Yuji Hatanaka - Academia.edu (original) (raw)
Papers by Yuji Hatanaka
Since the evaluation is influenced by the case selection, it is necessary to apply some common da... more Since the evaluation is influenced by the case selection, it is necessary to apply some common databases in the performance evaluation. To evaluate our computer-aided diag- nosis (CAD) system for detecting masses, 320 images from the Mammogaphy Image Analy- sis Society (MIAS) database in UK were applied in this study. Our algorithm for detecting masses was based on a standard adaptive thresholding technique which had been developed in our group. However, the preliminary result was not as well as those for a Japanese database. After we adjusted some thresholding values of our system, a 90% sensitivity with 0.8 false positive (FP) was achieved which indicated that our scheme was effective for the different databases in both Japan and UK. Moreover, the differences between the MIAS and a Japa- nese database were also discussed.
Automated blood vessels detection on retinal images is an important process in the development of... more Automated blood vessels detection on retinal images is an important process in the development of pathologies analysis systems. This paper describes about an automated blood vessel extraction using high-order local autocorrelation (HLAC) on retinal images. Although HLAC features are shift-invariant, HLAC features are weak to turned image. Therefore, a method was improved by the addition of HLAC features to a polar transformed image. We have proposed a method using HLAC, pixel-based-features and three filters. However, we have not investigated about feature selection and machine learning method. Therefore, this paper discusses about effective features and machine learning method. We tested eight methods by extension of HLAC features, addition of 4 kinds of pixel-based features, difference of preprocessing techniques, and 3 kinds of machine learning methods. Machine learning methods are general artificial neural network (ANN), a network using two ANNs, and Boosting algorithm. As a res...
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, Jul 1, 2017
The pattern of blood vessels in the eye is unique to each person because it rarely changes over t... more The pattern of blood vessels in the eye is unique to each person because it rarely changes over time. Therefore, it is well known that retinal blood vessels are useful for biometrics. This paper describes a biometrics method using the Jaccard similarity coefficient (JSC) based on blood vessel regions in retinal image pairs. The retinal image pairs were rough matched by the center of their optic discs. Moreover, the image pairs were aligned using the Iterative Closest Point algorithm based on detailed blood vessel skeletons. For registration, perspective transform was applied to the retinal images. Finally, the pairs were classified as either correct or incorrect using the JSC of the blood vessel region in the image pairs. The proposed method was applied to temporal retinal images, which were obtained in 2009 (695 images) and 2013 (87 images). The 87 images acquired in 2013 were all from persons already examined in 2009. The accuracy of the proposed method reached 100%.
SPIE Proceedings, 2017
Early detection of glaucoma is important to slow down progression of the disease and to prevent t... more Early detection of glaucoma is important to slow down progression of the disease and to prevent total vision loss. We have been studying an automated scheme for detection of a retinal nerve fiber layer defect (NFLD), which is one of the earliest signs of glaucoma on retinal fundus images. In our previous study, we proposed a multi-step detection scheme which consists of Gabor filtering, clustering and adaptive thresholding. The problems of the previous method were that the number of false positives (FPs) was still large and that the method included too many rules. In attempt to solve these problems, we investigated the end-to-end learning system without pre-specified features. A deep convolutional neural network (DCNN) with deconvolutional layers was trained to detect NFLD regions. In this preliminary investigation, we investigated effective ways of preparing the input images and compared the detection results. The optimal result was then compared with the result obtained by the previous method. DCNN training was carried out using original images of abnormal cases, original images of both normal and abnormal cases, ellipse-based polar transformed images, and transformed half images. The result showed that use of both normal and abnormal cases increased the sensitivity as well as the number of FPs. Although NFLDs are visualized with the highest contrast in green plane, the use of color images provided higher sensitivity than the use of green image only. The free response receiver operating characteristic curve using the transformed color images, which was the best among seven different sets studied, was comparable to that of the previous method. Use of DCNN has a potential to improve the generalizability of automated detection method of NFLDs and may be useful in assisting glaucoma diagnosis on retinal fundus images.
