Tivnan Matthew, Gang Grace J, Wang Wenying, Noël Peter, Sulam Jeremias, Webster Stayman J
Johns Hopkins University, Department of Biomedical Engineering, Baltimore, Maryland, United States.
Hospital of the University of Pennsylvania, Department of Radiology, Philadelphia, Pennsylvania, United States.
J Med Imaging (Bellingham). 2023 May;10(3):033501. doi: 10.1117/1.JMI.10.3.033501. Epub 2023 May 4.
Optimization of CT image quality typically involves balancing variance and bias. In traditional filtered back-projection, this trade-off is controlled by the filter cutoff frequency. In model-based iterative reconstruction, the regularization strength parameter often serves the same function. Deep neural networks (DNNs) typically do not provide this tunable control over output image properties. Models are often trained to minimize the expected mean squared error, which penalizes both variance and bias in image outputs but does not offer any control over the trade-off between the two. We propose a method for controlling the output image properties of neural networks with a new loss function called weighted covariance and bias (WCB). Our proposed method uses multiple noise realizations of the input images during training to allow for separate weighting matrices for the variance and bias penalty terms. Moreover, we show that tuning these weights enables targeted penalization of specific image features with spatial frequency domain penalties. To evaluate our method, we present a simulation study using digital anthropomorphic phantoms, physical simulation of CT measurements, and image formation with various algorithms. We show that the WCB loss function offers a greater degree of control over trade-offs between variance and bias, whereas mean-squared error provides only one specific image quality configuration. We also show that WCB can be used to control specific image properties including variance, bias, spatial resolution, and the noise correlation of neural network outputs. Finally, we present a method to optimize the proposed weights for a spiculated lung nodule shape discrimination task. Our results demonstrate this new image quality can control the image properties of DNN outputs and optimize image quality for task-specific performance.
CT图像质量的优化通常涉及平衡方差和偏差。在传统的滤波反投影中，这种权衡由滤波器截止频率控制。在基于模型的迭代重建中，正则化强度参数通常起到相同的作用。深度神经网络（DNN）通常无法对输出图像属性进行这种可调节的控制。模型通常被训练以最小化预期的均方误差，这会惩罚图像输出中的方差和偏差，但无法对两者之间的权衡进行任何控制。我们提出了一种使用称为加权协方差和偏差（WCB）的新损失函数来控制神经网络输出图像属性的方法。我们提出的方法在训练期间使用输入图像的多个噪声实现，以便为方差和偏差惩罚项设置单独的加权矩阵。此外，我们表明调整这些权重能够通过空间频域惩罚对特定图像特征进行有针对性的惩罚。为了评估我们的方法，我们进行了一项模拟研究，使用数字人体模型、CT测量的物理模拟以及各种算法进行图像生成。我们表明，WCB损失函数在方差和偏差之间的权衡方面提供了更大程度的控制，而均方误差仅提供一种特定的图像质量配置。我们还表明，WCB可用于控制特定的图像属性，包括方差、偏差、空间分辨率以及神经网络输出的噪声相关性。最后，我们提出了一种针对毛刺状肺结节形状判别任务优化所提出权重的方法。我们的结果表明，这种新的图像质量控制方法能够控制DNN输出图像的属性，并针对特定任务性能优化图像质量。
