Search

Background: In recent years, low‐dose computed tomography (LDCT) has played an important role in the diagnosis CT to reduce the potential adverse effects of X‐ray radiation on patients, while maintaining the same diagnostic image quality. Purpose: Deep learning (DL)‐based methods have played an increasingly important role in the field of LDCT imaging. However, its performance is highly dependent on the consistency of feature distributions between training data and test data. Due to patient's breathing movements during data acquisition, the paired LDCT and normal dose CT images are difficult to obtain from realistic imaging scenarios. Moreover, LDCT images from simulation or clinical CT examination often have different feature distributions due to the pollution by different amounts and types of image noises. If a network model trained with a simulated dataset is used to directly test clinical patients' LDCT data, its denoising performance may be degraded. Based on this, we propose a novel domain‐adaptive denoising network (DADN) via noise estimation and transfer learning to resolve the out‐of‐distribution problem in LDCT imaging. Methods: To overcome the previous adaptation issue, a novel network model consisting of a reconstruction network and a noise estimation network was designed. The noise estimation network based on a double branch structure is used for image noise extraction and adaptation. Meanwhile, the U‐Net‐based reconstruction network uses several spatially adaptive normalization modules to fuse multi‐scale noise input. Moreover, to facilitate the adaptation of the proposed DADN network to new imaging scenarios, we set a two‐stage network training plan. In the first stage, the public simulated dataset is used for training. In the second transfer training stage, we will continue to fine‐tune the network model with a torso phantom dataset, while some parameters are frozen. The main reason using the two‐stage training scheme is based on the fact that the feature distribution of image content from the public dataset is complex and diverse, whereas the feature distribution of noise pattern from the torso phantom dataset is closer to realistic imaging scenarios. Results: In an evaluation study, the trained DADN model is applied to both the public and clinical patient LDCT datasets. Through the comparison of visual inspection and quantitative results, it is shown that the proposed DADN network model can perform well in terms of noise and artifact suppression, while effectively preserving image contrast and details. Conclusions: In this paper, we have proposed a new DL network to overcome the domain adaptation problem in LDCT image denoising. Moreover, the results demonstrate the feasibility and effectiveness of the application of our proposed DADN network model as a new DL‐based LDCT image denoising method. [ABSTRACT FROM AUTHOR]

Chronic inflammation (CI) in older adults is associated with reduced health span and life span. Interleukin-6 (IL-6) is one CI marker that is strongly associated with adverse health outcomes and mortality in aging. We have previously characterized a mouse model of frailty and chronic inflammatory pathway activation (IL-10tm/tm, IL-10 KO) that demonstrates the upregulation of numerous proinflammatory cytokines, including IL-6. We sought to identify a more specific role for IL-6 within the context of CI and aging and developed a mouse with targeted deletion of both IL-10 and IL-6 (IL-10tm/tm/IL-6tm/tm, DKO). Phenotypic characteristics, cytokine measurements, cardiac myocardial oxygen consumption, physical function, and survival were measured in DKO mice and compared to age- and gender-matched IL-10 KO and wild-type mice. Our findings demonstrate that selective knockdown of IL-6 in a frail mouse with CI resulted in the reversal of some of the CI-associated changes. We observed increased protective mitochondrial-associated lipid metabolites, decreased cardiac oxaloacetic acid, improved myocardial oxidative metabolism, and better short-term functional performance in DKO mice. However, the DKO mice also demonstrated higher mortality. This work shows the pleiotropic effects of IL-6 on aging and frailty. [ABSTRACT FROM AUTHOR]

LumaGEM is a newly developed gamma camera for dedicated, small field of view, high spatial resolution imaging. The system consists of an array of 2 x 2 x 6 [mm.sup.3] NaI(Tl) pixels coupled to an array of position-sensitive photomultiplier tubes. It has a 125 x 125 [mm.sup.2] field of view. A pinhole collimator was used on LumaGEM to acquire SPECT images of mice that had transgenic modifications so as to model various diseases. Pinhole apertures of 1, 2, and 3 mm are interchangeable on the collimator and were used to acquire images. An iterative MLEM algorithm for pinhole SPECT was used to reconstruct the 128 projection images that covered 360 [degrees] rotation. The reconstruction algorithm is based on a projector and backprojector pair implemented using a ray-tracing algorithm. The crucial reconstruction input parameters are the radius of rotation, center of rotation, and pinhole focal length. Ideal pinhole geometry is assumed, and no correction for attenuation has been made. The preliminary images presented here show detailed uptake in the mice subjects and are a convincing sign that animal SPECT can reach submiilimeter spatial resolution and be a valuable tool in the study of diseases and the development of pharmaceuticals in animal models. Index Terms--Animal SPECT, high-resolution SPECT, Lumagem, small animal imaging, small gamma camera.