Medical Imaging 2016: Computer-Aided Diagnosis, 2016
Early detection of glaucoma is important to slow down or cease progression of the disease and for... more Early detection of glaucoma is important to slow down or cease progression of the disease and for preventing total blindness. We have previously proposed an automated scheme for detection of retinal nerve fiber layer defect (NFLD), which is one of the early signs of glaucoma observed on retinal fundus images. In this study, a new multi-step detection scheme was included to improve detection of subtle and narrow NFLDs. In addition, new features were added to distinguish between NFLDs and blood vessels, which are frequent sites of false positives (FPs). The result was evaluated with a new test dataset consisted of 261 cases, including 130 cases with NFLDs. Using the proposed method, the initial detection rate was improved from 82% to 98%. At the sensitivity of 80%, the number of FPs per image was reduced from 4.25 to 1.36. The result indicates the potential usefulness of the proposed method for early detection of glaucoma.
SPIE Proceedings, 2008
Biometric technique has been implemented instead of conventional identification methods such as p... more Biometric technique has been implemented instead of conventional identification methods such as password in computer, automatic teller machine (ATM), and entrance and exit management system. We propose a personal identification (PI) system using color retinal fundus images which are unique to each individual. The proposed procedure for identification is based on comparison of an input fundus image with reference fundus images in the database. In the first step, registration between the input image and the reference image is performed. The step includes translational and rotational movement. The PI is based on the measure of similarity between blood vessel images generated from the input and reference images. The similarity measure is defined as the cross-correlation coefficient calculated from the pixel values. When the similarity is greater than a predetermined threshold, the input image is identified. This means both the input and the reference images are associated to the same person. Four hundred sixty-two fundus images including forty-one same-person's image pairs were used for the estimation of the proposed technique. The false rejection rate and the false acceptance rate were 9.9×10-5 % and 4.3×10-5 %, respectively. The results indicate that the proposed method has a higher performance than other biometrics except for DNA. To be used for practical application in the public, the device which can take retinal fundus images easily is needed. The proposed method is applied to not only the PI but also the system which warns about misfiling of fundus images in medical facilities.
International Congress Series, 2001
We have been developing an automated detection algorithm for masses on mammograms. There were the... more We have been developing an automated detection algorithm for masses on mammograms. There were the cases of missed detections, in most of which the masses existed around thick-mammarygland region or near chest-wall region. The purpose of this study is to develop an algorithm for detecting such masses based on a sector-shape mass-pattern model. After applying the new technique along with our previous method to 3575 digitized mammograms, the detection sensitivity was changed from 78% to 93% at the number of false positives with varying from 1.2 to 1.9 per image. This result indicated that the new technique improved the overall performance of our CAD system for detecting the masses on mammograms effectively.
IEEE Transactions on Medical Imaging, 2010
The detection of microaneurysms in digital color fundus photographs is a critical first step in a... more The detection of microaneurysms in digital color fundus photographs is a critical first step in automated screening for diabetic retinopathy (DR), a common complication of diabetes. To accomplish this detection numerous methods have been published in the past but none of these was compared with each other on the same data. In this work we present the results of the first international microaneurysm detection competition, organized in
This paper describes the purpose and summary of our recent progress and achievement in the resear... more This paper describes the purpose and summary of our recent progress and achievement in the research work, which is a part of the research project “Multidisciplinary Computational Anatomy and Its Application to Highly Intelligent Diagnosis and Therapy,” founded by Grant-in-Aid for Scientific Research on Innovative Areas, MEXT, Japan. The main purposes of our research in this project are to establish a scientific principle of multidisciplinary computational anatomy and to develop computer-aided diagnosis (CAD) systems based on such anatomical models for organ and tissue functions. In FY2015, we proceeded with the preparation for sub projects, such as the database establishment, and started the investigation in earnest. Several progresses have been achieved with the promising results, which encourage the continual effort for success of this project.
