Skip to main content
Journal of Central South University Medical Sciences logoLink to Journal of Central South University Medical Sciences
. 2021 Oct 28;46(10):1147–1152. [Article in Chinese] doi: 10.11817/j.issn.1672-7347.2021.200999

前列腺癌骨转移诊断的研究进展

Progress in diagnosis of bone metastasis of prostate cancer

刘 俊 1,2,2, 董 永超 3, 徐 东波 2, 张 春雷 2, 蓝 天 4, 常 德辉 2,
Editor: 傅 希文
PMCID: PMC10930230  PMID: 34911846

Abstract

前列腺癌(prostate cancer,PC)骨转移的诊断对PC患者的治疗以及预后有重要的意义。骨转移的早期诊断最常用的为骨扫描,但骨扫描的诊断特异性较低,存在较高的假阳性。近年来随着对CT、MRI、发射计算机断层显像(emission-computed tomography,ECT)、正电子发射计算机断层显像/计算机断层扫描(positron emission computed tomography/computed tomography,PET/CT)以及深度学习算法-卷积神经网络(convolutional neural networks,CNN)在骨转移诊断应用中的深入研究,各种辅助参数联合应用于骨转移的诊断,显著提高了对骨转移诊断的准确性,还可以对PC骨转移患者的治疗效果进行评估,有望在诊断的同时实现骨转移灶的治疗。

Keywords: 前列腺癌, 骨转移, 诊断


在2018年世界20个地区的统计中[1],前列腺癌(prostate cancer,PC)的发病率在14个地区成为最常见的男性癌症,其中有5个地区PC成为了癌症死亡的首要原因。随着生活水平的提高及寿命的延长,我国PC的发病率也呈现出显著上升的趋势,大多数患者在确诊时就已经处于晚期[2]。有数据[3]表明70%以上的晚期PC患者会伴有骨转移的发生,而且骨是所有PC远处转移中最常见的转移部位[4]。目前,诊断PC患者骨转移的公认最常用的方法为99m锝标记的亚甲基双膦酸盐(99mtechnetium-methylene diphosphonate,99mTc-MDP)骨扫描(bone scan,BS),但是其特异性仅有57%[5],不能够有效区分出患者的PC骨转移与良性骨病变。骨转移PC患者一旦进展为转移性去势抵抗型PC(metastasize castration-resistant prostate cancer,mCRPC),其中位生存时间只有2年,是PC的主要致死原因[6],其一线治疗方案也只能延长4.4个月左右的总生存期[7]。如何准确早期诊断骨转移并提供个体化治疗方案对于临床医生来说仍是一个巨大的挑战,故本文对骨转移PC诊断的最新进展进行综述。

1. 单光子发射计算机断层显像

单光子发射计算机断层显像(single-photon emis-sion-computed tomography,SPECT)是目前广泛使用的BS,SPECT/CT是将SPECT与CT联合应用,从而提高骨显像的特异性,SPECT中摄取增加的部位可能与CT中形态的改变存在一定的联系[8-9]。最新的定量SPECT/CT可以实现对99mTc的摄取值进行定量,Tabotta等[10]学者对标准摄取值(standard uptake value,SUV)在骨转移病灶与良性关节病变鉴别中的诊断价值进行研究,入组39例患者中包括骨转移病灶265处,关节炎性改变24处。结果发现骨转移病灶的SUVmax及SUVmean均显著高于关节炎性病灶[SUVmax分别为(34.6±24.6)和(14.2±3.8) g/mL;SUVmean分别为(20.8±14.7)和(8.9±2.2) g/mL],通过受试者工作特性曲线(receiver operating characteristic curves,ROC)获得SUVmax的临界值为19.5 g/mL,在这个界值下对骨转移病灶的诊断敏感性为87%,特异性为92%。Motegi等[11]发现骨总摄取(total bone uptake,TBU)对于骨转移的诊断也有重要价值,并且发现TBU可以在骨扫描指数(bone scan index,BSI)为0的患者中有着额外的诊断价值,但TBU也不能避免对退行性变的假阳性诊断,ROC曲线下面积(area under curve,AUC)为0.968。

