The use of multivariate MR imaging intensities versus metabolic data from MR spectroscopic imaging for brain tumour classification

This study investigated the value of information from both magnetic resonance imaging and magnetic resonance spectroscopic imaging (MRSI) to automated discrimination of brain tumours. The influence of imaging intensities and metabolic data was tested by comparing the use of MR spectra from MRSI, MR imaging intensities, peak integration values obtained from the MR spectra and a combination of the latter two. Three classification techniques were objectively compared: linear discriminant analysis, least squares support vector machines (LS-SVM) with a linear kernel as linear techniques and LS-SVM with radial basis function kernel as a nonlinear technique. Classifiers were evaluated over 100 stratified random splittings of the dataset into training and test sets. The area under the receiver operating characteristic (ROC) curve (AUC) was used as a global performance measure on test data. In general, all techniques obtained a high performance when using peak integration values with or without MR imaging intensities. For example for low- versus high-grade tumours, low- versus high-grade gliomas and gliomas versus meningiomas, the mean test AUC was higher than 0.91, 0.94, and 0.99, respectively, when both MR imaging intensities and peak integration values were used. The use of metabolic data from MRSI significantly improved automated classification of brain tumour types compared to the use of MR imaging intensities solely.     

稀疏降噪自编码结合高斯过程的近红外光谱药品鉴别方法

提出一种稀疏降噪自编码结合高斯过程的近红外光谱药品鉴别方法。首先对近红外光谱数据进行小波变换以消除基线漂移,然后用稀疏降噪自编码(SDAE)网络提取光谱特征并降维表示,最后采用高斯过程(GP)进行二分类,其中GP选用光谱混合(SM)核函数作为协方差函数,记此分类网络为wSDAG_SM。自编码网络具有很强的模型表示能力,高斯过程分类器在处理小样本数据时具有优势。wSDAG_SM网络通过稀疏降噪自编码学习得到维数更低但更有价值的特征来表示输入数据,同时将具有很好表达力的光谱混合核作为高斯过程的协方差函数,有利于更准确的光谱数据分类。以琥乙红霉素及其他药品的近红外光谱为实验数据,将该方法与经过墨西哥帽小波变换的BP神经网络(wBP)、支持向量机(wSVM), SDAE结合Logistic二分类(wSDAL)、SDAE结合采用平方指数(SE)协方差核的GP二分类(wSDAG_SE),以及未采用小波变换的SDAG_SM网络等方法进行对比。实验结果表明,对光谱数据进行墨西哥帽小波变换预处理能有效提升SDAG_SM网络的分类准确率和稳定性。wSDAG_SM方法无论从分类准确率还是分类结果稳定性方面,都优于其他分类器。     

The theory of the quantum kernel-based binary classifier

Binary classification is a fundamental problem in machine learning. Recent development of quantum similarity-based binary classifiers and kernel method that exploit quantum interference and feature quantum Hilbert space opened up tremendous opportunities for quantum-enhanced machine learning. To lay the fundamental ground for its further advancement, this work extends the general theory of quantum kernel-based classifiers. Existing quantum kernel-based classifiers are compared and the connection among them is analyzed. Focusing on the squared overlap between quantum states as a similarity measure, the essential and minimal ingredients for the quantum binary classification are examined. The classifier is also extended concerning various aspects, such as data type, measurement, and ensemble learning. The validity of the Hilbert-Schmidt inner product, which becomes the squared overlap for pure states, as a positive definite and symmetric kernel is explicitly shown, thereby connecting the quantum binary classifier and kernel methods.     

Respiratory motion-corrected proton magnetic resonance spectroscopy of the liver

Purpose

To develop a post-processing, respiratory-motion correction algorithm for magnetic resonance spectroscopy (MRS) of the liver and to determine the incidence and impact of respiratory motion in liver MRS.

Materials and Methods

One hundred thirty-two subjects (27 healthy, 31 with nonalcoholic fatty liver disease and 74 HIV-infected with or without hepatitis C) were scanned with free breathing MRS at 1.5 T. Two spectral time series were acquired on an 8-ml single voxel using TR/TE=2500 ms/30 ms and (1) water suppression, 128 acquisitions, and (2) no water suppression, 8 acquisitions. Individual spectra were phased and frequency aligned to correct for intrahepatic motion. Next, water peaks more than 50% different from the median water peak area were identified and removed, and remaining spectra averaged to correct for presumed extrahepatic motion. Total CH₂+CH₃ lipids to unsuppressed water ratios were compared before and after corrections.

