Medical Imaging Research Center (MIRC), Department of Electrical Engineering, KU Leuven, Herestraat 49, 3000, Leuven, Belgium.
Department of Electronics and information systems, UGent - imec, Technologiepark 126, Zwijnaarde, 9052, Belgium.
Med Phys. 2021 May;48(5):2448-2457. doi: 10.1002/mp.14835. Epub 2021 Apr 3.
Three-dimensional (3D) reconstructions of the human anatomy have been available for surgery planning or diagnostic purposes for a few years now. The different image modalities usually rely on several consecutive two-dimensional (2D) acquisitions in order to reconstruct the 3D volume. Hence, such acquisitions are expensive, time-demanding and often expose the patient to an undesirable amount of radiation. For such reasons, along the most recent years, several studies have been proposed that extrapolate 3D anatomical features from merely 2D exams such as x rays for implant templating in total knee or hip arthroplasties.
The presented study shows an adaptation of a deep learning-based convolutional neural network to reconstruct 3D volumes from a mere 2D digitally reconstructed radiograph from one of the most extensive lower limb computed tomography datasets available. This novel approach is based on an encoder-decoder architecture with skip connections and a multidimensional Gaussian filter as data augmentation technique.
The results achieved promising values when compared against the ground truth volumes, quantitatively yielding an average of 0.77 ± 0.05 structured similarity index.
A novel deep learning-based approach to reconstruct 3D medical image volumes from a single x-ray image was shown in the present study. The network architecture was validated against the original scans presenting SSIM values of 0.77 ± 0.05 and 0.78 ± 0.06, respectively for the knee and the hip crop.
三维（3D）人体解剖结构重建已经可以用于手术规划或诊断目的已有几年了。不同的成像方式通常需要进行几次连续的二维（2D）采集，以重建 3D 体积。因此，这种采集既昂贵又耗时，而且往往会使患者暴露在不必要的辐射下。出于这些原因，近年来，已经提出了几项研究，这些研究从 X 射线等二维检查中推断出 3D 解剖特征，以便在全膝关节或髋关节置换术的植入物模板中使用。
本研究展示了一种基于深度学习的卷积神经网络的改编，该网络可以从最广泛的下肢计算机断层扫描数据集之一的仅二维数字重建射线照片中重建 3D 体积。这种新方法基于具有跳过连接的编码器-解码器架构和多维高斯滤波器作为数据增强技术。
与真实体积相比，该方法取得了有希望的结果，在定量方面平均产生了 0.77±0.05 的结构相似性指数。
本研究提出了一种从单张 X 射线图像重建 3D 医学图像体积的新的基于深度学习的方法。该网络架构针对原始扫描进行了验证，膝关节和髋关节裁剪的 SSIM 值分别为 0.77±0.05 和 0.78±0.06。