2. 正电子发射计算机断层显像

2.1. 18 氟标记氟化钠PET/CT 68 镓标记的前列腺特异性膜抗原PET/CT

正电子发射计算机断层显像/计算机断层扫描(positron emission computed tomography/computed tomography,PET/CT)在骨转移的诊断敏感性及特异性方面均明显高于BS[12],在临床诊疗活动中,PET/CT对BS诊断后的患者提供的价值受到越来越多学者的重视。Zacho等[13]在中高危PC患者中进行前瞻性研究,针对入组BS均显示无骨转移的患者进行18氟标记氟化钠(18fluorine-sodium fluoride,18F-NaF)PET/CT检查,结果显示1例存在骨转移,7例可疑骨转移,随后进行根治性前列腺切除术,在2年的临床及影像学随访期内,所有患者均没有出现骨转移。这项研究结果使临床医生对18F-NaF PET/CT有了新的认识:对18F-NaF-PET/CT的结果解读应更加仔细,以便减少假阳性及可疑阳性的出现,避免患者因诊断错误而失去积极治疗的机会及过度治疗的发生。Zacho等[14]68镓标记的前列腺特异性膜抗原(68gallium-prostate specific membrane anti-gen,68Ga-PSMA)PET/CT的额外诊断价值进行了研究,在112名受试者中BS显示81例无骨转移,22例有可疑骨转移。经过68Ga-PSMA PET/CT检查后无骨转移的患者中,8例有可疑转移,其中6例经各项检验、检查及随访证实存在骨转移,2例为假阳性;BS可疑骨转移患者中9例诊断存在骨转移,其中假阳性2例,可疑2例,阴性11例。BS与68Ga-PSMA PET/CT的骨转移阳性结果一致,证明68Ga-PSMA PET/CT对于BS阴性及可疑骨转移患者的进一步诊断有重要意义。值得注意的是,68Ga-PSMA PET/CT存在4例假阳性患者,4例患者的部位均为肋骨,所以对于只有肋骨PSMA阳性的病变应进行仔细区分。虽然68Ga-PSMA PET/CT对于骨转移的诊断优于BS,但是其价格比较昂贵,限制了它的普及。有学者将BS与不同实验室检查指标相结合发现:在碱性磷酸酶(alkaline phosphatase,ALP)≥120 U/L和前列腺特异性抗原(prostate specific antigen,PSA)≥50 ng/mL的情况下,BS诊断骨转移的准确率分别为95.8%和87.5%,这一发现大大提高了BS单独使用对骨转移诊断的准确率[15]。通过辅助指标对患者进行分层可以提高BS的诊断效能,对于BS诊断效能较差的分层再进一步推荐68Ga-PSMA PET/CT,这样既可以提高患者BS的诊断效能,又可以减少患者的经济负担,还可以减少资源的浪费。