Results

Intrahepatic-motion correction increased the signal to noise ratio (S/N) in all cases (median=11-fold). Presumed extrahepatic motion was present in 41% (54/132) of the subjects. Its correction altered the lipids/water magnitude (magnitude change: median=2.6%, maximum=290%, and was >5% in 25% of these subjects). The incidence and effect of respiratory motion on lipids/water magnitude were similar among the three groups.

Conclusion

Respiratory-motion correction of free breathing liver MRS greatly increased the S/N and, in a significant number of subjects, changed the lipids/water ratios, relevant for monitoring subjects.     

Accurate 1H tumor spectra quantification from acquisitions without water suppression

Monovoxel magnetic resonance spectroscopy (MRS) is a technique extensively used for the study of brain tumors in many imaging centers. However, given the fact that monovoxel spectrum quality depends upon voxel size, region of acquisition and the presence of metal and/or blood residue after surgery can make the comparison of MRS brain tumor spectra more difficult than that of other pathologies. This study was conducted in order to evaluate whether it is possible to predict in which cases a tumor spectrum will be quantifiable from acquisitions obtained without water suppression, allowing comparison to other spectra. Three different methods were employed: a qualitative, clinical method and two quantitative ones (Amares and Quest). It was found that by using Quest, it is possible to estimate the number of acquisitions needed to obtain a quantifiable spectrum before its acquisition, something which was not feasible with Amares (given the base used). On examining the spectra as physicians would, it was found that after a certain number of acquisitions, they did not change. The study shows that it is possible to optimize MRS acquisition time in brain tumors and guarantee spectrum quantification for comparison of different MRS studies, obtained both from a single patient or different patients.     

PCA和SPA的近红外光谱识别白菜种子品种研究

为了实现对不同品种白菜种子的快速无损鉴别,应用近红外光谱技术获取白菜种子的光谱反射率,首先采用变量标准化校正和多元散射校正对原始光谱进行预处理;其次,采用主成分分析法(PCA)对光谱数据进行聚类分析,从定性分析的角度得到三种不同白菜种子的特征差异,并采用连续投影算法(SPA)选取特征波长;最后,分别基于全波段光谱、PCA分析得到的前3个主成分变量以及SPA算法选取的特征波长,建立了最小二乘支持向量机(LS-SVM)和偏最小二乘判别(PLS-DA)模型进行白菜种子不同品种的鉴别。从主成分PC1、PC2得分图中可以看出,主成分1和2对不同种类白菜种子具有很好的聚类作用。基于特征波长建立的PLS-DA和LS-SVM模型的判别结果优于基于主成分变量建立的模型,其中基于特征波长建立的LS-SVM模型识别效果最优,建模集和预测集的品种识别率均达到100%。结果表明,通过SPA算法选取的6个特征波长变量能够很好的反映光谱信息,提出的SPA算法结合LS-SVM预测模型能获得满意的分类结果,为白菜种子品种的识别提供了一种新方法。     

基于图像和光谱融合的脐橙货架期高光谱成像无损检测研究

水果货架期是影响水果品质的重要因素之一,快速无损检测货架期是消费者、食品加工企业日益关心的问题,为了探讨水果不同货架期的预测判别方法的可行性,以不同货架期脐橙为实验样品,运用高光谱成像技术并结合化学计量学方法对不同货架期脐橙进行了预测判别。分别采集脐橙货架期第0天、第7天、14天后的脐橙样本高光谱图像,并进行高光谱图像校正。从光谱角度,提取脐橙样本的平均光谱,每条光谱有176个波长点;从图像角度,先提取脐橙样本的RGB和HSI颜色空间中R,G,B,H,S和I特征值,得到6个分量的均值,然后提取灰度共生矩阵的能量、熵、对比度、逆差矩、相关性的5个图像纹理信息,一共11个图像特征值,并将图像特征进行归一化处理;结合光谱和图像信息,即176个原始光谱和11个图像信息一共187个特征值。利用光谱信息、图像信息、光谱和图像融合信息进行建模,分别建立偏最小二乘支持向量机（LS-SVM）和偏最小二乘判别（PLS-DA）模型。当原始176个光谱变量作为输入变量,核函数为LIN-Kernel时,LS-SVM模型预测效果最佳,预测集误判率为5.33%。当11个图像特征变量作为输入变量,核函数为LIN-Kernel时,LS-SVM模型预测效果最佳,预测集误判率较高为20%。当原始176个光谱变量和11个图像特征变量的融合特征作为输入变量,核函数为LIN-Kernel时,LS-SVM模型预测效果最佳,预测集误判率为1.33%。实验结果表明,以光谱和图像融合信息建立LS-SVM模型效果最优,提高了对不同货架期脐橙识别的正确率,可实现对不同货架期的脐橙准确有效分类识别,误判率为1.33%。利用高光谱成像技术对不同货架期脐橙进行快速判别,对消费者购买新鲜水果和水果深加工企业具有一定程度的理论指导,也为后期相关仪器研发奠定了基础。     

