Hongki Lim

Prof. Hongki Lim

Electronic Engineering, Inha University, Incheon, Republic of Korea
orcid

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
29 July 2020 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
ICIl Yong Chun
Il Yong Chun
ZHZhengyu Huang
Zhengyu Huang
HLHongki Lim
Hongki Lim
JFJeffrey A. Fessler
Jeffrey A. Fessler
Iterative neural networks (INN) are rapidly gaining attention for solving inverse problems in imaging, image processing, and computer vision. INNs combine regression NNs and an iterative model-based image reconstruction (MBIR) algorithm, often leading to both good generalization capability and outperforming reconstruction quality over existing MBIR optimization models. This paper proposes the first fast and convergent INN architecture, Momentum-Net , by generalizing a block-wise MBIR algorithm that uses momentum and majorizers with regression NNs. For fast MBIR, Momentum-Net uses momentum terms in extrapolation modules, and noniterative MBIR modules at each iteration by using majorizers , where each iteration of Momentum-Net consists of three core modules: image refining, extrapolation, and MBIR. Momentum-Net guarantees convergence to a fixed-point for general differentiable (non)convex MBIR functions (or data-fit terms) and convex feasible sets, under two asymptomatic conditions. To consider data-fit variations across training and testing samples, we also propose a regularization parameter selection scheme based on the “spectral spread” of majorization matrices. Numerical experiments for light-field photography using a focal stack and sparse-view computational tomography demonstrate that, given identical regression NN architectures, Momentum-Net significantly improves MBIR speed and accuracy over several existing INNs; it significantly improves reconstruction quality compared to a state-of-the-art MBIR method in each application.

28 October 2020 in IEEE Transactions on Medical Imaging
Verified
Journal Article
Closed access
HLHongki Lim
Hongki Lim
ICIl Yong Chun
Il Yong Chun
YDYuni K. Dewaraja
Yuni K. Dewaraja
JFJeffrey A. Fessler
Jeffrey A. Fessler
Image reconstruction in low-count PET is particularly challenging because gammas from natural radioactivity in Lu-based crystals cause high random fractions that lower the measurement signal-to-noise-ratio (SNR). In model-based image reconstruction (MBIR), using more iterations of an unregularized method may increase the noise, so incorporating regularization into the image reconstruction is desirable to control the noise. New regularization methods based on learned convolutional operators are emerging in MBIR. We modify the architecture of an iterative neural network, BCD-Net, for PET MBIR, and demonstrate the efficacy of the trained BCD-Net using XCAT phantom data that simulates the low true coincidence count-rates with high random fractions typical for Y-90 PET patient imaging after Y-90 microsphere radioembolization. Numerical results show that the proposed BCD-Net significantly improves CNR and RMSE of the reconstructed images compared to MBIR methods using non-trained regularizers, total variation (TV) and non-local means (NLM). Moreover, BCD-Net successfully generalizes to test data that differs from the training data. Improvements were also demonstrated for the clinically relevant phantom measurement data where we used training and testing datasets having very different activity distributions and count-levels.

15 May 2020 in European Journal of Nuclear Medicine and Molecular Imaging
Verified
Journal Article
Closed access
PDF file
HXHaowei Xiang
Haowei Xiang
HLHongki Lim
Hongki Lim
JFJeffrey A. Fessler
Jeffrey A. Fessler
YDYuni K Dewaraja
Yuni K Dewaraja