2.2. 89 锆标记的去铁胺MSTP2109A抗体PET/CT

前列腺六跨膜上皮抗原-1(six-transmembrane epithelial antigen of prostate-1,STEAP1)是一个含339个氨基酸的细胞表面标志,它们在PC中高表达,也存在于其他癌症中,但是与其他正常组织几乎没有交叉反应[16-17]。MSTP2109A是一种人源化IgG1单克隆抗体,可与STEAP1特异性结合。Carrasquillo等[18]89锆标记的去铁胺(89zirconium-defetoxamine,89Zr-DFO)-MSTP2109A在mCRPC患者中的显像能力进行前瞻性单中心研究,结果发现骨转移灶对 89Zr-DFO-MSTP2109A的中位SUVmax为20.6,明显优于之前报道的89Zr-DFO-huJ591抗PSMA药物(平均SUVmax为8.9)和89Zr-IAB2M抗PSMA微体(平均SUVmax为13.8)[19-20]89Zr-DFO-MSTP2109A显示骨转移的真阳性率为86%(95%可信区间为75%~100%)。此项研究入组患者只有19例,需要更大的多中心研究来进一步证实89Zr-DFO-MSTP2109A的诊断效能。根据目前研究结果,89Zr-DFO-MSTP2109A值得深入研究,并且基于MSTP2109A的抗体-药物结合物(antibody-drug conjugate,ADC)的治疗也为mCRPC患者提供了新的治疗方案,在Carrasquillo等[18]的研究中15名患者接受ADC治疗,其中6名由于不良反应终止治疗,最后因患者数量过少而无法对疗效进行准确评估。MSTP2109A在骨转移的诊断与治疗中都存在潜在的价值,有希望为mCRPC患者提供新的希望,后续相关研究成果值得期待。

3. 深度学习算法-卷积神经网络

在近几年中,深度学习在信息处理领域取得了飞快的发展,尤其是语音识别和图像识别等方面[21-22],深度学习算法-卷积神经网络(convolutional neural networks,CNN)在医学成像中发挥越来越重要的作用[23-24]。Aoki等[25]分别将基于深度学习算法的软件[26]和核医学专家对139例PC患者的BS结果进行分析,将患者MRI或临床资料作为诊断是否存在骨转移的金标准。结果显示在12个身体区域中核医学专家诊断的敏感性、特异性和准确性分别为100%、94.9%和97.1%,使用软件的检测效能分别为91.7%、87.3%和89.2%。虽然软件与专家的检测效能存在一定的差异,但是在10个身体区域中(除了颈椎和骶骨区域)专家的检测率与软件之间没有统计学差异,充分显示出深度学习算法应用于评估PC患者骨转移的辅助诊断的价值。Papandrianos等[27]提出了红绿蓝(red green blue,RGB)-CNN模型,RGB-CNN不但体系结构简单,运行时间短,而且通过对778名接受BS的PC患者进行验证发现,其鉴别骨转移病灶、退行性病变及正常骨组织的准确性为(91.42±1.64)%。Papandrianos等[28]还发现CNN模型具有以下特征:当批次=16,丢弃=0.7,节点(3个转换层)=8,16,32,密集节点=64,历元=200,像素=256,256,3时,各项指标均优于其他现有的CNN模型(准确率为0.97、精确度为0.969、召回率为0.974、灵敏度为0.965、特异性为0.968)。表面增强拉曼光谱(surface enhanced raman spectroscopy,SERS)是一种强有力的血液分析技术,具有快速响应、高敏感度、高特异度的优点,并具有不同分子结构特有的峰[29],但是识别复杂光谱中存在差异的因素并进行分类仍是一项极大的挑战。CNN的飞速发展使得对高精度的SERS识别成为了可能[30]。Shao等[31]将CNN与SERS联合用于对PC骨转移患者的筛查,在用SERS数据训练出的CNN模型中,当迭代次数到达57时CNN模型的训练准确率最高为(99.51±0.23)%,测试准确率为(81.7±2.83)%,测试灵敏度为(80.63±5.07)%,测试特异度为(82.82±2.94)%。CNN的使用可以在BS的基础上显著提高骨转移诊断的准确率,并可以减少不同临床医师诊断时的个体差异,提高对骨转移PC患者诊断的准确性。今后通过大数据对CNN模型进行训练,有望进一步对现有的CNN模型进行改进,从而提高模型对PC患者骨转移诊断的准确率。

