Deep neural networks have been shown to be effective adaptive beamformers for ultrasound imaging. However, when training with traditional L p norm loss functions, model selection is difficult because lower loss values are not always associated with higher image quality. This ultimately limits the maximum achievable image quality with this approach and raises concerns about the optimization objective. In an effort to align the optimization objective with the image quality metrics of interest, we implemented a novel ultrasound-specific loss function based on the spatial lag-one coherence and signal-to-noise ratio of the delayed channel data in the short-time Fourier domain. We employed the R-Adam optimizer with look ahead and cyclical learning rate to make the training more robust to initialization and local minima, leading to better model performance and more reliable convergence. With our custom loss function and optimization scheme, we achieved higher contrast-to-noise-ratio, higher speckle signal-to-noise-ratio, and more accurate contrast ratio reconstruction than with previous deep learning and delay-and-sum beamforming approaches.
more »
« less
Application of the generalized contrast-to-noise ratio to assess photoacoustic image quality
The generalized contrast-to-noise ratio (gCNR) is a relatively new image quality metric designed to assess the probability of lesion detectability in ultrasound images. Although gCNR was initially demonstrated with ultrasound images, the metric is theoretically applicable to multiple types of medical images. In this paper, the applicability of gCNR to photoacoustic images is investigated. The gCNR was computed for both simulated and experimental photoacoustic images generated by amplitude-based (i.e., delay-and-sum) and coherence-based (i.e., short-lag spatial coherence) beamformers. These gCNR measurements were compared to three more traditional image quality metrics (i.e., contrast, contrast-to-noise ratio, and signal-to-noise ratio) applied to the same datasets. An increase in qualitative target visibility generally corresponded with increased gCNR. In addition, gCNR magnitude was more directly related to the separability of photoacoustic signals from their background, which degraded with the presence of limited bandwidth artifacts and increased levels of channel noise. At high gCNR values (i.e., 0.95-1), contrast, contrast-to-noise ratio, and signal-to-noise ratio varied by up to 23.7-56.2 dB, 2.0-3.4, and 26.5-7.6×1020, respectively, for simulated, experimental phantom, and
- Award ID(s):
- 1751522
- NSF-PAR ID:
- 10160786
- Publisher / Repository:
- Optical Society of America
- Date Published:
- Journal Name:
- Biomedical Optics Express
- Volume:
- 11
- Issue:
- 7
- ISSN:
- 2156-7085
- Page Range / eLocation ID:
- Article No. 3684
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
more » « less
Objectives Ultrasound users are advised to observe the ALARA (as low as reasonably achievable) principle, but studies have shown that most do not monitor acoustic output metrics. We developed an adaptive ultrasound method that could suggest acoustic output levels based on real‐time image quality feedback using lag‐one coherence (LOC).
Methods Lag‐one coherence as a function of the mechanical index (MI) was assessed in 35 healthy volunteers in their second trimester of pregnancy. While imaging the placenta or the fetal abdomen, the system swept through 16 MI values ranging from 0.15 to 1.20. The LOC‐versus‐MI data were fit with a sigmoid curve, and the ALARA MI was selected as the point at which the fit reached 98% of its maximum.
Results In this study, the ALARA MI values were between 0.35 and 1.03, depending on the acoustic window. Compared to a default MI of 0.8, the pilot acquisitions suggested a lower ALARA MI 80% of the time. The contrast, contrast‐to‐noise ratio, generalized contrast‐to‐noise ratio, and LOC all followed sigmoidal trends with an increasing MI. The
R 2of the fit was statistically significantly greater for LOC than the other metrics (P < .017).Conclusions These results suggest that maximum image quality can be achieved with acoustic output levels lower than the US Food and Drug Administration limits in many cases, and an automated tool could be used in real time to find the ALARA MI for specific imaging conditions. Our results support the feasibility of an automated, LOC‐based implementation of the ALARA principle for obstetric ultrasound.
