Deep Learning-Based Identification of Spinal Metastasis in Lung Cancer Using Spectral CT Images

In this study, deep learning algorithm-based energy/spectral computed tomography (CT) for the spinal metastasis from lung cancer was used. A dilated convolutional U-Net model (DC-U-Net model) was first proposed, which was used to segment the energy/spectral CT image of patients with the spinal metastasis from lung cancer. Subsequently, energy/spectral CT images under different energy levels were collected for the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) comparison. It was found the learning rate of the model decreased exponentially as the number of training increased, with the lung contour segmented out of the image. Under 40–65 keV, the CT value of bone metastasis from lung cancer decreased with increasing energy, as with the average rank sum test result. +e SNR and CNR values were the highest under 60 keV.+e detection rate of the deep learning algorithm below 60 keV was 81.41%, and that of professional doctors was 77.56%. +e detection rate of the deep learning algorithm below 140 keV was 66.03%, and that of professional doctors was 64.74%. In conclusion, the DC-U-Net model demonstrates better segmentation effects versus the convolutional neutral networ k (CNN), with the lung contour segmented. Further, a higher energy level leads to worse segmentation effects on the energy/spectral CT image.

[1]  Z. Wang,et al.  Laparoscopic Varicocelectomy: Virtual Reality Training and Learning Curve , 2014, JSLS : Journal of the Society of Laparoendoscopic Surgeons.

[2]  Jae‐Hong Lee,et al.  Detection and diagnosis of dental caries using a deep learning-based convolutional neural network algorithm. , 2018, Journal of dentistry.

[3]  A. Bergmann,et al.  Bone metastases and skeletal-related events: incidence and prognosis according to histological subtype of lung cancer. , 2019, Future oncology.

[4]  Xiaosong Li,et al.  Automatic classification of lung nodule candidates based on a novel 3D convolution network and knowledge transferred from a 2D network. , 2019, Medical physics.

[5]  H. Popper,et al.  Progression and metastasis of lung cancer , 2016, Cancer and Metastasis Reviews.

[6]  Massoud Saidijam,et al.  Application of Artificial Neural Network in miRNA Biomarker Selection and Precise Diagnosis of Colorectal Cancer , 2019, Iranian biomedical journal.

[7]  Shuai Liu,et al.  Energy Spectrum CT Image Detection Based Dimensionality Reduction with Phase Congruency , 2018, Journal of Medical Systems.

[8]  S. Bennis,et al.  [Management of spinal metastases of lung cancer]. , 2013, Revue des maladies respiratoires.

[9]  吕晓琪 Lü Xiaoqi,et al.  Detection of low dose CT pulmonary nodules based on 3D convolution neural network , 2018 .

[10]  Yang Luo,et al.  Clinical features and treatment of patients with lung adenocarcinoma with bone marrow metastasis , 2019, Tumori.

[11]  Zaisheng Ling,et al.  Evaluation of the best single-energy scanning in energy spectrum CT in lower extremity arteriography. , 2019, Experimental and therapeutic medicine.