4. BSI

BSI是为了量化骨骼受累情况而被提出的,通过观察骨骼的受累情况,获得受累骨骼重量占总骨骼重量的比值,从而量化骨骼受累情况[32]。Petersen等[33]发现:当BSI截断为0%时,诊断PC患者无骨转移的敏感性为96%,但特异性只有38%;而当BSI截断为1%时,诊断时敏感性为98%,但特异性只有58%。Wuestemann等[34]通过对951例PC患者的BSI与是否存在骨转移进行研究,其中BSI的值由接受骨转移性PC患者BS结果训练过的CNN模型自动计算获得,结果发现在截断值为0.27%时,骨转移诊断的敏感性及特异性达到最高(分别为87%和99%)。将CNN与BSI联合使用可显著提高BS对骨转移诊断的敏感性及特异性,Motegi等[11]也提出将BSI与TUB联合使用有望提高对PC患者骨转移的诊断效能。Alshehri等[35]设计了一套转移性骨扫描工具(metastatic bone scan tool,MetsBST)对BS结果进行解读,并与现有的自动BSI(automated bone scan index,ABSI)软件结果对比,发现MetsBST对骨转移的判读效能与ABSI有良好的一致性(R 2=0.9189)。更重要的是,MetsBST可以评估不同区域骨转移病灶对于治疗疗效的反应,并且发现在前列腺与盆腔区域223镭联合外照射放射治疗效果较好,而脊柱区域单独接受223镭、不接受外照射放射治疗的效果较好。这也为今后新治疗方案的评估提供了依据,更有利于指导个体化治疗方案的制订。

5. CTMRI

在鉴别骨转移病灶的过程中,CT与MRI有着独特的作用,尤其是当ECT结果不能明确是否存在骨转移病灶时。正常的红骨髓在MRI中纵向弛豫时间加权像(longitudinal relaxation time weighted image,T1WI)上的信号高于T1WI上的肌肉或椎间盘,而骨转移病灶在T1WI上的信号低于肌肉或椎间盘,但良性红骨髓沉积在T1WI上的信号与骨转移灶相似[36]。Park等[37]对36例PC患者进行分析,结果为骨转移病灶22例,良性红骨髓沉积病灶14例。通过获得T1WI、弥散加权成像(diffusion weighted imaging,DWI)和动态增强(dynamic contrast enhanced,DCE)MRI中的5个参数,即病变-肌肉比(lesion-muscle ratio,LMR)、表观弥散系数(apparent diffusion coefficient,ADC)和DCE-MRI药代动力学参数—容量转移常数(volume transfer constant,Ktrans)、反流率(reflux rate,Kep)和血管外细胞外基质体积分数(volume fraction of the extravascular extracellular matrix,Ve),来区分骨转移与良性红骨髓沉积。结果发现:Ktrans、Kep、Ve、ADC的AUC值分别为0.896、0.844、0.812、0.724,具有较好的鉴别效能。结果还表明DCE MRI药代动力学参数对于鉴别骨转移与良性红骨髓沉积效能高于其他指标,如果将不同指标进行联合参考,有望能进一步提高鉴别诊断效能。Arita等[38]对合成磁共振成像(synthetic magnetic resonance imaging,SyMRI)在PC患者中诊断骨转移的可行性进行了前瞻性研究,SyMRI可以产生各种图像的对比度,通过调整参数值使病变更加直观,还可以对纵向弛豫时间(longitudinal relaxation time,T1)、横向弛豫时间(transverse relaxation time,T2)和质子密度(proton density,PD)进行定量分析[39]。结果发现PD值对于区分存在成骨改变的CT成像中的骨转移患者有显著的作用,将结果分别与两位专家得出的结果相比,PD的诊断效能分别为0.93与0.92,有较高的鉴别准确性。在骨转移的诊断中,MRI与PET/CT的诊断效能明显优于CT[40-41],而骨转移PC患者在接受治疗后可能会出现闪烁现象和成骨细胞的活化,使原有骨转移病灶进一步扩大,从而导致CT的诊断效能进一步降低。但CT作为PC患者随访的常规检查,相比MRI与PET/CT,提高CT图像对骨转移病灶的诊断效能更有意义。Acar等[42]通过机器学习算法联合CT纹理分析(纹理分析可以反映组织的异质性,能够评价图像中的灰度强度分布和空间结构[43])对治疗后的骨转移灶进行区分,分别应用决策树、判别分析、支持向量机(support vector machine,SVM)、k近邻(k nearest neighbor,KNN)、集成分类器等学习方法区分骨转移灶治疗后的反应,以了解是进一步进展还是好转。在骨转移组和治疗好转组中获得了28种有统计学意义的纹理数据,其中灰度区长度矩阵-短区高灰度级因子(gray level zone-length matrix_short zone high gray-level emphasis,GLZLM_SZHGE)和基于直方图的峰度值是区分骨转移病灶进展与好转的最佳参数。结合加权KNN算法与CT纹理分析的诊断效能为0.76,具有较好的诊断效能。虽然其效能低于MRI与PET/CT,但是仍高于单纯CT与BS,随着人工智能的飞快发展,CT成像在骨转移PC患者中的随访价值有望进一步增加。Wang等[44]研究者对多参数MRI(multiparameter-MRI,mp-MRI)在PC患者发生骨转移的预测中的作用进行研究,通过长期随访,将176例患者分为骨转移组与无骨转移组(无骨转移组随访>24个月),分别提取T2加权像(weighted image,WI)与DCE T1WI中的976个纹理特征,最终获得15个有统计学意义的纹理特征,结果发现纹理特征、游离前列腺癌特异性抗原(free-prostate specific antigen,F-PSA)与Gleason评分与预测PC患者发生骨转移独立相关,并且将mp-MRI纹理特征与F-PSA、年龄、Gleason评分结合预测PC患者发生骨转移的效能为0.916。值得注意的是,该研究[44]发现PSA水平与骨转移没有显著相关性,而F-PSA与骨转移的发生密切相关(P<0.01)。随着研究的不断深入,发现MRI与CT的临床价值越来越大,在预测骨转移的发生、诊断以及随访中都有重要意义,并且通过新技术、不同指标的联合应用有望进一步提升诊断效能。

