基于红外热成像技术与BP神经网络的心肌缺血预诊断方法研究
|
摘要
心肌缺血(MI)是诸多心脏疾病的基础性疾病,可引发多种致命性心脏病,然而在体检筛查中极难发现,病人出现症状时往往错过了最佳治疗时间,因此,心肌缺血的早期发现和早期干预是控制心脏功能衰减或疾病恶化的关键。采集165位健康和有不同程度心肌缺血患者的红外热图像,将所有样本集分为训练集和测试集,通过对红外热图像人体几何定位,提取心前区左右两侧温差集合,并使用多种卷积核对温差集合做降维处理,最终通过反向传播(BP)神经网络在留一法交叉验证下对训练集训练,并确定网络参数,建立分类模型。3×3尺寸的高斯核算子对温差集合卷积后,测试集在BP神经网络上分类准确率达到95.56%,可以为新样本做出准确预测。该方法能够快捷、准确地辅助临床体检对心肌缺血的早期预警,为心肌缺血预诊断提供了新的思路。
Myocardial ischemia(MI)is a heart disease that can cause various types of fatal heart attacks.Patients often miss the best treatment time when they develop the symptoms of a heart attack.Early detection of MI is considered to be necessary for curbing the deterioration of heart diseases because it is difficult to observe the symptoms of a heart attack through a medical check-up.Infrared thermal images of 165 healthy patients with different degrees of MI are collected,and all the samples are divided into training set and test set.Further,the geometrical differences between the left and right sides of the precordial area are extracted based on the geometric positioning of the infrared thermal image of a particular human body.Additionally,several convolutional kernels are used to reduce the dimensionality of the temperature difference set.The training set is trained using the backpropagation(BP)neural network based on the cross-validation method,and the network parameters are determined for establishing a classification model.After the 3×3 size Gaussian kernel operator is convolved on the temperature difference set,the classification accuracy of the test set with respect to the BP neural network becomes 95.56%,thereby denoting that the predictions for the new sample are considerably accurate.Further,the proposed method can rapidly and accurately assist during the early detection of MI in a clinical examination and provide a new methodology for the pre-diagnosis of MI.
Myocardial ischemia(MI)is a heart disease that can cause various types of fatal heart attacks.Patients often miss the best treatment time when they develop the symptoms of a heart attack.Early detection of MI is considered to be necessary for curbing the deterioration of heart diseases because it is difficult to observe the symptoms of a heart attack through a medical check-up.Infrared thermal images of 165 healthy patients with different degrees of MI are collected,and all the samples are divided into training set and test set.Further,the geometrical differences between the left and right sides of the precordial area are extracted based on the geometric positioning of the infrared thermal image of a particular human body.Additionally,several convolutional kernels are used to reduce the dimensionality of the temperature difference set.The training set is trained using the backpropagation(BP)neural network based on the cross-validation method,and the network parameters are determined for establishing a classification model.After the 3×3 size Gaussian kernel operator is convolved on the temperature difference set,the classification accuracy of the test set with respect to the BP neural network becomes 95.56%,thereby denoting that the predictions for the new sample are considerably accurate.Further,the proposed method can rapidly and accurately assist during the early detection of MI in a clinical examination and provide a new methodology for the pre-diagnosis of MI.
引文
[1]Ling T T,Wang L F,Huang M X,et al.Value of3Dcolor-coded imaging device in myocardial ischemia[J].Acta Universitatis Medicinalis Anhui,2015,50(12):1813-1816.凌婷婷,王联发,黄猛珣,等.三维彩色显像仪对心肌缺血诊断的价值[J].安徽医科大学学报,2015,50(12):1813-1816.
[2]Ruzsics B,Lee H,Powers E R,et al.Myocardial ischemia diagnosed by dual-energy computed tomography[J].Circulation,2008,117(9):1244-1245.
[3]Weininger M,Schoepf U J,Ramachandra A,et al.Adenosine-stress dynamic real-time myocardial perfusion CT and adenosine-stress first-pass dualenergy myocardial perfusion CT for the assessment of acute chest pain:initial results[J].European Journal of Radiology,2012,81(12):3703-3710.
[4]Liu X L,Hong W X,Song J L,et al.Using infrared thermal asymmetry analysis for objective assessment of the lesion of facial nerve function[J].Spectroscopy and Spectral Analysis,2012,32(3):647-650.刘旭龙,洪文学,宋佳霖,等.红外热不对称分析用于面神经功能损伤的客观评估[J].光谱学与光谱分析,2012,32(3):647-650.
[5]Wu Z Y,Liu X L,Hong W X,et al.Research on the correlation between the temperature asymmetry at acupoints of healthy and affected side and the severity index of facial paralysis[J].Chinese Acupuncture&Moxibustion,2010,30(11):953-956.吴振英,刘旭龙,洪文学,等.面瘫健患侧穴位温度不对称性与面瘫严重程度相关性研究[J].中国针灸,2010,30(11):953-956.