IEEE Transactions on Medical Imaging, 2001
Abstract—Recently, we have been developing several automated algorithms for detecting masses on m... more Abstract—Recently, we have been developing several automated algorithms for detecting masses on mammograms.,For our algo- rithm, we devised an adaptive thresholding technique for detecting masses, but our system failed to detect masses with a partial loss of region that were located on the edge of the film. This is a common issue in all of the algorithms developed so far
Glaucoma is the second leading cause of vision loss in the world. Detection of retinal nerve fibe... more Glaucoma is the second leading cause of vision loss in the world. Detection of retinal nerve fiber layer defects (NFLDs), which is one of the early glaucomatous changes, on retinal fundus images obtained in mass screening may prevent patients from becoming permanently blind. In this study, a technique for contrast enhancement and automated detection of NFLDs was investigated. Images used in this study were obtained from the Tajimi general screening database, which includes retinal fundus images obtained in the population glaucoma screening study at Tajimi, Japan. In this preliminary investigation, images with at least one identified NFLD were included. Lesions of NFLDs were individually marked by two ophthalmologists, and those identified by both ophthalmologists were considered as the target NFLDs. First, the blood vessel regions in retinal fundus images were identified and interpolated by the surrounding pixels for creating “blood-vessel-erased” images. The resulted color images w...
Proceedings of SPIE - The International Society for Optical Engineering
Glaucoma is one of the leading causes of blindness in Japan and the US. One of the indices for di... more Glaucoma is one of the leading causes of blindness in Japan and the US. One of the indices for diagnosis of glaucoma is the cup-to-disc ratio (CDR). We have been developing a computerized method for measuring CDR on stereo fundus photographs. Although our previous study indicated that the method may be useful, cup determination was not always successful, especially for the normal eyes. In this study, we investigated a new method to quantify the likelihood of glaucomatous disc based on the similarity scores to the glaucoma and non-glaucoma models. Eighty-seven images, including 40 glaucomatous eyes, were used in this study. Only one eye from each patient was used. Using a stereo fundus camera, two images were captured from different angles, and the depth image was created by finding the local corresponding points. One of the characteristics of a glaucomatous disc can be not only that the cup is enlarged but it has an acute slope. On the other hand, a non-glaucomatous cup generally ha...
Microaneurysm is one of diabetic retinopathy (DR) findings. Visual loss can be prevented by early... more Microaneurysm is one of diabetic retinopathy (DR) findings. Visual loss can be prevented by early detection and treatment of DR. The fundus examination is effective for early detection of DR. We also proposed an automated MAs detection method using double-ring filter and shape index based on Hessian matrix in the retinal fundus images. However, it had a problem in that many MAs were lost, although the performance of the detectors had peaked. In this study, we applied a density gradient vector concentration with a point concentration. The microaneurysm candidates included so many false positives, that the candidates were classified into microaneurysms and FPs by using support vector machine with 48 features. For evaluation of this study, images from the ROC (Retinopathy Online Challenge) database were used. This database comprises 50 training images and 50 test images. When this method was evaluated by using the 50 test images, the sensitivity was 67% of 8 false positives per image. ...
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2005
We have developed a computer-aided diagnosis system (CAD) to detect abnormalities in fundus image... more We have developed a computer-aided diagnosis system (CAD) to detect abnormalities in fundus images. In Japan, ophthalmologists usually detect hypertensive changes by identifying arteriolar narrowing and focal arteriolar narrowing. The purpose of this study is to develop an automated method for detecting arteriolar narrowing and focal arteriolar narrowing on fundus images. The blood vessel candidates were detected by the density analysis method. In blood vessel tracking, a local detection function was used to determine the centerline of the blood vessel. A direction comparison function using three vectors was designed to optimally estimate the next possible location of a blood vessel. After the connectivity of vessel segments was adjusted based on the recognized intersections, the true tree-like structure of the blood vessels was established. The blood vessels were recognized as arteries or veins by hue of HSV color space and their diameters. The arteriolar narrowing was detected by ...
