Ouwen Huang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ouwen Huang has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Mark L. Palmeri
Department of Biomedical Engineering, Duke University, Durham, NC, USA
3 shared publications Mark L. Palmeri
Department of Biomedical Engineering, Duke University, Durham, NC, USA
2 shared publications- Nick Bottenus
Department of Mechanical Engineering, University of Colorado at Boulder, Boulder, CO, USA
2 shared publications - Sina Farsiu
Duke University
2 shared publications - Will Long
Department of Biomedical Engineering, Duke University, Durham, USA
2 shared publications Corey Jensen
Department of Abdominal Imaging, MD Anderson Cancer Center, Houston, TX, USA
1 shared publication
Latest Publications
- 8 March 2024
- Published by Elsevier in Ultrasound in Medicine & Biology
- Vol. 50 (6) , 788-796
Automated Spontaneous Echo Contrast Detection Using a Multisequence Attention Convolutional Neural Network
- 8 March 2024
- Published by Elsevier in Ultrasound in Medicine & Biology
- Vol. 50 (6) , 788-796
Automated Spontaneous Echo Contrast Detection Using a Multisequence Attention Convolutional Neural Network
- 8 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Imaging
- Vol. 10 (25730436) , 461-468
TPU Based Deep Learning Image Enhancement for Real-Time Point-of-Care Ultrasound
Handheld ultrasound devices are becoming more prevalent in point-of-care ultrasound workflows. However, these devices are computationally constrained which challenges the advances in deep learning methodology for real-time use on mobile Point-of-care ultrasound (POCUS) devices. In this work, we explore the feasibility of running MimickNet, a deep learning clinical post-processing model, on Tensor Processing Units (TPUs), hardware designed for deep learning operations capable of running on only 2 watts of power at 1.8 V with a form factor of 10 mm x 15 mm. We show that real-time deep learning based post-processing is feasible at 20 - 120 FPS for 1472x160 to 224x224 axial sample x B-mode scan line configurations. We refer to the TPU based model as MimickNet Mobile. MimickNet Mobile achieves outputs nearly identical to the original MimickNet with a structural similarity index measurement (SSIM) of 0.98 $\pm$ 0.001 and a mean squared error (MSE) of 0.0001 $\pm$ 0.0 over our test set of 588 frames consisting of 168 phantom frames and 420 prospectively acquired human liver frames. We investigate the latency of other common mobile architectures such as separable convolution. Finally, we investigate the distribution of model parameter error when quantizing MimickNet $\mathtt{float32}$ weights to MimickNet Mobile $\mathtt{int8}$ weights. This work demonstrates that real-time POCUS deep learning image enhancement is feasible using TPUs. Future ultrasound device manufacturers can consider incorporating a TPU for the added flexibility of supporting several deep learning architectures without compromising on power management and form factor. We make code and models available at: https://github.com/ouwen/mimicknetmobile .
- 8 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computational Imaging
- Vol. 10 (25730436) , 461-468
TPU Based Deep Learning Image Enhancement for Real-Time Point-of-Care Ultrasound
Handheld ultrasound devices are becoming more prevalent in point-of-care ultrasound workflows. However, these devices are computationally constrained which challenges the advances in deep learning methodology for real-time use on mobile Point-of-care ultrasound (POCUS) devices. In this work, we explore the feasibility of running MimickNet, a deep learning clinical post-processing model, on Tensor Processing Units (TPUs), hardware designed for deep learning operations capable of running on only 2 watts of power at 1.8 V with a form factor of 10 mm x 15 mm. We show that real-time deep learning based post-processing is feasible at 20 - 120 FPS for 1472x160 to 224x224 axial sample x B-mode scan line configurations. We refer to the TPU based model as MimickNet Mobile. MimickNet Mobile achieves outputs nearly identical to the original MimickNet with a structural similarity index measurement (SSIM) of 0.98 $\pm$ 0.001 and a mean squared error (MSE) of 0.0001 $\pm$ 0.0 over our test set of 588 frames consisting of 168 phantom frames and 420 prospectively acquired human liver frames. We investigate the latency of other common mobile architectures such as separable convolution. Finally, we investigate the distribution of model parameter error when quantizing MimickNet $\mathtt{float32}$ weights to MimickNet Mobile $\mathtt{int8}$ weights. This work demonstrates that real-time POCUS deep learning image enhancement is feasible using TPUs. Future ultrasound device manufacturers can consider incorporating a TPU for the added flexibility of supporting several deep learning architectures without compromising on power management and form factor. We make code and models available at: https://github.com/ouwen/mimicknetmobile .
- 31 January 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 39 (6) , 2277-2286
MimickNet, Mimicking Clinical Image Post- Processing Under Black-Box Constraints
Image post-processing is used in clinical-grade ultrasound scanners to improve image quality (e.g., reduce speckle noise and enhance contrast). These post-processing techniques vary across manufacturers and are generally kept proprietary, which presents a challenge for researchers looking to match current clinical-grade workflows. We introduce a deep learning framework, MimickNet, that transforms conventional delay-and-summed (DAS) beams into the approximate Dynamic Tissue Contrast Enhanced (DTCE™) post-processed images found on Siemens clinical-grade scanners. Training MimickNet only requires post-processed image samples from a scanner of interest without the need for explicit pairing to DAS data. This flexibility allows MimickNet to hypothetically approximate any manufacturer's post-processing without access to the pre-processed data. MimickNet post-processing achieves a 0.940 ± 0.018 structural similarity index measurement (SSIM) compared to clinical-grade post-processing on a 400 cine-loop test set, 0.937 ± 0.025 SSIM on a prospectively acquired dataset, and 0.928 ± 0.003 SSIM on an out-of-distribution cardiac cine-loop after gain adjustment. To our knowledge, this is the first work to establish deep learning models that closely approximate ultrasound post-processing found in current medical practice. MimickNet serves as a clinical post-processing baseline for future works in ultrasound image formation to compare against. Additionally, it can be used as a pretrained model for fine-tuning towards different post-processing techniques. To this end, we have made the MimickNet software, phantom data, and permitted in vivo data open-source at https://github.com/ouwen/MimickNet.
