Screening and diagnosing of abnormal Leukocytes are crucial for the diagnosis of immune diseases and Acute Lymphoblastic Leukemia(ALL).As the deterioration of abnormal leukocytes is mainly due to the changes in the ch...Screening and diagnosing of abnormal Leukocytes are crucial for the diagnosis of immune diseases and Acute Lymphoblastic Leukemia(ALL).As the deterioration of abnormal leukocytes is mainly due to the changes in the chromatin distribution,which signicantly affects the absorption and reflection of light,the spectral feature is proved to be important for leukocytes classication and identication.This paper proposes an accurate identication method for healthy and abnormal leukocytes based on microscopic hyperspectral imaging(HSI)technology which combines the spectral information.The segmentation of nucleus and cytoplasm is obtained by the morphological watershed algorithm.Then,the spectral features are extracted and combined with the spatial features.Based on this,the support vector machine(SVM)is applied for classication ofve types of leukocytes and abnormal leukocytes.Compared with different classication methods,the proposed method utilizes spectral features which highlight the differences between healthy leukocytes and abnormal leukocytes,improving the accuracy in the classication and identication of leukocytes.This paper only selects one subtype of ALL for test,and the proposed method can be applied for detection of other leukemia in the future.展开更多
Accurate histopathology classification is a crucial factor in the diagnosis and treatment of Cholangiocarcinoma(CCA).Hyperspectral images(HSI)provide rich spectral information than ordinary RGB images,making them more...Accurate histopathology classification is a crucial factor in the diagnosis and treatment of Cholangiocarcinoma(CCA).Hyperspectral images(HSI)provide rich spectral information than ordinary RGB images,making them more useful for medical diagnosis.The Convolutional Neural Network(CNN)is commonly employed in hyperspectral image classification due to its remarkable capacity for feature extraction and image classification.However,many existing CNN-based HSI classification methods tend to ignore the importance of image spatial context information and the interdependence between spectral channels,leading to unsatisfied classification performance.Thus,to address these issues,this paper proposes a Spatial-Spectral Joint Network(SSJN)model for hyperspectral image classification that utilizes spatial self-attention and spectral feature extraction.The SSJN model is derived from the ResNet18 network and implemented with the non-local and Coordinate Attention(CA)modules,which extract long-range dependencies on image space and enhance spatial features through the Branch Attention(BA)module to emphasize the region of interest.Furthermore,the SSJN model employs Conv-LSTM modules to extract long-range depen-dencies in the image spectral domain.This addresses the gradient disappearance/explosion phenom-ena and enhances the model classification accuracy.The experimental results show that the pro-posed SSJN model is more efficient in leveraging the spatial and spectral information of hyperspec-tral images on multidimensional microspectral datasets of CCA,leading to higher classification accuracy,and may have useful references for medical diagnosis of CCA.展开更多
A microscopic hyperspectral imager was developed based on the microscopic technology and the spectral imaging technology. Some microscopic hyperspectral images of retina sections of the normal, the diabetic, and the t...A microscopic hyperspectral imager was developed based on the microscopic technology and the spectral imaging technology. Some microscopic hyperspectral images of retina sections of the normal, the diabetic, and the treated rats were collected by the new imager. Single-band images and pseudo-color images of each group were obtained and the typical transmittance spectrums were ex-tracted. The results showed that the transmittance of outer nuclear layer cells of the diabetic group was generally higher than that of the normal. A small absorption peak appeared near the 180th band in the spectrum of the diabetic group and this peak weakened or disappeared in the spectrum of the treated group. Our findings indicate that the microscopic hyperspectral images include wealthy information of retina sections which is helpful for the ophthalmologist to reveal the pathogenesis of diabetic reti-nopathy and explore the therapeutic effect of drugs.展开更多