![Research paper thumbnail of [A comparison between physicians' interpretation and a CAD system's cancer detection by using a mammogram database in a physicians' self-learning course]](https://attachments.academia-assets.com/54881432/thumbnails/1.jpg)
](Nihon Hoshasen Gijutsu Gakkai zasshi, 2002
We have been developing automated detection algorithms for masses and clustered microcalcificatio... more We have been developing automated detection algorithms for masses and clustered microcalcifications in a mammography computer-aided diagnosis (CAD) system. In this study, we investigated the potential of our CAD system by comparing 579 physicians' interpretation results with that of the CAD system's cancer detection for 100 mammograms (21 malignant and 29 benign cases) employed in a physicians' self-learning course. As a result, our CAD system detected 7 out of 8 malignant lesions whose physicians' averaged sensitivity was less than 60%. Although the average of physicians' sensitivities were 76% (about 16 cases), the CAD system's detection rate was 90% (19 cases). Sensitivity was raised up to 97% if the physicians' interpretation and the CAD system's detection result were treated in a matter of logical OR. Thus, it was raised the possibility that even the less-experienced physicians would diagnose with a higher sensitivity by using the computer output...
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2012
Glaucoma is the first leading cause of vision loss in Japan, thus developing a scheme for helping... more Glaucoma is the first leading cause of vision loss in Japan, thus developing a scheme for helping glaucoma diagnosis is important. For this problem, automated nerve fiber layer defects (NFLDs) detection method was proposed, but glaucoma risk assessment using this method was not evaluated. In this paper, computerized risk assessment for having glaucoma was attempted by use of the patients' clinical information, and the performances of the NFLDs detection and the glaucoma risk assessment were compared. The clinical data includes the systemic data, ophthalmologic data, and right and left retinal images. Glaucoma risk assessment was built by using machine learning technique, which were artificial neural network, radial basis function (RBF) network, k-nearest neighbor algorithm, and support vector machine. The inputting parameter was ten clinical ones with/without the results of NFLDs detection. As a result, proposed glaucoma risk assessment showed the higher performance than the NFL...
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2011
Presence of peripapillary chorioretinal atrophy (PPA) is considered one of the risk factors for g... more Presence of peripapillary chorioretinal atrophy (PPA) is considered one of the risk factors for glaucoma. It can be identified as bright regions in retinal fundus images, and therefore, incorrectly included as the part of the optic disc regions in the automated disc detection scheme. For potential risk assessment and use in improving optic disc segmentation, a computerized detection of PPA was investigated. By using texture analysis, the sensitivity for detecting the moderate to severe PPA regions in the test dataset was 73% with the specificity of 95%. The proposed method may be useful for identifying the cases with the PPA in retinal fundus images.
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2007
It is important for diagnosis of glaucoma to grasp 3-D structure of an optic nerve head (ONH). Th... more It is important for diagnosis of glaucoma to grasp 3-D structure of an optic nerve head (ONH). The quantitative 3-D reconstruction of the ONH is required for the diagnosis. We propose a technique to obtain the depth value from stereo image pair of a retinal fundus for the 3-D reconstruction of the ONH. Our technique mainly consists of four steps: (1) cutout of the ONH region from the fundus images, (2) registration of the stereo pair, (3) disparity detection, and (4) depth calculation. For quantitative estimation of the depth value measured by using this method, the depth value was compared with the measurement results determined from the Heidelberg Retina Tomograph (HRT), which is a confocal laser-scanning microscope. As a result, the depth value of the ONH obtained using the stereo retinal image pair was in accordance with that obtained using the HRT (r=0.91). These results indicate that the stereo fundus images could be useful for assessing the depth value of the ONH for the diag...
2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2013
Diabetic retinopathy (DR) is the most frequent cause of blindness. Microaneurysm (MA) is an early... more Diabetic retinopathy (DR) is the most frequent cause of blindness. Microaneurysm (MA) is an early symptom of DR. Therefore, the detection of MA is important for the early detection of DR. We have proposed an automated MA detection method based on double-ring filter, but it has given many false positives. In this paper, we propose an MA detection method based on eigenvalue analysis using a Hessian matrix, with an aim to improve MA detection. After image preprocessing, the MA candidate regions were detected by eigenvalue analysis using the Hessian matrix in green-channeled retinal fundus images. Then, 126 features were calculated for each candidate region. By a threshold operation based on feature analysis, false positive candidates were removed. The candidate regions were then classified either as MA or false positive using artificial neural networks (ANN) based on principal component analysis (PCA). The 126 features were reduced to 25 components by PCA, and were then inputted to ANN. When the method was evaluated on visible MAs using 25 retinal images from the retinopathy online challenge (ROC) database, the true positive rate was 73%, with eight false positives per image.