综上所述,PC患者正确诊断是否存在早期骨转移对治疗方案的制订有重要意义,与患者的预后显著相关。对于骨转移患者的诊断最常用的是ECT,随着影像学技术及人工智能的发展,越来越多的方法被用来弥补ECT检查的缺点。目前诊断效能最高的是PSMA PET/CT,随着纹理计算、深度学习计算研究的深入,MRI与CT对于诊断骨转移灶的效能也明显提升,尤其是在患者随访过程中可以有效评价骨转移灶的反应,通过BS中的SUV也可进一步提升BS的诊断效能。随着MSTP2109A研究的深入,有望在提高骨转移诊断效能的同时对摄取灶达到治疗的效果。在PET/CT还未能普及的今天,CT、MRI与BS诊断效能的提升有重要意义,而且PET/CT价格昂贵,在患者治疗后的随访过程中很难成为常规随访项目用来评价治疗后效果。诊断技术的进步可以提升CT、MRI与BS对骨转移灶的诊断效能,既可以帮助临床医生更加准确地判断PC患者是否存在骨转移,又可以对PC患者的治疗效果进行有效评估,从而指导治疗方案的制订及调整。

基金资助

甘肃省自然科学基金(18JR3RA404)。

This work was supported by the Natural Science Foundation of Gansu Province, China (18JR3RA404).

