert R. Sabuncu
Using concurrent eye-tracking and functional Magnetic Resonance Imaging (fMRI) recordings from a large cohort of human subjects watching movies, we first demonstrate that leveraging gaze information, in the form of attentional masking, can significantly improve brain response prediction accuracy in a neural encoding model.
no code implementations • 17 Aug 2020 • Mert R. Sabuncu
Many fields view stochasticity as a way to gain computational efficiency, while often having to trade off accuracy.
2 code implementations • 7 Aug 2020 • Gia H. Ngo, Meenakshi Khosla, Keith Jamison, Amy Kuceyeski, Mert R. Sabuncu
Resting-state functional MRI (rsfMRI) yields functional connectomes that can serve as cognitive fingerprints of individuals.
1 code implementation • 7 Aug 2020 • Gia H. Ngo, Meenakshi Khosla, Keith Jamison, Amy Kuceyeski, Mert R. Sabuncu
Resting-state functional MRI (rsfMRI) yields functional connectomes that can serve as cognitive fingerprints of individuals.
no code implementations • 31 Jul 2020 • Yichen Shen, Zhilu Zhang, Mert R. Sabuncu, Lin Sun
We propose a simple, easy-to-optimize distillation method for learning the conditional predictive distribution of a pre-trained dropout model for fast, sample-free uncertainty estimation in computer vision tasks.
1 code implementation • 29 Jul 2020 • Alan Q. Wang, Adrian V. Dalca, Mert R. Sabuncu
In this paper, we explore a novel strategy to train an unrolled reconstruction network in an unsupervised fashion by adopting a loss function widely-used in classical optimization schemes.
1 code implementation • 29 Jun 2020 • Meenakshi Khosla, Gia H. Ngo, Keith Jamison, Amy Kuceyeski, Mert R. Sabuncu
The increasing popularity of naturalistic paradigms in fMRI (such as movie watching) demands novel strategies for multi-subject data analysis, such as use of neural encoding models.
1 code implementation • NeurIPS 2020 • Zhilu Zhang, Mert R. Sabuncu
It has been recently demonstrated that multi-generational self-distillation can improve generalization.
1 code implementation • MIDL 2019 • Evan M. Yu, Juan Eugenio Iglesias, Adrian V. Dalca, Mert R. Sabuncu
Thus there is a strong need for deep learning-based segmentation tools that do not require heavy supervision and can continuously adapt.
1 code implementation • IEEE 17th International Symposium on Biomedical Imaging (ISBI) 2020 • Tianyu Ma, Ajay Gupta, Mert R. Sabuncu
Deep neural networks yield promising results in a wide range of computer vision applications, including landmark detection.
1 code implementation • 3 Mar 2020 • Tianyu Ma, Ajay Gupta, Mert R. Sabuncu
Deep neural networks yield promising results in a wide range of computer vision applications, including landmark detection.
1 code implementation • 27 Feb 2020 • Hang Zhang, Jinwei Zhang, Qihao Zhang, Jeremy Kim, Shun Zhang, Susan A. Gauthier, Pascal Spincemaille, Thanh D. Nguyen, Mert R. Sabuncu, Yi Wang
Brain lesion volume measured on T2 weighted MRI images is a clinically important disease marker in multiple sclerosis (MS).
no code implementations • 25 Sep 2019 • Zhilu Zhang, Adrian V. Dalca, Mert R. Sabuncu
Motivated by this, we explore the use of various structured dropout techniques to promote model diversity and improve the quality of probabilistic predictions.
no code implementations • 16 Aug 2019 • Meenakshi Khosla, Keith Jamison, Amy Kuceyeski, Mert R. Sabuncu
Resting-state functional MRI (rs-fMRI) is a rich imaging modality that captures spontaneous brain activity patterns, revealing clues about the connectomic organization of the human brain.
no code implementations • 13 Aug 2019 • Hyeon Woo Lee, Mert R. Sabuncu, Adrian V. Dalca
We tackle biomedical image segmentation in the scenario of only a few labeled brain MR images.
1 code implementation • NeurIPS 2019 • Adrian V. Dalca, Marianne Rakic, John Guttag, Mert R. Sabuncu
We develop a learning framework for building deformable templates, which play a fundamental role in many image analysis and computational anatomy tasks.
1 code implementation • 26 Jul 2019 • Cagla D. Bahadir, Alan Q. Wang, Adrian V. Dalca, Mert R. Sabuncu
In our experiments, we demonstrate that LOUPE-optimized under-sampling masks are data-dependent, varying significantly with the imaged anatomy, and perform well with different reconstruction methods.
no code implementations • 23 Jun 2019 • Zhilu Zhang, Adrian V. Dalca, Mert R. Sabuncu
Motivated by this, we explore the use of structured dropout to promote model diversity and improve confidence calibration.