基于LS-SVM紫外可见光谱检测水产养殖水体COD研究

采用紫外可见(ultraviolet/visible,UV/Vis)光谱技术对水体中有机物浓度的指标化学需氧量(chemical oxygen demand,COD)进行快速检测,将收集到的135份水样进行UV/VIS波段全光谱扫描,应用Savitzky-Golay (SG)平滑算法,经验模态分解算法(empirical modedecomposition,EMD)和小波分析(wavelet transform,WT)对提取出的光谱数据进行去除噪声处理,为了简化模型,PLSR建模得到的6个潜在变量(LVs)作为偏最小二乘支持向量机(LS-SVM)的输入建立COD预测模型,LS-SVM模型的预测集决定系数r2为0.82,预测均方根误差RMSEP为14.82 mg·L-1。说明使用LVs作为LS-SVM建模输入,可以准确快速检测水产养殖水体中的COD含量,为将来实现水产养殖水质COD含量的在线检测以及其他水质参数的快速测定奠定了基础。     

利用基于互信息的不平衡分类方法识别稀有光谱

从海量恒星光谱中发现稀有光谱是天文学研究的重要课题之一。与一般光谱相比,稀有光谱数量较少,因此,传统分类方法无法正常工作。究其原因是这些方法不仅在分类决策时并未对稀有光谱予以更多关注,而且只关注分类的准确率。鉴于此,在总结当前分类方法的基础上,深入分析互信息与决策树之间的关系,提出基于互信息的代价缺失决策树。SDSS DR8中K型、F型、G型以及M型恒星光谱上的比较实验表明,与传统分类方法相比,所提方法能够较好地完成稀有光谱识别的任务。     

LS-SVM的梨可溶性固形物近红外光谱检测的特征波长筛选

为提高梨可溶性固形物含量(soluble solids content,SSC)的近红外光谱模型的精度和稳定性,以160个梨样品为实验对象,分别对原始光谱、多元散射校正(MSC)和标准正态变量变换(SNV)处理后的光谱,经无信息变量消除算法(UVE)挑选后,再结合遗传算法(GA)和连续投影算法(SPA),筛选梨可溶性固形物的近红外光谱特征波长。将筛选后的波长作为输入变量建立梨可溶性固形物的最小二乘支持向量机(LS-SVM)模型。结果表明经过SNV-UVE-GA-SPA从全波段3112个波长中筛选出的30个特征波长建立的梨可溶性固形物LS-SVM模型效果最好,该模型的预测集相关系数(R_p)和预测均方根误差(RMSEP)分别为0.956和0.271。该模型简单可靠,预测效果好,能满足梨的可溶性固形物含量的快速检测,为在线检测和便携式设备开发提供了理论基础。     

Feasibility of in-vivo semi-LASER renal magnetic resonance spectroscopy (MRS): Pilot study in healthy volunteers

PurposeTo evaluate the feasibility of semi-LASER renal magnetic resonance spectroscopy (MRS) in healthy volunteers and establish signature chemical composition of normal renal tissue towards future application for renal carcinoma characterization and grading.Materials and methods14 healthy volunteers were recruited after informed consent. Single voxel ¹H spectra were acquired on a 3 T MRI system using a semi-LASER sequence, employing outer-volume suppression and VAPOR water suppression with multiple averages in multiple breath-holds. Off-line processing and automatic correction for zero-order phase and frequency using the water resonance or residual water resonance for water-suppressed acquisitions was performed.Results11 volunteers successfully completed the entire examination. Phase and frequency correction was necessary to obtain optimal data quality prior to signal summation in few datasets. No lipid resonance was observed in any spectra from the unsuppressed water acquisitions, either in individual transients or in corrected summed spectra opposed to previously reported studies. No signal from other metabolites, such as choline-containing compounds, was observed in any dataset.ConclusionSemi-LASER renal MRS is technically feasible. Normal renal parenchyma does not demonstrate detectable levels of lipid or choline. This may provide a reference point for future application of this technique for noninvasive renal carcinoma histologic subtype characterization and grade.     