利益冲突声明

作者声称无任何利益冲突。

原文网址

http://xbyxb.csu.edu.cn/xbwk/fileup/PDF/2021101147.pdf

参考文献

  • 1. Ferlay J, Colombet M, Soerjomataram I, et al. Estimating the global cancer incidence and mortality in 2018: GLOBOCAN sources and methods[J]. Int J Cancer, 2019, 144(8): 1941-1953. [DOI] [PubMed] [Google Scholar]
  • 2. Liu X, Yu C, Bi Y, et al. Trends and age-period-cohort effect on incidence and mortality of prostate cancer from 1990 to 2017 in China[J]. Public Health, 2019, 172: 70-80. [DOI] [PubMed] [Google Scholar]
  • 3. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019[J]. CA Cancer J Clin,  2019,  69(1):  7-34. [DOI] [PubMed] [Google Scholar]
  • 4. Shou J, Zhang Q, Wang S, et al. The prognosis of different distant metastases pattern in prostate cancer: A population based retrospective study[J]. Prostate, 2018, 78(7): 491-497. [DOI] [PubMed] [Google Scholar]
  • 5. Evensapir E, Metser U, Mishani E, et al. The detection of bone metastases in patients with high-risk prostate cancer: 99mTc-MDP Planar bone scintigraphy, single- and multi-field-of-view SPECT, 18F-fluoride PET, and 18F-fluoride PET/CT[J]. J Nucl Med, 2006, 47(2): 287-297. [PubMed] [Google Scholar]
  • 6. Cornford P, Bellmunt J, Bolla M, et al. EAU-ESTRO-SIOG guidelines on prostate cancer. Part II: Treatment of relapsing, metastatic, and castration-resistant prostate cancer[J]. Eur Urol, 2017, 71(4): 630-642. [DOI] [PubMed] [Google Scholar]
  • 7. Ryan CJ, Smith MR, Fizazi K, et al. Abiraterone acetate plus prednisone versus placebo plus prednisone in chemotherapy-naive men with metastatic castration-resistant prostate cancer (COU-AA-302): final overall survival analysis of a randomised, double-blind, placebo-controlled phase 3 study[J]. Lancet Oncol, 2015, 16(2): 152-160. [DOI] [PubMed] [Google Scholar]
  • 8. Saha S, Burke C, Desai A, et al. SPECT-CT: applications in musculoskeletal radiology[J]. Br J Radiol, 2013, 86(1031): 20120519. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Strobel K, Burger C, Seifert B, et al. Characterization of focal bone lesions in the axial skeleton: performance of planar bone scintigraphy compared with SPECT and SPECT fused with CT[J]. Am J Roentgenol, 2007, 188(5): 467-474. [DOI] [PubMed] [Google Scholar]
  • 10. Tabotta F, Jreige M, Schaefer N, et al. Quantitative bone SPECT/CT: high specificity for identification of prostate cancer bone metastases[J]. BMC Musculoskel Dis, 2019, 20(1): 619-625. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Motegi K, Matsutomo N, Yamamoto T, et al. Evaluation of bone metastasis burden as an imaging biomarker by quantitative single-photon emission computed tomography/computed tomography for assessing prostate cancer with bone metastasis: a phantom and clinical study[J]. Radiol Phys Technol, 2020, 13(3): 219-229. [DOI] [PubMed] [Google Scholar]
  • 12. Zhou J, Gou Z, Wu R, et al. Comparison of PSMA-PET/CT, choline-PET/CT, NaF-PET/CT, MRI, and bone scintigraphy in the diagnosis of bone metastases in patients with prostate cancer: a systematic review and meta-analysis[J]. Skeletal Radiol, 2019, 48(12): 1915-1924. [DOI] [PubMed] [Google Scholar]
  • 13. Zacho HD, Jochumsen MR, Langkilde NC, et al. No added value of 18F-sodium fluoride PET/CT for the detection of bone metastases in patients with newly diagnosed prostate cancer with normal bone scintigraphy[J]. J Nucl Med, 2019, 60(12): 1713-1716. [DOI] [PubMed] [Google Scholar]
  • 14. Zacho HD, Ravn S, Afshar-Oromieh A, et al. Added value of 68Ga-PSMA PET/CT for the detection of bone metastases in patients with newly diagnosed prostate cancer and a previous 99mTc bone scintigraphy[J]. EJNMMI Res, 2020, 10(1): 31-39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Caglar M, Tuncel M, Yildiz E, et al. Bone scintigraphy as a gatekeeper for the detection of bone metastases in patients with prostate cancer: comparison with Ga-68 PSMA PET/CT[J]. Ann Nucl Med, 2020, 34(12): 932-941. [DOI] [PubMed] [Google Scholar]