In dendroclimatology,tree ring chronology is ordinarily established to reveal the fluctuation law of climate change on the interannual,interdecadal,and centennial scales.However,since traditional dendrochronology can ...In dendroclimatology,tree ring chronology is ordinarily established to reveal the fluctuation law of climate change on the interannual,interdecadal,and centennial scales.However,since traditional dendrochronology can only use one variable(tree ring width)to reflect environmentally related information,this causes the richer information recorded in the tree rings to be discarded.In this study,we examined the potential of hyperspectral chronological indices(shortened as“hyperspectral index/indices”)with samples collected in Shennongjia woodland in central China.The correlation analysis of the tree ring series on different samples indicated that hyperspectral indices outperform the traditional width index in chronology statistics including Signal-to-noise Ratio(SNR)and Expressed Population Signal(EPS).The reliability test shows that hyperspectral chronologies have more periods reaching the threshold of EPS or Subsample Signal Strength(SSS)>0.85,which means that hyperspectral chronologies provide more reliable periods for accurate climate reconstruction.Based on this,chronologies built by the three dendroclimatic indices were used to reconstruct the average temperature changes in Shennongjia over the last 103 years.The reconstruction results indicate that in our study area,the traditional width index model failed the split-sample calibration test and exhibited a low reconstruction accuracy,while the hyperspectral index model has a higher explained variance of 46.4%(p<0.01),compared to the width index(21.4%)and the grayscale index(38.3%).Our research results show that hyperspectral indices have greater potential for climate reconstruction in regions with lower susceptibility to climate stress.This is attributed to their ability to effectively extract subtle climate signals from the spectral variations on the surface of tree rings.Such ring spectral changes may be caused by complex and currently unknown responses of the trees to the climate.展开更多
Bladder urothelial carcinoma is the most common malignant tumor disease in urinary system,and its incidence rate ranks ninth in the world.In recent years,the continuous development of hyperspectral imaging technology ...Bladder urothelial carcinoma is the most common malignant tumor disease in urinary system,and its incidence rate ranks ninth in the world.In recent years,the continuous development of hyperspectral imaging technology has provided a new tool for the auxiliary diagnosis of bladder cancer.In this study,based on microscopic hyperspectral data,an automatic detection algorithm of bladder tumor cells combining color features and shape features is proposed.Support vector machine(SVM)is used to build classification models and compare the classification performance of spectral feature,spectral and shape fusion feature,and the fusion feature proposed in this paper on the same classifier.The results show that the sensitivity,specificity,and accuracy of our classification algorithm based on shape and color fusion features are 0.952,0.897,and 0.920,respectively,which are better than the classification algorithm only using spectral features.Therefore,this study can effectively extract the cell features of bladder urothelial carcinoma smear,thus achieving automatic,real-time,and noninvasive detection of bladder tumor cells,and then helping doctors improve the efficiency of pathological diagnosis of bladder urothelial cancer,and providing a reliable basis for doctors to choose treatment plans and judge the prognosis of the disease.展开更多
The incidence of breast cancer is tending younger globally,and tumor development,clinical treatment,and prognosis are largely influenced by histopathological diagnosis.For diagnosed patients,the distinction between th...The incidence of breast cancer is tending younger globally,and tumor development,clinical treatment,and prognosis are largely influenced by histopathological diagnosis.For diagnosed patients,the distinction between the cancer nests and normal tissue is the basis of breast cancer treatment.Microscopic hyperspectral imaging technology has shown its potential in auxiliary pathological examinations due to the superior imaging modality and data characteristics.This paper presents a method for cancer nest segmentation from hyperspectral images of breast cancer tissue microarray samples.The scheme combines the strengths of the U-Net neural network and unsupervised principal component analysis,which reduces the amount of calculation and improves the recognition accuracy.The experimental accuracy of cancer nest segmentation reaches 87.14%.Furthermore,a set of quantitative pathological characteristic parameters reflects the degree of breast cancer lesions from multiple angles,providing a relatively comprehensive reference for the pathologist’s diagnosis.In-depth exploration of the combined development of deep learning and microscopic hyperspectral imaging technology is worthy to promote efficient diagnosis of breast tumors and concern for human health.展开更多
基金supported in part by the National Natural Science Foundation of China(Grant Nos.61975056 and 61901173)the Shanghai Natural Science Foundation(Grant No.19ZR1416000)the Science and Technology Commission of Shanghai Municipality(Grant Nos.14DZ2260800 and 18511102500).