基于共焦显微拉曼的柑橘黄龙病无损检测研究

黄龙病危害柑橘果树日益严重,对柑橘黄龙病进行快速检测研究具有重大意义。采用拉曼光谱技术,结合偏最小二乘判别分析(PLS-DA)方法探讨快速诊断柑橘黄龙病及病情类别的可行性。获取柑橘叶片拉曼光谱并进行普通PCR鉴别分为轻度、中度、重度、缺素和正常5类。在715～1 639.5 cm-1范围内采用一阶导,基线校正(Baseline)和多项式拟合三种方法扣除光谱背景,突显叶片拉曼光谱特征峰。多项式拟合方法分别进行了2次,3次和4次拟合,与一阶导和基线校正两种扣除背景方法进行比较,结合最小二乘支持向量机(LS-SVM)和偏最小二乘判别分析(PLS-DA)建立判别模型。经比较发现,多项式拟合方法扣除光谱背景效果均好于另外两种方法,其中用2次多项式拟合的PLS-DA模型的效果最好,预测相关系数(R_P)为0.98,预测均方根误差(RMSEP)为0.67,总误判率最小为0。基线校正扣除光谱背景的LS-SVM模型效果最差,总误判率最大为40%。研究结果表明,利用拉曼光谱技术对柑橘黄龙病进行快速识别研究具有一定的可行性,为柑橘黄龙病无损检测研究提供一种新途径。     

Combined multi-kernel support vector machine and wavelet analysis for hyperspectral remote sensing image classification

Many remote sensing image classifiers are limited in their ability to combine spectral features with spatial features.Multi-kernel classifiers,however,are capable of integrating spectral features with spatial or structural features using multiple kernels and summing them for final outputs.Using a support vector machine(SVM) as classifier,different multi-kernel classifiers are constructed and tested using 64-band Operational Modular Imaging Spectrometer II hyperspectral image of Changping Area,Beijing City.Results show that by integrating spectral and wavelet texture information,multi-kernel SVM classifiers can obtain more accurate classification results than sole-kernel SVM classifiers and cross-information SVM kernel classifiers.Moreover,when the multi-kernel SVM classifier is used,the combination of the first four principal components from principal component analysis and wavelet texture provides the highest accuracy(97.06%).Multi-kernel SVM is therefore an effective approach to improve the accuracy of hyperspectral image classification and to expand possibilities for remote sensing image interpretation and application.     

基于二维相关光谱分析的纯棉与丝光棉制品鉴别分析

纯棉与丝光棉制品是日常生活中常用的两种纤维制品,但是由于二者在物理结构和化学结构上非常相似,以至于使用一些简单的方法难以准确识别一部分纯棉与丝光棉制品。提出一种使用水含量作为扰动的二维相关光谱结合机器学习方法来对二者进行鉴别的新方法。共使用从专业机构获得的200个标准样本来设计实验对新方法进行验证,其中包括100个纯棉样本与100个丝光棉样本。对每一个样本,使用水含量作为扰动,分4次改变样本水含量并采集该水含量下样本的一维光谱,其中4次的水含量分别为20.20%,14.52%,7.77%与0%。根据四条不同的一维构造每一个样本的动态光谱,再通过二维相关算法来计算其同步二维相关光谱,从该同步二维相关光谱中使用移动窗口技术提取三组不同的分类特征,每组特征分别对应一个设计好的支持向量机(SVM)分类器。之后本文提出一种基于信息熵的多分类器融合方法,根据权值不同,将三个分类器融合为一个具有更优效果的强分类器。为了验证方法的准确性与有效性,设计了严谨的实验对方法进行验证。实验首先按照传统的从一维光谱中提取特征的方法对纯棉与丝光棉样本进行鉴别,使用两种样本各50个来进行分类模型建立,剩余的进行模型验证,分类效果最高只有76%。但是基于从二维相关光谱中提取的三组特征设计的三个支持向量机(SVM)分类器的准确率分别可以达到88%,90%,88%,最后根据提出的基于信息熵的多分类器信息融合方法将三个分类器进行融合同一可以得到92%的分类准确率,比三个基础分类器准确率都有提升。与从一维光谱中提取特征并设计分类器进行分别鉴别相比,从二维相关光谱中提取特征设计多个分类器并使用基于信息熵的多分类器信息融合方法进行分类鉴别具有更高的分类准确率。二维相关光谱将光谱信息扩展到更高的维度,将一维光谱中隐藏的折叠峰进行展开,因此具有更高的分类准确率。提出的方法是一种快速准确鉴别纯棉与丝光棉制品的新方法。     