-
Flexible array transducer for photoacoustic-guided interventions: phantom and ex vivo demonstrationsmore » « less
Photoacoustic imaging has demonstrated recent promise for surgical guidance, enabling visualization of tool tips during surgical and non-surgical interventions. To receive photoacoustic signals, most conventional transducers are rigid, while a flexible array is able to deform and provide complete contact on surfaces with different geometries. In this work, we present photoacoustic images acquired with a flexible array transducer in multiple concave shapes in phantom and
ex vivo bovine liver experiments targeted toward interventional photoacoustic applications. We validate our image reconstruction equations for known sensor geometries with simulated data, and we provide empirical elevation field-of-view, target position, and image quality measurements. The elevation field-of-view was 6.08 mm at a depth of 4 cm and greater than 13 mm at a depth of 5 cm. The target depth agreement with ground truth ranged 98.35-99.69%. The mean lateral and axial target sizes when imaging 600μ m-core-diameter optical fibers inserted within the phantoms ranged 0.98-2.14 mm and 1.61-2.24 mm, respectively. The mean ± one standard deviation of lateral and axial target sizes when surrounded by liver tissue were 1.80±0.48 mm and 2.17±0.24 mm, respectively. Contrast, signal-to-noise, and generalized contrast-to-noise ratios ranged 6.92–24.42 dB, 46.50–67.51 dB, and 0.76–1, respectively, within the elevational field-of-view. Results establish the feasibility of implementing photoacoustic-guided surgery with a flexible array transducer. -
null (Ed.)Ultrasound B-Mode images are created from data obtained from each element in the transducer array in a process called beamforming. The beamforming goal is to enhance signals from specified spatial locations, while reducing signal from all other locations. On clinical systems, beamforming is accomplished with the delay-and-sum (DAS) algorithm. DAS is efficient but fails in patients with high noise levels, so various adaptive beamformers have been proposed. Recently, deep learning methods have been developed for this task. With deep learning methods, beamforming is typically framed as a regression problem, where clean, ground-truth data is known, and usually simulated. For in vivo data, however, it is extremely difficult to collect ground truth information, and deep networks trained on simulated data underperform when applied to in vivo data, due to domain shift between simulated and in vivo data. In this work, we show how to correct for domain shift by learning deep network beamformers that leverage both simulated data, and unlabeled in vivo data, via a novel domain adaption scheme. A challenge in our scenario is that domain shift exists both for noisy input, and clean output. We address this challenge by extending cycle-consistent generative adversarial networks, where we leverage maps between synthetic simulation and real in vivo domains to ensure that the learned beamformers capture the distribution of both noisy and clean in vivo data. We obtain consistent in vivo image quality improvements compared to existing beamforming techniques, when applying our approach to simulated anechoic cysts and in vivo liver data.more » « less
-
more » « less
Abstract Background Magnetic resonance imaging (MRI) scans are known to suffer from a variety of acquisition artifacts as well as equipment‐based variations that impact image appearance and segmentation performance. It is still unclear whether a direct relationship exists between magnetic resonance (MR) image quality metrics (IQMs) (e.g., signal‐to‐noise, contrast‐to‐noise) and segmentation accuracy.
Purpose Deep learning (DL) approaches have shown significant promise for automated segmentation of brain tumors on MRI but depend on the quality of input training images. We sought to evaluate the relationship between IQMs of input training images and DL‐based brain tumor segmentation accuracy toward developing more generalizable models for multi‐institutional data.
Methods We trained a 3D DenseNet model on the BraTS 2020 cohorts for segmentation of tumor subregions enhancing tumor (ET), peritumoral edematous, and necrotic and non‐ET on MRI; with performance quantified via a 5‐fold cross‐validated Dice coefficient. MRI scans were evaluated through the open‐source quality control tool MRQy, to yield 13 IQMs per scan. The Pearson correlation coefficient was computed between whole tumor (WT) dice values and IQM measures in the training cohorts to identify quality measures most correlated with segmentation performance. Each selected IQM was used to group MRI scans as “better” quality (BQ) or “worse” quality (WQ), via relative thresholding. Segmentation performance was re‐evaluated for the DenseNet model when (i) training on BQ MRI images with validation on WQ images, as well as (ii) training on WQ images, and validation on BQ images. Trends were further validated on independent test sets derived from the BraTS 2021 training cohorts.
Results For this study, multimodal MRI scans from the BraTS 2020 training cohorts were used to train the segmentation model and validated on independent test sets derived from the BraTS 2021 cohort. Among the selected IQMs, models trained on BQ images based on inhomogeneity measurements (coefficient of variance, coefficient of joint variation, coefficient of variation of the foreground patch) and the models trained on WQ images based on noise measurement peak signal‐to‐noise ratio (SNR) yielded significantly improved tumor segmentation accuracy compared to their inverse models.
Conclusions Our results suggest that a significant correlation may exist between specific MR IQMs and DenseNet‐based brain tumor segmentation performance. The selection of MRI scans for model training based on IQMs may yield more accurate and generalizable models in unseen validation.