[6]Jones B F.A reappraisal of the use of infrared thermal image analysis in medicine[J].IEEETransactions on Medical Imaging,1998,17(6):1019-1027.
[7]Xiao Z J,Wan Y W,Li K Y.Clinical application of asymmetry analysis in breast infrared thermal image[J].Chinese Journal of Modern Imaging,2010.肖子健,万颖文,李凯扬.乳腺热图像不对称性分析的临床应用价值探讨[J].中华现代影像学杂志,2010.
[8]Liu X L,Hong W X,Liu J M.Objective assessment of facial paralysis using infrared thermography and formal concept analysis[J].Spectroscopy and Spectral Analysis,2014,34(4):932-936.刘旭龙,洪文学,刘杰民.基于红外热成像与形式概念分析的面瘫病情客观评估方法[J].光谱学与光谱分析,2014,34(4):932-936.
[9]Tang X W,Ding H S,Teng Y C.Pseudo color method for the infrared thermogram display of local breast focus tissue[J].Spectroscopy and Spectral Analysis,2009,29(3):611-615.唐先武,丁海曙,腾轶超.用于显示乳房局部病灶组织红外热图像的伪彩色方法[J].光谱学与光谱分析,2009,29(3):611-615.
[10]Bagavathiappan S,Philip J,Jayakumar T,et al.Correlation between plantar foot temperature and diabetic neuropathy:a case study by using an infrared thermal imaging technique[J].Journal of Diabetes Science and Technology,2010,4(6):1386-1392.
[11]Pan Y H,Chen W L,Xu S B,et al.Clinical report of infrared thermal imaging in screening heart insufficiency and coronary heart disease[J].Clinical Journal of Chinese Medicine,2018,10(13):111-113.潘跃红,陈文良,许少波,等.红外热成像技术筛查心供血不足、冠心病患者的临床价值探讨[J].中医临床研究,2018,10(13):111-113.
[12]Fang Z Z,Shu F,Yuan S Z,et al.Application of infrared thermography in clinical pain evaluation[J].Chinese Journal of Medical Imaging,2011,19(12):931-934.方镇洙,舒帆,袁绍忠,等.红外热成像技术在临床疼痛评定标准中的应用进展[J].中国医学影像学杂志,2011,19(12):931-934.
[13]Gupta S,Markey M K,Bovik A C.Anthropometric3Dface recognition[J].International Journal of Computer Vision,2010,90(3):331-349.
[14]Huttenlocher D P,Klanderman G A,Rucklidge W J.Comparing images using the Hausdorff distance[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1993,15(9):850-863.
[15]Paramanand C,Rajagopalan A N.Efficient geometric matching with higher-order features[C]∥19th International Conference on Pattern Recognition,2008:1-4.
[16]Chen Y,Fan R S,Wang J X,et al.High resolution image classification method combining with minimum noise fraction rotation and convolution neural network[J].Laser&Optoelectronics Progress,2017,54(10):102801.陈洋,范荣双,王竞雪,等.结合最小噪声分离变换和卷积神经网络的高分辨影像分类方法[J].激光与光电子学进展,2017,54(10):102801.
[17]Xu Y,Sun M S.Convolution neural network image defogging based on multi-feature fusion[J].Laser&Optoelectronics Progress,2018,55(3):031012.徐岩,孙美双.基于多特征融合的卷积神经网络图像去雾算法[J].激光与光电子学进展,2018,55(3):031012.
[18]Qian M Y,Yu Y L.Tactile sensing of fiber bragg grating based on back propagation neural network[J].Chinese Journal of Lasers,2017,44(8):0806001.钱牧云,余有龙.基于逆传播神经网络的光纤布拉格光栅触觉传感[J].中国激光,2017,44(8):0806001.
[19]Hagan M T,Demuth H B,Beale M H.Neural network design[M].Dai K,et al.Transl.Beijing:China Machine Press,2002.哈根,德姆斯,比勒.神经网络设计[M].戴葵,等,译.北京:机械工业出版社,2002.
[2]Ruzsics B,Lee H,Powers E R,et al.Myocardial ischemia diagnosed by dual-energy computed tomography[J].Circulation,2008,117(9):1244-1245.
[3]Weininger M,Schoepf U J,Ramachandra A,et al.Adenosine-stress dynamic real-time myocardial perfusion CT and adenosine-stress first-pass dualenergy myocardial perfusion CT for the assessment of acute chest pain:initial results[J].European Journal of Radiology,2012,81(12):3703-3710.