An Approach to Growth Delimitation of Straight Line Segment Classifiers Based on a Minimum Bounding Box

Several supervised machine learning algorithms focused on binary classification for solving daily problems can be found in the literature. The straight-line segment classifier stands out for its low complexity and competitiveness, compared to well-knownconventional classifiers. This binary classifier is based on distances between points and two labeled sets of straight-line segments. Its training phase consists of finding the placement of labeled straight-line segment extremities (and consequently, their lengths) which gives the minimum mean square error. However, during the training phase, the straight-line segment lengths can grow significantly, giving a negative impact on the classification rate. Therefore, this paper proposes an approach for adjusting the placements of labeled straight-line segment extremities to build reliable classifiers in a constrained search space (tuned by a scale factor parameter) in order to restrict their lengths. Ten artificial and eight datasets from the UCI Machine Learning Repository were used to prove that our approach shows promising results, compared to other classifiers. We conclude that this classifier can be used in industry for decision-making problems, due to the straightforward interpretation and classification rates.     

基于图像局部区域梯度特征描述的红外人体识别算法

提出了一种新的红外图像中人体目标识别方案并进行了算法实现。通过直方图聚类分析对红外图像进行分割,根据二值化图像团块的特点,确定图像中的候选目标图像区域。将候选目标图像按比例划分为多个区域,使用梯度位置朝向直方图(GLOH,Gradient location-orientation histogram)对候选目标图像进行描述。与其它红外图像中人体识别算法相比,不需要多种特征提取算法组合进行分步骤识别,仅使用单个SVM分类器即可达到满意的识别率,避免了分类器的级联,算法简单有效。     

变量优选补正算法的鲜枣可溶性固形物检测模型传递方法研究

在水果的品质检测和分级分选中,存在不同仪器所建检测模型难以共享的难题。为此,以壶瓶枣为研究对象,利用可见/近红外光谱技术探讨仪器间可溶性固形物含量（SSC）检测模型的传递方法。首先,采用美国ASD（Analytical Spectral Device）公司生产的两台仪器采集样本的光谱信息,采用最小二乘支持向量机（LS-SVM）建立原始光谱、Savitzky-Golay一阶导数处理、标准正态变量变换后的SSC检测模型,预测不同仪器采集的光谱时3种方法的预测能力均较差。预测同一台仪器的光谱时,基于原始光谱的主仪器所建模型最优,预测集的决定系数（R2_p）和均方根误差（RMSEP）分别为0.73和1.36%。在此基础上,采用Kennard/Stone算法选取标样,利用专利算法（Shenk’s）、直接标准化（DS）、斜率/偏差算法（S/B）进行模型传递。然后,根据回归系数提取主仪器（24个）和从仪器（28个）的特征波长,优选出单一变量（SV）24个、共性变量（CV）23个、融合变量（FV）29个,均涵盖了SSC的主要吸收谱带。利用优选的变量分别建立主仪器的LS-SVM检测模型,采用主仪器的预测结果（R2_p=0.78~0.80,RMSEP=1.07%~1.13%）明显好于全波段所建模型,但预测从仪器时RMSEP为6.62%~7.88%,模型失效。最后,基于波长位置偏移和分子振动的吸收特性提出了共性变量优选结合差值补正（CV-MC）、单一变量优选结合差值补正、融合变量优选结合差值补正、共性变量优选结合波长补正算法（CV-WC）进行模型传递,并与SV-Shenk’s,CV-Shenk’s,FV-Shenk’s,SV-DS,CV-DS,FV-DS,SV-S/B,CV-S/B和FV-S/B进行对比分析。结果表明,基于全波段进行模型传递时,预测结果均较差（R2_p=0.03~0.34,RMSEP=2.44%~4.67%）;基于优选变量所建模型经SV-Shenk’s,CV-Shenk’s,FV-Shenk’s传递后的结果较差,经其他算法传递后的结果（R2_p=0.47~0.73,RMSEP=1.30%~1.90%）好于全波段;基于共性变量传递后的结果好于单一变量和融合变量,CV-MC结果最佳（R2_p=0.73,RMSEP=1.30%）,CV-WC传递后的预测结果（RMSEP=1.62%）与CV-DS和CV-S/B相近。研究表明,CV-MC和CV-WC均是一种有效模型传递算法,对建立不同仪器间通用的鲜枣品质检测模型具有重要意义。     

