NMNikolai J. Mickevicius

Nikolai J. Mickevicius

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Nikolai J. Mickevicius has a total of 28 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Variable Resolution Sampling and Deep Learning Image Recovery for Accelerated Multi-Spectral MRI Near Metal Implants

    by Azadeh Sharafi, Nikolai J. Mickevicius, Mehran Baboli, Andrew S. Nencka, Kevin M. Koch

    Purpose: This study presents a variable resolution (VR) sampling and deep learning reconstruction approach for multi-spectral MRI near metal implants, aiming to reduce scan times while maintaining image quality. Background: The rising use of metal implants has increased MRI scans affected by metal artifacts. Multi-spectral imaging (MSI) reduces these artifacts but sacrifices acquisition efficiency. Methods: This retrospective study on 1.5T MSI knee and hip data from patients with metal hardware used a novel spectral undersampling scheme to improve acquisition efficiency by ~40%. U-Net-based deep learning models were trained for reconstruction. Image quality was evaluated using SSIM, PSNR, and RESI metrics. Results: Deep learning reconstructions of undersampled VR data (DL-VR) showed significantly higher SSIM and PSNR values (p<0.001) compared to conventional reconstruction (CR-VR), with improved edge sharpness. Edge sharpness in DL-reconstructed images matched fully sampled references (p=0.5). Conclusion: This approach can potentially enhance MRI examinations near metal implants by reducing scan times or enabling higher resolution. Further prospective studies are needed to assess clinical value.

  • Open Access

    Variable Resolution Sampling and Deep Learning-Based Image Recovery for Faster Multi-Spectral Imaging Near Metal Implants

    by Nikolai J. Mickevicius, Azadeh Sharafi, Andrew S. Nencka, Kevin M. Koch

    Purpose: In multi-spectral imaging (MSI), several fast spin echo volumes with discrete Larmor frequency offsets are acquired in an interleaved fashion with multiple concatenations. Here, a variable resolution (VR) method to nearly halve scan time is proposed by only acquiring low resolution autocalibrating signal in half of the concatenations. Methods: Knee MSI datasets were retrospectively undersampled with the proposed variable resolution sampling scheme. A U-Net model was trained to predict the full-resolution images from the VR input. Image quality was assessed in 10 test subjects. Results: Spectral bin-combined images produced with the proposed variable resolution sampling with deep learning reconstruction appear to be of high quality and exhibited a median structural image similarity of 0.984 across test subjects and slices. Conclusion: The proposed variable resolution sampling method shows promise for drastically reducing the time it takes to collect multi-spectral imaging data near metallic implants. Further studies will rigorously examine its clinical utility across multiple implant scenarios.

  • Open Access

    Low‐rank inversion reconstruction for through‐plane accelerated radial MR fingerprinting applied to relaxometry at 0.35 T

    by Nikolai J. Mickevicius, Carri K. Glide‐Hurst

    Purpose: To reduce scan time, methods to accelerate phase‐encoded/non‐Cartesian MR fingerprinting (MRF) acquisitions for variable density spiral acquisitions have recently been developed. These methods are not applicable to MRF acquisitions, wherein a single k‐space spoke is acquired per frame. Therefore, we propose a low‐rank inversion method to resolve MRF contrast dynamics from through‐plane accelerated Cartesian/radial measurements applied to quantitative relaxation‐time mapping on a 0.35T system.Methods: An algorithm was implemented to reconstruct through‐plane aliased low‐rank images describing the contrast dynamics occurring because of the transient‐state MRF acquisition. T1 and T2 times from accelerated acquisitions were compared with those from unaccelerated linear reconstructions in a standardized system phantom and within in vivo brain and prostate experiments on a hybrid 0.35T MRI/linear accelerator.Results: No significant differences between T1 and T2 times for the accelerated reconstructions were observed compared to fully sampled acquisitions (p = 0.41 and p = 0.36, respectively). The mean absolute errors in T1 and T2 were 5.6% and 2.9%, respectively, between the full and accelerated acquisitions. The SDs in T1 and T2 decreased with the advanced accelerated reconstruction compared with the unaccelerated reconstruction (p = 0.02 and p = 0.03, respectively). The quality of the T1 and T2 maps generated with the proposed approach are comparable to those obtained using the unaccelerated data sets.Conclusions: Through‐plane accelerated MRF with radial k‐space coverage was demonstrated at a low field strength of 0.35 T. This method enabled 3D T1 and T2 mapping at 0.35 T with a 3‐min scan.


  • Split‐slice training and hyperparameter tuning of RAKI networks for simultaneous multi‐slice reconstruction

    by Andrew S. Nencka, Volkan E. Arpinar, Sampada Bhave, Baolian Yang, Suchandrima Banerjee, Michael McCrea, Nikolai J. Mickevicius, L. Tugan Muftuler, Kevin M. Koch

    Purpose: Simultaneous multi‐slice acquisitions are essential for modern neuroimaging research, enabling high temporal resolution functional and high‐resolution q‐space sampling diffusion acquisitions. Recently, deep learning reconstruction techniques have been introduced for unaliasing these accelerated acquisitions, and robust artificial‐neural‐networks for k‐space interpolation (RAKI) have shown promising capabilities. This study systematically examines the impacts of hyperparameter selections for RAKI networks, and introduces a novel technique for training data generation which is analogous to the split‐slice formalism used in slice‐GRAPPA.Methods: RAKI networks were developed with variable hyperparameters and with and without split‐slice training data generation. Each network was trained and applied to five different datasets including acquisitions harmonized with Human Connectome Project lifespan protocol. Unaliasing performance was assessed through L1 errors computed between unaliased and calibration frequency‐space data.Results: Split‐slice training significantly improved network performance in nearly all hyperparameter configurations. Best unaliasing results were achieved with three layer RAKI networks using at least 64 convolutional filters with receptive fields of 7 voxels, 128 single‐voxel filters in the penultimate RAKI layer, batch normalization, and no training dropout with the split‐slice augmented training dataset. Networks trained without the split‐slice technique showed symptoms of network over‐fitting.Conclusions: Split‐slice training for simultaneous multi‐slice RAKI networks positively impacts network performance. Hyperparameter tuning of such reconstruction networks can lead to further improvements in unaliasing performance.


  • Generalized simultaneous multi‐orientation 2D imaging

    by Nikolai J. Mickevicius, Andrew S. Nencka, Eric S. Paulson

    Purpose: Flexibility in slice prescription is critical for precise motion monitoring during MR‐guided therapies. Adding more slices to improve spatial coverage during rapid 2D cine imaging often hampers temporal resolution. This work describes a framework to simultaneously acquire multiple arbitrarily oriented slices which share a common frequency encoding axis. This framework allows for higher frame rates for a given number of slices compared to conventional interleaved‐slice multi‐orientation cine imaging.Theory and Methods: A framework to calculate zeroth gradient moments to be played out between sequentially excited slices with multiple orientations is described here. Experiments were performed in phantom, and in vivo in the head/neck and abdomen of patients.Results: Images arbitrarily rotated relative to one another were successfully obtained in phantom and in vivo. Simultaneous multi‐orientation (SMO) images were also acquired with additional in‐plane acceleration to demonstrate the capability of this method to rapidly image objects moving with physiological motion.Conclusions: The technical feasibility of the generalized SMO imaging framework was tested in this study. It shows promise for continued development for motion monitoring during MR‐guided therapies.