[4]Liu X L,Hong W X,Song J L,et al.Using infrared thermal asymmetry analysis for objective assessment of the lesion of facial nerve function[J].Spectroscopy and Spectral Analysis,2012,32(3):647-650.刘旭龙,洪文学,宋佳霖,等.红外热不对称分析用于面神经功能损伤的客观评估[J].光谱学与光谱分析,2012,32(3):647-650.
[5]Wu Z Y,Liu X L,Hong W X,et al.Research on the correlation between the temperature asymmetry at acupoints of healthy and affected side and the severity index of facial paralysis[J].Chinese Acupuncture&Moxibustion,2010,30(11):953-956.吴振英,刘旭龙,洪文学,等.面瘫健患侧穴位温度不对称性与面瘫严重程度相关性研究[J].中国针灸,2010,30(11):953-956.
[6]Jones B F.A reappraisal of the use of infrared thermal image analysis in medicine[J].IEEETransactions on Medical Imaging,1998,17(6):1019-1027.
[7]Xiao Z J,Wan Y W,Li K Y.Clinical application of asymmetry analysis in breast infrared thermal image[J].Chinese Journal of Modern Imaging,2010.肖子健,万颖文,李凯扬.乳腺热图像不对称性分析的临床应用价值探讨[J].中华现代影像学杂志,2010.
[8]Liu X L,Hong W X,Liu J M.Objective assessment of facial paralysis using infrared thermography and formal concept analysis[J].Spectroscopy and Spectral Analysis,2014,34(4):932-936.刘旭龙,洪文学,刘杰民.基于红外热成像与形式概念分析的面瘫病情客观评估方法[J].光谱学与光谱分析,2014,34(4):932-936.
[9]Tang X W,Ding H S,Teng Y C.Pseudo color method for the infrared thermogram display of local breast focus tissue[J].Spectroscopy and Spectral Analysis,2009,29(3):611-615.唐先武,丁海曙,腾轶超.用于显示乳房局部病灶组织红外热图像的伪彩色方法[J].光谱学与光谱分析,2009,29(3):611-615.
[10]Bagavathiappan S,Philip J,Jayakumar T,et al.Correlation between plantar foot temperature and diabetic neuropathy:a case study by using an infrared thermal imaging technique[J].Journal of Diabetes Science and Technology,2010,4(6):1386-1392.
[11]Pan Y H,Chen W L,Xu S B,et al.Clinical report of infrared thermal imaging in screening heart insufficiency and coronary heart disease[J].Clinical Journal of Chinese Medicine,2018,10(13):111-113.潘跃红,陈文良,许少波,等.红外热成像技术筛查心供血不足、冠心病患者的临床价值探讨[J].中医临床研究,2018,10(13):111-113.
[12]Fang Z Z,Shu F,Yuan S Z,et al.Application of infrared thermography in clinical pain evaluation[J].Chinese Journal of Medical Imaging,2011,19(12):931-934.方镇洙,舒帆,袁绍忠,等.红外热成像技术在临床疼痛评定标准中的应用进展[J].中国医学影像学杂志,2011,19(12):931-934.
[13]Gupta S,Markey M K,Bovik A C.Anthropometric3Dface recognition[J].International Journal of Computer Vision,2010,90(3):331-349.
[14]Huttenlocher D P,Klanderman G A,Rucklidge W J.Comparing images using the Hausdorff distance[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1993,15(9):850-863.
[15]Paramanand C,Rajagopalan A N.Efficient geometric matching with higher-order features[C]∥19th International Conference on Pattern Recognition,2008:1-4.
[16]Chen Y,Fan R S,Wang J X,et al.High resolution image classification method combining with minimum noise fraction rotation and convolution neural network[J].Laser&Optoelectronics Progress,2017,54(10):102801.陈洋,范荣双,王竞雪,等.结合最小噪声分离变换和卷积神经网络的高分辨影像分类方法[J].激光与光电子学进展,2017,54(10):102801.
[17]Xu Y,Sun M S.Convolution neural network image defogging based on multi-feature fusion[J].Laser&Optoelectronics Progress,2018,55(3):031012.徐岩,孙美双.基于多特征融合的卷积神经网络图像去雾算法[J].激光与光电子学进展,2018,55(3):031012.
[18]Qian M Y,Yu Y L.Tactile sensing of fiber bragg grating based on back propagation neural network[J].Chinese Journal of Lasers,2017,44(8):0806001.钱牧云,余有龙.基于逆传播神经网络的光纤布拉格光栅触觉传感[J].中国激光,2017,44(8):0806001.
[19]Hagan M T,Demuth H B,Beale M H.Neural network design[M].Dai K,et al.Transl.Beijing:China Machine Press,2002.哈根,德姆斯,比勒.神经网络设计[M].戴葵,等,译.北京:机械工业出版社,2002.