文摘Screening and diagnosing of abnormal Leukocytes are crucial for the diagnosis of immune diseases and Acute Lymphoblastic Leukemia(ALL).As the deterioration of abnormal leukocytes is mainly due to the changes in the chromatin distribution,which signicantly affects the absorption and reflection of light,the spectral feature is proved to be important for leukocytes classication and identication.This paper proposes an accurate identication method for healthy and abnormal leukocytes based on microscopic hyperspectral imaging(HSI)technology which combines the spectral information.The segmentation of nucleus and cytoplasm is obtained by the morphological watershed algorithm.Then,the spectral features are extracted and combined with the spatial features.Based on this,the support vector machine(SVM)is applied for classication ofve types of leukocytes and abnormal leukocytes.Compared with different classication methods,the proposed method utilizes spectral features which highlight the differences between healthy leukocytes and abnormal leukocytes,improving the accuracy in the classication and identication of leukocytes.This paper only selects one subtype of ALL for test,and the proposed method can be applied for detection of other leukemia in the future.
基金supported by National Natural Science Foundation of China(No.62101040).
文摘Accurate histopathology classification is a crucial factor in the diagnosis and treatment of Cholangiocarcinoma(CCA).Hyperspectral images(HSI)provide rich spectral information than ordinary RGB images,making them more useful for medical diagnosis.The Convolutional Neural Network(CNN)is commonly employed in hyperspectral image classification due to its remarkable capacity for feature extraction and image classification.However,many existing CNN-based HSI classification methods tend to ignore the importance of image spatial context information and the interdependence between spectral channels,leading to unsatisfied classification performance.Thus,to address these issues,this paper proposes a Spatial-Spectral Joint Network(SSJN)model for hyperspectral image classification that utilizes spatial self-attention and spectral feature extraction.The SSJN model is derived from the ResNet18 network and implemented with the non-local and Coordinate Attention(CA)modules,which extract long-range dependencies on image space and enhance spatial features through the Branch Attention(BA)module to emphasize the region of interest.Furthermore,the SSJN model employs Conv-LSTM modules to extract long-range depen-dencies in the image spectral domain.This addresses the gradient disappearance/explosion phenom-ena and enhances the model classification accuracy.The experimental results show that the pro-posed SSJN model is more efficient in leveraging the spatial and spectral information of hyperspec-tral images on multidimensional microspectral datasets of CCA,leading to higher classification accuracy,and may have useful references for medical diagnosis of CCA.
文摘A microscopic hyperspectral imager was developed based on the microscopic technology and the spectral imaging technology. Some microscopic hyperspectral images of retina sections of the normal, the diabetic, and the treated rats were collected by the new imager. Single-band images and pseudo-color images of each group were obtained and the typical transmittance spectrums were ex-tracted. The results showed that the transmittance of outer nuclear layer cells of the diabetic group was generally higher than that of the normal. A small absorption peak appeared near the 180th band in the spectrum of the diabetic group and this peak weakened or disappeared in the spectrum of the treated group. Our findings indicate that the microscopic hyperspectral images include wealthy information of retina sections which is helpful for the ophthalmologist to reveal the pathogenesis of diabetic reti-nopathy and explore the therapeutic effect of drugs.
基金supported by the National Natural Science Foundation of China(NSFC)Projects[grant numbers 42271476 and 41771227]Key Technology Projects of the Hubei Provincial Company of the China National Tobacco Corporation(grant number 027Y2021-020 and 027Y2022-006)Young Scholar of Wuhan University 351 Talent Program[grant number 202017].