  • 16. Hubert RS, Vivanco I, Chen E, et al. STEAP: A prostate-specific cell-surface antigen highly expressed in human prostate tumors[J]. Proc Natl Acad Sci USA, 1999, 96(25): 14523-14528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Moreaux J, Kassambara A, Hose D, et al. STEAP1 is overexpressed in cancers: A promising therapeutic target[J]. Biochem Bioph Res Co, 2012, 429(3/4): 148-155. [DOI] [PubMed] [Google Scholar]
  • 18. Carrasquillo JA, Fine BM, Pandit-Taskar N, et al. Imaging metastatic castration-resistant prostate cancer patients with 89Zr-DFO-MSTP2109A anti-STEAP1 antibody[J]. J Nucl Med, 2019, 60(11): 1517-1523. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Pandit-Taskar N, O'Donoghue JA, Durack JC, et al. A Phase I/II study for analytic validation of 89Zr-J591 immunoPET as a molecular imaging agent for metastatic prostate cancer[J]. Clin Cancer Res, 2015, 21(23): 5277-5285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Pandit-Taskar N, O'Donoghue JA, Ruan ST, et al. First-in-human imaging with 89Zr-Df-IAB2M anti-PSMA minibody in patients with metastatic prostate cancer: Pharmacokinetics, biodistribution, dosimetry, and lesion uptake[J]. J Nucl Med, 2016, 57(12): 1858-1864. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Sklan JE, Plassard AJ, Fabbri D, et al. Toward content based image retrieval with deep convolutional neural networks[J]. Proc SPIE Int Soc Opt Eng, 2015, 9417: 94172C. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Toledano DT, Fernández-Gallego MP, Lozano-Diez A. Multi-resolution speech analysis for automatic speech recognition using deep neural networks: Experiments on TIMIT[J]. PLoS One, 2018, 13(10): e0205355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Biswas M, Kuppili V, Saba L, et al. State-of-the-art review on deep learning in medical imaging[J]. Front Biosci (Landmark Ed), 2019. ,1(24): 392-426. [DOI] [PubMed] [Google Scholar]
  • 24. Shen D, Wu G, Suk HI. Deep learning in medical image analysis[J]. Annu Rev Biomed Eng, 2017, 19(1): 221-248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Aoki Y, Nakayama M, Nomura K, et al. The utility of a deep learning-based algorithm for bone scintigraphy in patient with prostate cancer[J]. Ann Nucl Med, 2020, 34(12): 926-931. [DOI] [PubMed] [Google Scholar]
  • 26. Shimizu A, Wakabayashi H, Kanamori T, et al. Automated measurement of bone scan index from a whole-body bone scintigram[J]. Int J Comput Ass Rad, 2020, 15(3): 389-400. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Papandrianos N, Papageorgiou EI, Anagnostis A. Development of Convolutional Neural Networks to identify bone metastasis for prostate cancer patients in bone scintigraphy[J]. Ann Nucl Med, 2020, 34(11): 824-832. [DOI] [PubMed] [Google Scholar]
  • 28. Papandrianos N, Papageorgiou E, Anagnostis A, et al. Bone metastasis classification using whole body images from prostate cancer patients based on convolutional neural networks application[J]. PLoS One, 2020. ,15(8): e0237213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Zhang YY, Mi X, Tan X, et al. Recent progress on liquid biopsy analysis using surface-enhanced raman spectroscopy[J]. Theranostics, 2019, 9(2): 491-525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Liu J, Osadchy M, Ashton L, et al. Deep convolutional neural networks for Raman spectrum recognition: a unified solution[J]. Analyst, 2017, 142(21): 4067-4074. [DOI] [PubMed] [Google Scholar]