- 31 January 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 39 (6) , 2277-2286
MimickNet, Mimicking Clinical Image Post- Processing Under Black-Box Constraints
Image post-processing is used in clinical-grade ultrasound scanners to improve image quality (e.g., reduce speckle noise and enhance contrast). These post-processing techniques vary across manufacturers and are generally kept proprietary, which presents a challenge for researchers looking to match current clinical-grade workflows. We introduce a deep learning framework, MimickNet, that transforms conventional delay-and-summed (DAS) beams into the approximate Dynamic Tissue Contrast Enhanced (DTCE™) post-processed images found on Siemens clinical-grade scanners. Training MimickNet only requires post-processed image samples from a scanner of interest without the need for explicit pairing to DAS data. This flexibility allows MimickNet to hypothetically approximate any manufacturer's post-processing without access to the pre-processed data. MimickNet post-processing achieves a 0.940 ± 0.018 structural similarity index measurement (SSIM) compared to clinical-grade post-processing on a 400 cine-loop test set, 0.937 ± 0.025 SSIM on a prospectively acquired dataset, and 0.928 ± 0.003 SSIM on an out-of-distribution cardiac cine-loop after gain adjustment. To our knowledge, this is the first work to establish deep learning models that closely approximate ultrasound post-processing found in current medical practice. MimickNet serves as a clinical post-processing baseline for future works in ultrasound image formation to compare against. Additionally, it can be used as a pretrained model for fine-tuning towards different post-processing techniques. To this end, we have made the MimickNet software, phantom data, and permitted in vivo data open-source at https://github.com/ouwen/MimickNet.
- Open Access
- 27 September 2019
- Published by Duke University Libraries
UltraDuke Version 1
DukeUltrasound is an ultrasound dataset collected at Duke University with a Verasonics c52v probe. It contains delay-and-sum (DAS) beamformed data as well as data post-processed with Siemens Dynam...
- 27 September 2019
- Published by Duke University Libraries
Open AccessUltraDuke Version 1
DukeUltrasound is an ultrasound dataset collected at Duke University with a Verasonics c52v probe. It contains delay-and-sum (DAS) beamformed data as well as data post-processed with Siemens Dynam...
- Open Access
MimickNet, Matching Clinical Post-Processing Under Realistic Black-Box Constraints
Image post-processing is used in clinical-grade ultrasound scanners to improve image quality (e.g., reduce speckle noise and enhance contrast). These post-processing techniques vary across manufacturers and are generally kept proprietary, which presents a challenge for researchers looking to match current clinical-grade workflows. We introduce a deep learning framework, MimickNet, that transforms raw conventional delay-and-summed (DAS) beams into the approximate post-processed images found on clinical-grade scanners. Training MimickNet only requires post-processed image samples from a scanner of interest without the need for explicit pairing to raw DAS data. This flexibility allows it to hypothetically approximate any manufacturer's post-processing without access to the pre-processed data. MimickNet generates images with an average similarity index measurement (SSIM) of 0.930$\pm$0.0892 on a 300 cineloop test set, and it generalizes to cardiac cineloops outside of our train-test distribution achieving an SSIM of 0.967$\pm$0.002. We also explore the theoretical SSIM achievable by evaluating MimickNet performance when trained under gray-box constraints (i.e., when both pre-processed and post-processed images are available). To our knowledge, this is the first work to establish deep learning models that closely approximate current clinical-grade ultrasound post-processing under realistic black-box constraints where before and after post-processing data is unavailable. MimickNet serves as a clinical post-processing baseline for future works in ultrasound image formation to compare against. To this end, we have made the MimickNet software open source.
Open AccessMimickNet, Matching Clinical Post-Processing Under Realistic Black-Box Constraints
Image post-processing is used in clinical-grade ultrasound scanners to improve image quality (e.g., reduce speckle noise and enhance contrast). These post-processing techniques vary across manufacturers and are generally kept proprietary, which presents a challenge for researchers looking to match current clinical-grade workflows. We introduce a deep learning framework, MimickNet, that transforms raw conventional delay-and-summed (DAS) beams into the approximate post-processed images found on clinical-grade scanners. Training MimickNet only requires post-processed image samples from a scanner of interest without the need for explicit pairing to raw DAS data. This flexibility allows it to hypothetically approximate any manufacturer's post-processing without access to the pre-processed data. MimickNet generates images with an average similarity index measurement (SSIM) of 0.930$\pm$0.0892 on a 300 cineloop test set, and it generalizes to cardiac cineloops outside of our train-test distribution achieving an SSIM of 0.967$\pm$0.002. We also explore the theoretical SSIM achievable by evaluating MimickNet performance when trained under gray-box constraints (i.e., when both pre-processed and post-processed images are available). To our knowledge, this is the first work to establish deep learning models that closely approximate current clinical-grade ultrasound post-processing under realistic black-box constraints where before and after post-processing data is unavailable. MimickNet serves as a clinical post-processing baseline for future works in ultrasound image formation to compare against. To this end, we have made the MimickNet software open source.