稻干尖线虫病胁迫水稻叶片波谱响应特征及识别研究

对植被病害的精确识别是采取植保措施的前提,同时对喷施农药也具有积极的指导作用。比较了受稻干尖线虫胁迫水稻叶片和健康叶片色素含量、光谱反射率、高光谱特征参数,受害水稻叶片与健康叶片相比,叶绿素和类胡萝卜素含量分别降低18%和22%;光谱反射率在蓝紫光、绿光和红光谱段分别增加1.5,1和2.3倍,在近红外和短波红外区域分别降低约28.9%和26.3%,红边和蓝边分别蓝移约8和10nm,绿峰和红谷分别红移约8.5和6 nm。以红边面积和红边位置作为C-SVC(非线性软间隔分类机)的输入向量,对受害和健康叶片进行识别,精度为100%。研究表明,水稻叶片光谱对病害胁迫具有显著的响应特征,利用C-SVC对受害和健康叶片进行辨别的方法是可行的。     

Radiomic prediction of mutation status based on MR imaging of lung cancer brain metastases

Lung cancer metastases comprise most of all brain metastases in adults and most brain metastases are diagnosed by magnetic resonance (MR) scans. The purpose of this study was to conduct an MR imaging-based radiomic analysis of brain metastatic lesions from patients with primary lung cancer to classify mutational status of the metastatic disease. We retrospectively identified lung cancer patients with brain metastases treated at our institution between 2009 and 2017 who underwent genotype testing of their primary lung cancer. Brain MR Images were used for segmentation of enhancing tumors and peritumoral edema, and for radiomic feature extraction. The most relevant radiomic features were identified and used with clinical data to train random forest classifiers to classify the mutation status. Of 110 patients in the study cohort (mean age 57.51 ± 12.32 years; M: F = 37:73), 75 had an EGFR mutation, 21 had an ALK translocation, and 15 had a KRAS mutation. One patient had both ALK translocation and EGFR mutation. Majority of radiomic features most relevant for mutation classification were textural. Model building using both radiomic features and clinical data yielded more accurate classifications than using either alone. For classification of EGFR, ALK, and KRAS mutation status, the model built with both radiomic features and clinical data resulted in area-under-the-curve (AUC) values based on cross-validation of 0.912, 0.915, and 0.985, respectively. Our study demonstrated that MR imaging-based radiomic analysis of brain metastases in patients with primary lung cancer may be used to classify mutation status. This approach may be useful for devising treatment strategies and informing prognosis.     

Experimental Study and Comparison of Imbalance Ensemble Classifiers with Dynamic Selection Strategy

Imbalance ensemble classification is one of the most essential and practical strategies for improving decision performance in data analysis. There is a growing body of literature about ensemble techniques for imbalance learning in recent years, the various extensions of imbalanced classification methods were established from different points of view. The present study is initiated in an attempt to review the state-of-the-art ensemble classification algorithms for dealing with imbalanced datasets, offering a comprehensive analysis for incorporating the dynamic selection of base classifiers in classification. By conducting 14 existing ensemble algorithms incorporating a dynamic selection on 56 datasets, the experimental results reveal that the classical algorithm with a dynamic selection strategy deliver a practical way to improve the classification performance for both a binary class and multi-class imbalanced datasets. In addition, by combining patch learning with a dynamic selection ensemble classification, a patch-ensemble classification method is designed, which utilizes the misclassified samples to train patch classifiers for increasing the diversity of base classifiers. The experiments’ results indicate that the designed method has a certain potential for the performance of multi-class imbalanced classification.