Since the evaluation is influenced by the case selection, it is necessary to apply some common da... more Since the evaluation is influenced by the case selection, it is necessary to apply some common databases in the performance evaluation. To evaluate our computer-aided diag- nosis (CAD) system for detecting masses, 320 images from the Mammogaphy Image Analy- sis Society (MIAS) database in UK were applied in this study. Our algorithm for detecting masses was based on a standard adaptive thresholding technique which had been developed in our group. However, the preliminary result was not as well as those for a Japanese database. After we adjusted some thresholding values of our system, a 90% sensitivity with 0.8 false positive (FP) was achieved which indicated that our scheme was effective for the different databases in both Japan and UK. Moreover, the differences between the MIAS and a Japa- nese database were also discussed.
Automated blood vessels detection on retinal images is an important process in the development of... more Automated blood vessels detection on retinal images is an important process in the development of pathologies analysis systems. This paper describes about an automated blood vessel extraction using high-order local autocorrelation (HLAC) on retinal images. Although HLAC features are shift-invariant, HLAC features are weak to turned image. Therefore, a method was improved by the addition of HLAC features to a polar transformed image. We have proposed a method using HLAC, pixel-based-features and three filters. However, we have not investigated about feature selection and machine learning method. Therefore, this paper discusses about effective features and machine learning method. We tested eight methods by extension of HLAC features, addition of 4 kinds of pixel-based features, difference of preprocessing techniques, and 3 kinds of machine learning methods. Machine learning methods are general artificial neural network (ANN), a network using two ANNs, and Boosting algorithm. As a res...
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, Jul 1, 2017
The pattern of blood vessels in the eye is unique to each person because it rarely changes over t... more The pattern of blood vessels in the eye is unique to each person because it rarely changes over time. Therefore, it is well known that retinal blood vessels are useful for biometrics. This paper describes a biometrics method using the Jaccard similarity coefficient (JSC) based on blood vessel regions in retinal image pairs. The retinal image pairs were rough matched by the center of their optic discs. Moreover, the image pairs were aligned using the Iterative Closest Point algorithm based on detailed blood vessel skeletons. For registration, perspective transform was applied to the retinal images. Finally, the pairs were classified as either correct or incorrect using the JSC of the blood vessel region in the image pairs. The proposed method was applied to temporal retinal images, which were obtained in 2009 (695 images) and 2013 (87 images). The 87 images acquired in 2013 were all from persons already examined in 2009. The accuracy of the proposed method reached 100%.
SPIE Proceedings, 2017
Early detection of glaucoma is important to slow down progression of the disease and to prevent t... more Early detection of glaucoma is important to slow down progression of the disease and to prevent total vision loss. We have been studying an automated scheme for detection of a retinal nerve fiber layer defect (NFLD), which is one of the earliest signs of glaucoma on retinal fundus images. In our previous study, we proposed a multi-step detection scheme which consists of Gabor filtering, clustering and adaptive thresholding. The problems of the previous method were that the number of false positives (FPs) was still large and that the method included too many rules. In attempt to solve these problems, we investigated the end-to-end learning system without pre-specified features. A deep convolutional neural network (DCNN) with deconvolutional layers was trained to detect NFLD regions. In this preliminary investigation, we investigated effective ways of preparing the input images and compared the detection results. The optimal result was then compared with the result obtained by the previous method. DCNN training was carried out using original images of abnormal cases, original images of both normal and abnormal cases, ellipse-based polar transformed images, and transformed half images. The result showed that use of both normal and abnormal cases increased the sensitivity as well as the number of FPs. Although NFLDs are visualized with the highest contrast in green plane, the use of color images provided higher sensitivity than the use of green image only. The free response receiver operating characteristic curve using the transformed color images, which was the best among seven different sets studied, was comparable to that of the previous method. Use of DCNN has a potential to improve the generalizability of automated detection method of NFLDs and may be useful in assisting glaucoma diagnosis on retinal fundus images.