Respiratory motion can cause artifacts in myocardial SPECT and computed tomography (CT). We incorporate models of respiratory mechanics into the current 4D MCAT and into the next generation spline-based MCAT phantoms. In order to simulate respiratory motion in the current MCAT phantom, the geometric solids for the diaphragm, heart, ribs, and lungs were altered through manipulation of parameters defining them. Affine transformations were applied to the control points defining the same respiratory structures in the spline-based MCAT phantom to simulate respiratory motion. The Non-Uniform Rational B-Spline (NURBS) surfaces for the lungs and body outline were constructed in such a way as to be linked to the surrounding ribs. Expansion and contraction of the thoracic cage then coincided with expansion and contraction of the lungs and body. The changes both phantoms underwent were spline-interpolated over time to create time continuous 4D respiratory models. We then used the geometry-based and spline-based MCAT phantoms in an initial simulation study of the effects of respiratory motion on myocardial SPECT. The simulated reconstructed images demonstrated distinct artifacts in the inferior region of the myocardium. We conclude that both respiratory models can be effective tools for researching effects of respiratory motion.

We present an iterative transmission reconstruction algorithm based on principles used for block-iterative emission reconstruction. Filtered backprojection (FBP) is adequate for transmission CT image reconstruction under ideal conditions. In cases where insufficient count rates are obtained, FBP suffers from significant biases. Our block-iterative transmission algorithm is developed by analogy to ordered-subset emission CT algorithms. We show similarities between our algorithm and a previously-proposed Ordered-Subset Convex (OSC) algorithm. Under ideal conditions, our block-iterative transmission algorithm generates estimates of attenuation line integrals that are not significantly different from those from FBP or OSC. In the case of very low count-rate data, our block-iterative algorithm produces line integral estimates that are less biased than those from FBP and nearly identical to those from OSC. We show that our algorithm is more computationally efficient than OSC. We conclude that our block-iterative method is a viable alternative to FBP and other iterative algorithms, especially under poor imaging conditions.

In this work, we present a method for approximating constrained maximum entropy (ME) reconstructions of SPECT data with modifications to a block-iterative maximum a posteriori (MAP) algorithm. Maximum likelihood (ML)-based reconstruction algorithms require some form of noise smoothing. Constrained ME provides a more formal method of noise smoothing without requiring the user to select parameters. In the context of SPECT, constrained ME seeks the minimum-information image estimate among those whose projections are a given distance from the noisy measured data, with that distance determined by the magnitude of the Poisson noise. Images that meet the distance criterion are referred to as feasible images. We find that modeling of all principal degrading factors (attenuation, detector response, and scatter) in the reconstruction is critical because feasibility is not meaningful unless the projection model is as accurate as possible. Because the constrained ME solution is the same as a MAP solution for a particular value of the MAP weighting parameter, [Beta], the constrained ME solution can be found with a MAP algorithm if the correct value of [Beta] is found. We show that the RBI-MAP algorithm, if used with a dynamic scheme for estimating [Beta], can approximate constrained ME solutions in 20 or fewer iterations. We compare results for various methods of achieving feasible images on a simulation of Tl-201 cardiac SPECT data. Results show that the RBI-MAP ME approximation provides images and quantitative estimates close to those from a slower algorithm that gives the true ME solution. Also, we find that the ME results have higher spatial resolution and greater high-frequency noise content than a feasibility-based stopping rule, feasibility-based low-pass filtering, and a quadratic Gibbs prior with [Beta] selected according to the feasibility criterion. We conclude that fast ME approximation is possible using either RBI-MAP with the dynamic procedure or a feasibility-based stopping rule, and that such reconstructions may be particularly useful in applications where resolution is critical.