Deep-learning-based biomarker of spinal cartilage endplate health using ultra-short echo time magnetic resonance imaging.
Deep-learning-based biomarker of spinal cartilage endplate health using ultra-short echo time magnetic resonance imaging.
复制标题
DOI：
10.21037/qims-22-729
复制
发表时间：
2023-05-01
影响因子：
2.8
通讯作者：
Fields, Aaron J.
中科院分区：
文献类型：
--
作者：
Bonnheim, Noah B.;Wang, Linshanshan;Lazar, Ann A.;Chachad, Ravi;Zhou, Jiamin;Guo, Xiaojie;O'Neill, Conor;Castellanos, Joel;Du, Jiang;Jang, Hyungseok;Krug, Roland;Fields, Aaron J.
使用超短回波时间磁共振成像（UTE MRI）测量的脊柱软骨终板（CEP）中的T2* 弛豫时间反映了影响CEP对营养物的渗透性的生化组成的方面。在慢性腰痛（cLBP）患者中，使用UTE MRI的T2* 生物标志物测量的CEP组成缺陷与更严重的椎间盘退变相关。本研究的目标是开发一种客观、准确、高效的基于深度学习的方法，用于使用UTE图像计算CEP健康的生物标志物。腰椎的多回波UTE MRI采集自前瞻性入组的83例受试者的横截面和连续队列，这些受试者的年龄和cLBP相关疾病范围很广。L4-S1水平的CEP在6，972个UTE图像上手动分割，并用于利用u-net架构训练神经网络。使用Dice评分、灵敏度、特异性、Bland-Altman和受试者-操作者特征（ROC）分析比较CEP分割和手动和模型生成分割的平均CEP T2* 值。计算信噪比（SNR）和对比度噪声（CNR）比，并与模型性能相关。与手动CEP分割相比，模型生成的分割实现了0.80-0.91的灵敏度，0.99的特异性，0.77-0.85的Dice评分，0.99的受试者操作特征曲线下的面积值，以及0.56-0.77的精确度-召回（PR）AUC值，这取决于脊柱水平和矢状面图像位置。在未见过的测试数据集中，从模型预测分割中推导出的平均CEP T2* 值和主要CEP角度具有低偏差（T2* 偏差=0.33±2.37 ms，角度偏差=0.36±2.65°）。为了模拟假设的临床场景，使用预测的分割将CEP分层为高、中和低T2* 组。组预测的诊断敏感性为0.77-0.86，特异性为0.86-0.95。模型性能与图像SNR和CNR呈正相关。经过训练的深度学习模型能够实现准确的自动CEP分割和T2* 生物标志物计算，这些计算在统计学上与手动分割相似。这些模型解决了与手动方法相关的低效率和主观性的局限性。这些技术可用于阐明CEP组成在椎间盘退变病因学中的作用，并指导cLBP的新兴疗法。
T2* relaxation times in the spinal cartilage endplate (CEP) measured using ultra-short echo time magnetic resonance imaging (UTE MRI) reflect aspects of biochemical composition that influence the CEP’s permeability to nutrients. Deficits in CEP composition measured using T2* biomarkers from UTE MRI are associated with more severe intervertebral disc degeneration in patients with chronic low back pain (cLBP). The goal of this study was to develop an objective, accurate, and efficient deep-learning-based method for calculating biomarkers of CEP health using UTE images. Multi-echo UTE MRI of the lumbar spine was acquired from a prospectively enrolled cross-sectional and consecutive cohort of 83 subjects spanning a wide range of ages and cLBP-related conditions. CEPs from the L4-S1 levels were manually segmented on 6,972 UTE images and used to train neural networks utilizing the u-net architecture. CEP segmentations and mean CEP T2* values derived from manually- and model-generated segmentations were compared using Dice scores, sensitivity, specificity, Bland-Altman, and receiver-operator characteristic (ROC) analysis. Signal-to-noise (SNR) and contrast-to-noise (CNR) ratios were calculated and related to model performance. Compared with manual CEP segmentations, model-generated segmentations achieved sensitives of 0.80–0.91, specificities of 0.99, Dice scores of 0.77–0.85, area under the receiver-operating characteristic curve values of 0.99, and precision-recall (PR) AUC values of 0.56–0.77, depending on spinal level and sagittal image position. Mean CEP T2* values and principal CEP angles derived from the model-predicted segmentations had low bias in an unseen test dataset (T2* bias =0.33±2.37 ms, angle bias =0.36±2.65°). To simulate a hypothetical clinical scenario, the predicted segmentations were used to stratify CEPs into high, medium, and low T2* groups. Group predictions had diagnostic sensitivities of 0.77–0.86 and specificities of 0.86–0.95. Model performance was positively associated with image SNR and CNR. The trained deep learning models enable accurate, automated CEP segmentations and T2* biomarker computations that are statistically similar to those from manual segmentations. These models address limitations with inefficiency and subjectivity associated with manual methods. Such techniques could be used to elucidate the role of CEP composition in disc degeneration etiology and guide emerging therapies for cLBP.
登录
查看更多内容
影响因子：
3.7
作者：
Dolor, Aaron;Sampson, Sara L.;Fields, Aaron J.
通讯作者：
Fields, Aaron J.
影响因子：
19.7
作者：
Fields, Aaron J.;Han, Misung;Lotz, Jeffrey C.
通讯作者：
Lotz, Jeffrey C.
影响因子：
2.8
作者：
Bonnheim, Noah B.;Wang, Linshanshan;Lazar, Ann A.;Zhou, Jiamin;Chachad, Ravi;Sollmann, Nico;Guo, Xiaojie;Iriondo, Claudia;O'Neill, Conor;Lotz, Jeffrey C.;Link, Thomas M.;Krug, Roland;Fields, Aaron J.
通讯作者：
Fields, Aaron J.
DOI：
10.2307/2987937
发表时间：
1983-01-01
期刊：
JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES D-THE STATISTICIAN
影响因子：
--
作者：
ALTMAN, DG;BLAND, JM
通讯作者：
BLAND, JM
影响因子：
4.3
作者：
Hsu LM;Wang S;Walton L;Wang TW;Lee SH;Shih YI
通讯作者：
Shih YI