文摘In dendroclimatology,tree ring chronology is ordinarily established to reveal the fluctuation law of climate change on the interannual,interdecadal,and centennial scales.However,since traditional dendrochronology can only use one variable(tree ring width)to reflect environmentally related information,this causes the richer information recorded in the tree rings to be discarded.In this study,we examined the potential of hyperspectral chronological indices(shortened as“hyperspectral index/indices”)with samples collected in Shennongjia woodland in central China.The correlation analysis of the tree ring series on different samples indicated that hyperspectral indices outperform the traditional width index in chronology statistics including Signal-to-noise Ratio(SNR)and Expressed Population Signal(EPS).The reliability test shows that hyperspectral chronologies have more periods reaching the threshold of EPS or Subsample Signal Strength(SSS)>0.85,which means that hyperspectral chronologies provide more reliable periods for accurate climate reconstruction.Based on this,chronologies built by the three dendroclimatic indices were used to reconstruct the average temperature changes in Shennongjia over the last 103 years.The reconstruction results indicate that in our study area,the traditional width index model failed the split-sample calibration test and exhibited a low reconstruction accuracy,while the hyperspectral index model has a higher explained variance of 46.4%(p<0.01),compared to the width index(21.4%)and the grayscale index(38.3%).Our research results show that hyperspectral indices have greater potential for climate reconstruction in regions with lower susceptibility to climate stress.This is attributed to their ability to effectively extract subtle climate signals from the spectral variations on the surface of tree rings.Such ring spectral changes may be caused by complex and currently unknown responses of the trees to the climate.
基金Bethune Medical Engineering and Instrument Center Fund(E10133Y8H0)Jilin province science and technology development plan project(20210204216YY,20210204146YY).
文摘Bladder urothelial carcinoma is the most common malignant tumor disease in urinary system,and its incidence rate ranks ninth in the world.In recent years,the continuous development of hyperspectral imaging technology has provided a new tool for the auxiliary diagnosis of bladder cancer.In this study,based on microscopic hyperspectral data,an automatic detection algorithm of bladder tumor cells combining color features and shape features is proposed.Support vector machine(SVM)is used to build classification models and compare the classification performance of spectral feature,spectral and shape fusion feature,and the fusion feature proposed in this paper on the same classifier.The results show that the sensitivity,specificity,and accuracy of our classification algorithm based on shape and color fusion features are 0.952,0.897,and 0.920,respectively,which are better than the classification algorithm only using spectral features.Therefore,this study can effectively extract the cell features of bladder urothelial carcinoma smear,thus achieving automatic,real-time,and noninvasive detection of bladder tumor cells,and then helping doctors improve the efficiency of pathological diagnosis of bladder urothelial cancer,and providing a reliable basis for doctors to choose treatment plans and judge the prognosis of the disease.
基金funded by National Natural Science Foundation of China(Grant No.61975056)the Shanghai Natural Science Foundation(Grant No.19ZR1416000)the Science and Technology Commission of Shanghai Municipality(Grants No.20440713100,19511120100,18DZ2270800).
文摘The incidence of breast cancer is tending younger globally,and tumor development,clinical treatment,and prognosis are largely influenced by histopathological diagnosis.For diagnosed patients,the distinction between the cancer nests and normal tissue is the basis of breast cancer treatment.Microscopic hyperspectral imaging technology has shown its potential in auxiliary pathological examinations due to the superior imaging modality and data characteristics.This paper presents a method for cancer nest segmentation from hyperspectral images of breast cancer tissue microarray samples.The scheme combines the strengths of the U-Net neural network and unsupervised principal component analysis,which reduces the amount of calculation and improves the recognition accuracy.The experimental accuracy of cancer nest segmentation reaches 87.14%.Furthermore,a set of quantitative pathological characteristic parameters reflects the degree of breast cancer lesions from multiple angles,providing a relatively comprehensive reference for the pathologist’s diagnosis.In-depth exploration of the combined development of deep learning and microscopic hyperspectral imaging technology is worthy to promote efficient diagnosis of breast tumors and concern for human health.