1 February 2018 in Physics in Medicine & Biology
Verified
Journal Article
Open access
PDF file
HLHongki Lim
Hongki Lim
YDYuni K Dewaraja
Yuni K Dewaraja
JFJeffrey A Fessler
Jeffrey A Fessler
Most existing PET image reconstruction methods impose a nonnegativity constraint in the image domain that is natural physically, but can lead to biased reconstructions. This bias is particularly problematic for Y-90 PET because of the low probability positron production and high random coincidence fraction. This paper investigates a new PET reconstruction formulation that enforces nonnegativity of the projections instead of the voxel values. This formulation allows some negative voxel values, thereby potentially reducing bias. Unlike the previously reported NEG-ML approach that modifies the Poisson log-likelihood to allow negative values, the new formulation retains the classical Poisson statistical model. To relax the non-negativity constraint embedded in the standard methods for PET reconstruction, we used an alternating direction method of multipliers (ADMM). Because choice of ADMM parameters can greatly influence convergence rate, we applied an automatic parameter selection method to improve the convergence speed. We investigated the methods using lung to liver slices of XCAT phantom. We simulated low true coincidence count-rates with high random fractions corresponding to the typical values from patient imaging in Y-90 microsphere radioembolization. We compared our new methods with standard reconstruction algorithms and NEG-ML and a regularized version thereof. Both our new method and NEG-ML allow more accurate quantification in all volumes of interest while yielding lower noise than the standard method. The performance of NEG-ML can degrade when its user-defined parameter is tuned poorly, while the proposed algorithm is robust to any count level without requiring parameter tuning.

22 May 2018 in Physics in Medicine & Biology
Verified
Journal Article
Closed access
PDF file
HLHongki Lim
Hongki Lim
JFJeffrey A Fessler
Jeffrey A Fessler
SWScott J Wilderman
Scott J Wilderman
ABAllen Fredrick Brooks
Allen Fredrick Brooks
YDYuni K Dewaraja
Yuni K Dewaraja
While the yield of positrons used in Y-90 PET is independent of tissue media, Y-90 SPECT imaging is complicated by the tissue dependence of bremsstrahlung photon generation. The probability of bremsstrahlung production is proportional to the square of the atomic number of the medium. Hence, the same amount of activity in different tissue regions of the body will produce different numbers of bremsstrahlung photons. Existing reconstruction methods disregard this tissue-dependency, potentially impacting both qualitative and quantitative imaging of heterogeneous regions of the body such as bone with marrow cavities. In this proof-of-concept study, we propose a new maximum-likelihood method that incorporates bremsstrahlung generation probabilities into the system matrix, enabling images of the desired Y-90 distribution to be reconstructed instead of the ‘bremsstrahlung distribution’ that is obtained with existing methods. The tissue-dependent probabilities are generated by Monte Carlo simulation while bone volume fractions for each SPECT voxel are obtained from co-registered CT. First, we demonstrate the tissue dependency in a SPECT/CT imaging experiment with Y-90 in bone equivalent solution and water. Visually, the proposed reconstruction approach better matched the true image and the Y-90 PET image than the standard bremsstrahlung reconstruction approach. An XCAT phantom simulation including bone and marrow regions also demonstrated better agreement with the true image using the proposed reconstruction method. Quantitatively, compared with the standard reconstruction, the new method improved estimation of the liquid bone:water activity concentration ratio by 40% in the SPECT measurement and the cortical bone:marrow activity concentration ratio by 58% in the XCAT simulation.

25 September 2023 in IEEE Transactions on Computational Imaging
Verified
Journal Article
Closed access
HLHongki Lim
Hongki Lim
YDYuni K. Dewaraja
Yuni K. Dewaraja
JFJeffrey A. Fessler
Jeffrey A. Fessler
Improving low-count SPECT can shorten scans and support pre-therapy theranostic imaging for dosimetry-based treatment planning, especially with radionuclides like $^{177}$ Lu known for low photon yields. Conventional methods often underperform in low-count settings, highlighting the need for trained regularization in model-based image reconstruction. This article introduces a trained regularizer for SPECT reconstruction that leverages segmentation based on CT imaging. The regularizer incorporates CT-side information via a segmentation mask from a pre-trained network (nnUNet). In this proof-of-concept study, we used patient studies with $^{177}$ Lu DOTATATE to train and tested with phantom and patient datasets, simulating pre-therapy imaging conditions. Our results show that the proposed method outperforms both standard unregularized EM algorithms and conventional regularization with CT-side information. Specifically, our method achieved marked improvements in activity quantification, noise reduction, and root mean square error. The enhanced low-count SPECT approach has promising implications for theranostic imaging, post-therapy imaging, whole body SPECT, and reducing SPECT acquisition times.