Medical Imaging 2016: Computer-Aided Diagnosis, 2016
Early detection of glaucoma is important to slow down or cease progression of the disease and for... more Early detection of glaucoma is important to slow down or cease progression of the disease and for preventing total blindness. We have previously proposed an automated scheme for detection of retinal nerve fiber layer defect (NFLD), which is one of the early signs of glaucoma observed on retinal fundus images. In this study, a new multi-step detection scheme was included to improve detection of subtle and narrow NFLDs. In addition, new features were added to distinguish between NFLDs and blood vessels, which are frequent sites of false positives (FPs). The result was evaluated with a new test dataset consisted of 261 cases, including 130 cases with NFLDs. Using the proposed method, the initial detection rate was improved from 82% to 98%. At the sensitivity of 80%, the number of FPs per image was reduced from 4.25 to 1.36. The result indicates the potential usefulness of the proposed method for early detection of glaucoma.
SPIE Proceedings, 2008
Biometric technique has been implemented instead of conventional identification methods such as p... more Biometric technique has been implemented instead of conventional identification methods such as password in computer, automatic teller machine (ATM), and entrance and exit management system. We propose a personal identification (PI) system using color retinal fundus images which are unique to each individual. The proposed procedure for identification is based on comparison of an input fundus image with reference fundus images in the database. In the first step, registration between the input image and the reference image is performed. The step includes translational and rotational movement. The PI is based on the measure of similarity between blood vessel images generated from the input and reference images. The similarity measure is defined as the cross-correlation coefficient calculated from the pixel values. When the similarity is greater than a predetermined threshold, the input image is identified. This means both the input and the reference images are associated to the same person. Four hundred sixty-two fundus images including forty-one same-person's image pairs were used for the estimation of the proposed technique. The false rejection rate and the false acceptance rate were 9.9×10-5 % and 4.3×10-5 %, respectively. The results indicate that the proposed method has a higher performance than other biometrics except for DNA. To be used for practical application in the public, the device which can take retinal fundus images easily is needed. The proposed method is applied to not only the PI but also the system which warns about misfiling of fundus images in medical facilities.
International Congress Series, 2001
We have been developing an automated detection algorithm for masses on mammograms. There were the... more We have been developing an automated detection algorithm for masses on mammograms. There were the cases of missed detections, in most of which the masses existed around thick-mammarygland region or near chest-wall region. The purpose of this study is to develop an algorithm for detecting such masses based on a sector-shape mass-pattern model. After applying the new technique along with our previous method to 3575 digitized mammograms, the detection sensitivity was changed from 78% to 93% at the number of false positives with varying from 1.2 to 1.9 per image. This result indicated that the new technique improved the overall performance of our CAD system for detecting the masses on mammograms effectively.
IEEE Transactions on Medical Imaging, 2010
The detection of microaneurysms in digital color fundus photographs is a critical first step in a... more The detection of microaneurysms in digital color fundus photographs is a critical first step in automated screening for diabetic retinopathy (DR), a common complication of diabetes. To accomplish this detection numerous methods have been published in the past but none of these was compared with each other on the same data. In this work we present the results of the first international microaneurysm detection competition, organized in
This paper describes the purpose and summary of our recent progress and achievement in the resear... more This paper describes the purpose and summary of our recent progress and achievement in the research work, which is a part of the research project “Multidisciplinary Computational Anatomy and Its Application to Highly Intelligent Diagnosis and Therapy,” founded by Grant-in-Aid for Scientific Research on Innovative Areas, MEXT, Japan. The main purposes of our research in this project are to establish a scientific principle of multidisciplinary computational anatomy and to develop computer-aided diagnosis (CAD) systems based on such anatomical models for organ and tissue functions. In FY2015, we proceeded with the preparation for sub projects, such as the database establishment, and started the investigation in earnest. Several progresses have been achieved with the promising results, which encourage the continual effort for success of this project.