1 code implementation • 25 Apr 2019 • Adrian V. Dalca, Evan Yu, Polina Golland, Bruce Fischl, Mert R. Sabuncu, Juan Eugenio Iglesias
To develop a deep learning-based segmentation model for a new image dataset (e. g., of different contrast), one usually needs to create a new labeled training dataset, which can be prohibitively expensive, or rely on suboptimal ad hoc adaptation or augmentation approaches.
1 code implementation • 8 Mar 2019 • Adrian V. Dalca, Guha Balakrishnan, John Guttag, Mert R. Sabuncu
We present a probabilistic generative model and derive an unsupervised learning-based inference algorithm that uses insights from classical registration methods and makes use of recent developments in convolutional neural networks (CNNs).
Ranked #2 on Diffeomorphic Medical Image Registration on OASIS+ADIBE+ADHD200+MCIC+PPMI+HABS+HarvardGSP
Constrained Diffeomorphic Image Registration Deformable Medical Image Registration +2
6 code implementations • 8 Mar 2019 • Adrian V. Dalca, John Guttag, Mert R. Sabuncu
In this work, we introduce a general probabilistic model that describes sparse high dimensional imaging data as being generated by a deep non-linear embedding.
2 code implementations • CVPR 2018 • Adrian V. Dalca, John Guttag, Mert R. Sabuncu
The integration of anatomical priors can facilitate CNN-based anatomical segmentation in a range of novel clinical problems, where few or no annotations are available and thus standard networks are not trainable.
1 code implementation • 7 Jan 2019 • Cagla Deniz Bahadir, Adrian V. Dalca, Mert R. Sabuncu
Acquisition of Magnetic Resonance Imaging (MRI) scans can be accelerated by under-sampling in k-space (i. e., the Fourier domain).
no code implementations • 30 Dec 2018 • Meenakshi Khosla, Keith Jamison, Gia H. Ngo, Amy Kuceyeski, Mert R. Sabuncu
Here, we present an overview of various unsupervised and supervised machine learning applications to rs-fMRI.
1 code implementation • 17 Oct 2018 • Evan M. Yu, Mert R. Sabuncu
We propose a novel machine learning strategy for studying neuroanatomical shape variation.
6 code implementations • 14 Sep 2018 • Guha Balakrishnan, Amy Zhao, Mert R. Sabuncu, John Guttag, Adrian V. Dalca
In contrast to this approach, and building on recent learning-based methods, we formulate registration as a function that maps an input image pair to a deformation field that aligns these images.
Ranked #1 on Diffeomorphic Medical Image Registration on OASIS+ADIBE+ADHD200+MCIC+PPMI+HABS+HarvardGSP (Dice metric)
Deformable Medical Image Registration Diffeomorphic Medical Image Registration +1
1 code implementation • 11 Sep 2018 • Meenakshi Khosla, Keith Jamison, Amy Kuceyeski, Mert R. Sabuncu
The specificty and sensitivity of resting state functional MRI (rs-fMRI) measurements depend on pre-processing choices, such as the parcellation scheme used to define regions of interest (ROIs).
2 code implementations • 17 Aug 2018 • Adrian V. Dalca, Katherine L. Bouman, William T. Freeman, Natalia S. Rost, Mert R. Sabuncu, Polina Golland
We present an algorithm for creating high resolution anatomically plausible images consistent with acquired clinical brain MRI scans with large inter-slice spacing.
4 code implementations • NeurIPS 2018 • Zhilu Zhang, Mert R. Sabuncu
Here, we present a theoretically grounded set of noise-robust loss functions that can be seen as a generalization of MAE and CCE.
Ranked #19 on Learning with noisy labels on CIFAR-10N-Worst
2 code implementations • 11 May 2018 • Adrian V. Dalca, Guha Balakrishnan, John Guttag, Mert R. Sabuncu
We demonstrate our method on a 3D brain registration task, and provide an empirical analysis of the algorithm.
no code implementations • 8 May 2018 • Sundaresh Ram, Vicky T. Nguyen, Kirsten H. Limesand, Mert R. Sabuncu
Mirroring the co-dependency of these tasks, our proposed model consists of two serial components: the first part computes a segmentation of cell bodies, while the second module identifies the centers of these cells.
1 code implementation • 13 Mar 2018 • Yingying Zhu, Mert R. Sabuncu
An additional layer of complexity is that, in real life, the amount and type of data available for each patient can differ significantly.
3 code implementations • CVPR 2018 • Guha Balakrishnan, Amy Zhao, Mert R. Sabuncu, John Guttag, Adrian V. Dalca
We define registration as a parametric function, and optimize its parameters given a set of images from a collection of interest.
2 code implementations • 3 Jan 2018 • Mohammad Haft-Javaherian, Linjing Fang, Victorine Muse, Chris B. Schaffer, Nozomi Nishimura, Mert R. Sabuncu
The health and function of tissue rely on its vasculature network to provide reliable blood perfusion.