  • 31. Shao XG, Zhang H, Wang YQ, et al. Deep convolutional neural networks combine Raman spectral signature of serum for prostate cancer bone metastases screening[J]. Nanomedicine, 2020, 29: 102245. [DOI] [PubMed] [Google Scholar]
  • 32. Imbriaco M, Larson SM, Yeung HW, et al. A new parameter for measuring metastatic bone involvement by prostate cancer: the bone scan index[J]. Clin Cancer Res, 1998, 4(7): 1765-1772. [PubMed] [Google Scholar]
  • 33. Petersen LJ, Mortensen JC, Bertelsen H, et al. Prospective evaluation of computer-assisted analysis of skeletal lesions for the staging of prostate cancer[J]. BMC Med Imaging, 2017, 17(1): 40-45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Wuestemann J, Hupfeld S, Kupitz D, et al. Analysis of bone scans in various tumor entities using a deep-learning-based artificial neural network algorithm-evaluation of diagnostic performance[J]. Cancers, 2020, 12(9): 2654-2666. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Alshehri AHD, Osman SOS, Prise KM, et al. A novel tool for improving the interpretation of isotope bone scans in metastatic prostate cancer[J]. Br J Radiol, 2020, 93(1115): 20200775-20200783. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Kim YP, Kannengiesser S, Paek MY, et al. Differentiation between focal malignant marrow-replacing lesions and benign red marrow deposition of the spine with T2*-corrected fat-signal fraction map using a three-echo volume interpolated breath-hold gradient echo dixon sequence[J]. Korean J Radiol, 2014, 15(6): 781-791. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Park S, Park JG, Jun S, et al. Differentiation of bone metastases from prostate cancer and benign red marrow depositions of the pelvic bone with multiparametric MRI[J]. Magn Reson Imaging, 2020, 73(3): 118-124. [DOI] [PubMed] [Google Scholar]
  • 38. Arita Y, Takahara T, Yoshida S, et al. Quantitative assessment of bone metastasis in prostate cancer using synthetic magnetic resonance imaging[J]. Invest Radiol, 2019, 54(10): 638-644. [DOI] [PubMed] [Google Scholar]
  • 39. Hagiwara A, Warntjes M, Hori M, et al. SyMRI of the brain: rapid quantification of relaxation rates and proton density, with synthetic MRI, automatic brain segmentation, and myelin measurement[J]. Invest Radiol, 2017, 52(10): 647-657. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Filograna L, Lenkowicz J, Cellini F, et al. Identification of the most significant magnetic resonance imaging (MRI) radiomic features in oncological patients with vertebral bone marrow metastatic disease: A feasibility study[J]. Radiol Med, 2019, 124(1): 50-57. [DOI] [PubMed] [Google Scholar]
  • 41. Arvola S, Jambor I, Kuisma A, et al. Comparison of standardized uptake values between 99mTc-HDP SPECT/CT and 18F-NaF PET/CT in bone metastases of breast and prostate cancer[J]. EJNMMI Res, 2019, 9(1): 6-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Acar E, Leblebici A, Ellidokuz BE, et al. Machine learning for differentiating metastatic and completely responded sclerotic bone lesion in prostate cancer: A retrospective radiomics study[J]. Br J Radiol, 2019, 92(1101): 20190286-20190294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Scalco E, Rizzo G. Texture analysis of medical images for radiotherapy applications[J]. Br J Radiol, 2017, 90(1070): 20160642-20160656. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Wang Y, Yu B, Zhong F, et al. MRI-based texture analysis of the primary tumor for pre-treatment prediction of bone metastases in prostate cancer[J]. Magn Reson Imaging, 2019, 60: 76-84. [DOI] [PubMed] [Google Scholar]

Articles from Journal of Central South University Medical Sciences are provided here courtesy of Central South University

RESOURCES