IEEE Transactions on Medical Imaging, 2001
Abstract—Recently, we have been developing several automated algorithms for detecting masses on m... more Abstract—Recently, we have been developing several automated algorithms for detecting masses on mammograms.,For our algo- rithm, we devised an adaptive thresholding technique for detecting masses, but our system failed to detect masses with a partial loss of region that were located on the edge of the film. This is a common issue in all of the algorithms developed so far
Glaucoma is the second leading cause of vision loss in the world. Detection of retinal nerve fibe... more Glaucoma is the second leading cause of vision loss in the world. Detection of retinal nerve fiber layer defects (NFLDs), which is one of the early glaucomatous changes, on retinal fundus images obtained in mass screening may prevent patients from becoming permanently blind. In this study, a technique for contrast enhancement and automated detection of NFLDs was investigated. Images used in this study were obtained from the Tajimi general screening database, which includes retinal fundus images obtained in the population glaucoma screening study at Tajimi, Japan. In this preliminary investigation, images with at least one identified NFLD were included. Lesions of NFLDs were individually marked by two ophthalmologists, and those identified by both ophthalmologists were considered as the target NFLDs. First, the blood vessel regions in retinal fundus images were identified and interpolated by the surrounding pixels for creating “blood-vessel-erased” images. The resulted color images w...
Proceedings of SPIE - The International Society for Optical Engineering
Glaucoma is one of the leading causes of blindness in Japan and the US. One of the indices for di... more Glaucoma is one of the leading causes of blindness in Japan and the US. One of the indices for diagnosis of glaucoma is the cup-to-disc ratio (CDR). We have been developing a computerized method for measuring CDR on stereo fundus photographs. Although our previous study indicated that the method may be useful, cup determination was not always successful, especially for the normal eyes. In this study, we investigated a new method to quantify the likelihood of glaucomatous disc based on the similarity scores to the glaucoma and non-glaucoma models. Eighty-seven images, including 40 glaucomatous eyes, were used in this study. Only one eye from each patient was used. Using a stereo fundus camera, two images were captured from different angles, and the depth image was created by finding the local corresponding points. One of the characteristics of a glaucomatous disc can be not only that the cup is enlarged but it has an acute slope. On the other hand, a non-glaucomatous cup generally ha...
Microaneurysm is one of diabetic retinopathy (DR) findings. Visual loss can be prevented by early... more Microaneurysm is one of diabetic retinopathy (DR) findings. Visual loss can be prevented by early detection and treatment of DR. The fundus examination is effective for early detection of DR. We also proposed an automated MAs detection method using double-ring filter and shape index based on Hessian matrix in the retinal fundus images. However, it had a problem in that many MAs were lost, although the performance of the detectors had peaked. In this study, we applied a density gradient vector concentration with a point concentration. The microaneurysm candidates included so many false positives, that the candidates were classified into microaneurysms and FPs by using support vector machine with 48 features. For evaluation of this study, images from the ROC (Retinopathy Online Challenge) database were used. This database comprises 50 training images and 50 test images. When this method was evaluated by using the 50 test images, the sensitivity was 67% of 8 false positives per image. ...
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2005
We have developed a computer-aided diagnosis system (CAD) to detect abnormalities in fundus image... more We have developed a computer-aided diagnosis system (CAD) to detect abnormalities in fundus images. In Japan, ophthalmologists usually detect hypertensive changes by identifying arteriolar narrowing and focal arteriolar narrowing. The purpose of this study is to develop an automated method for detecting arteriolar narrowing and focal arteriolar narrowing on fundus images. The blood vessel candidates were detected by the density analysis method. In blood vessel tracking, a local detection function was used to determine the centerline of the blood vessel. A direction comparison function using three vectors was designed to optimally estimate the next possible location of a blood vessel. After the connectivity of vessel segments was adjusted based on the recognized intersections, the true tree-like structure of the blood vessels was established. The blood vessels were recognized as arteries or veins by hue of HSV color space and their diameters. The arteriolar narrowing was detected by ...
Nihon Hoshasen Gijutsu Gakkai zasshi, 2002
We have been developing automated detection algorithms for masses and clustered microcalcificatio... more We have been developing automated detection algorithms for masses and clustered microcalcifications in a mammography computer-aided diagnosis (CAD) system. In this study, we investigated the potential of our CAD system by comparing 579 physicians' interpretation results with that of the CAD system's cancer detection for 100 mammograms (21 malignant and 29 benign cases) employed in a physicians' self-learning course. As a result, our CAD system detected 7 out of 8 malignant lesions whose physicians' averaged sensitivity was less than 60%. Although the average of physicians' sensitivities were 76% (about 16 cases), the CAD system's detection rate was 90% (19 cases). Sensitivity was raised up to 97% if the physicians' interpretation and the CAD system's detection result were treated in a matter of logical OR. Thus, it was raised the possibility that even the less-experienced physicians would diagnose with a higher sensitivity by using the computer output...
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2012
Glaucoma is the first leading cause of vision loss in Japan, thus developing a scheme for helping... more Glaucoma is the first leading cause of vision loss in Japan, thus developing a scheme for helping glaucoma diagnosis is important. For this problem, automated nerve fiber layer defects (NFLDs) detection method was proposed, but glaucoma risk assessment using this method was not evaluated. In this paper, computerized risk assessment for having glaucoma was attempted by use of the patients' clinical information, and the performances of the NFLDs detection and the glaucoma risk assessment were compared. The clinical data includes the systemic data, ophthalmologic data, and right and left retinal images. Glaucoma risk assessment was built by using machine learning technique, which were artificial neural network, radial basis function (RBF) network, k-nearest neighbor algorithm, and support vector machine. The inputting parameter was ten clinical ones with/without the results of NFLDs detection. As a result, proposed glaucoma risk assessment showed the higher performance than the NFL...
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2011
Presence of peripapillary chorioretinal atrophy (PPA) is considered one of the risk factors for g... more Presence of peripapillary chorioretinal atrophy (PPA) is considered one of the risk factors for glaucoma. It can be identified as bright regions in retinal fundus images, and therefore, incorrectly included as the part of the optic disc regions in the automated disc detection scheme. For potential risk assessment and use in improving optic disc segmentation, a computerized detection of PPA was investigated. By using texture analysis, the sensitivity for detecting the moderate to severe PPA regions in the test dataset was 73% with the specificity of 95%. The proposed method may be useful for identifying the cases with the PPA in retinal fundus images.
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2007
It is important for diagnosis of glaucoma to grasp 3-D structure of an optic nerve head (ONH). Th... more It is important for diagnosis of glaucoma to grasp 3-D structure of an optic nerve head (ONH). The quantitative 3-D reconstruction of the ONH is required for the diagnosis. We propose a technique to obtain the depth value from stereo image pair of a retinal fundus for the 3-D reconstruction of the ONH. Our technique mainly consists of four steps: (1) cutout of the ONH region from the fundus images, (2) registration of the stereo pair, (3) disparity detection, and (4) depth calculation. For quantitative estimation of the depth value measured by using this method, the depth value was compared with the measurement results determined from the Heidelberg Retina Tomograph (HRT), which is a confocal laser-scanning microscope. As a result, the depth value of the ONH obtained using the stereo retinal image pair was in accordance with that obtained using the HRT (r=0.91). These results indicate that the stereo fundus images could be useful for assessing the depth value of the ONH for the diag...
2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2013
Diabetic retinopathy (DR) is the most frequent cause of blindness. Microaneurysm (MA) is an early... more Diabetic retinopathy (DR) is the most frequent cause of blindness. Microaneurysm (MA) is an early symptom of DR. Therefore, the detection of MA is important for the early detection of DR. We have proposed an automated MA detection method based on double-ring filter, but it has given many false positives. In this paper, we propose an MA detection method based on eigenvalue analysis using a Hessian matrix, with an aim to improve MA detection. After image preprocessing, the MA candidate regions were detected by eigenvalue analysis using the Hessian matrix in green-channeled retinal fundus images. Then, 126 features were calculated for each candidate region. By a threshold operation based on feature analysis, false positive candidates were removed. The candidate regions were then classified either as MA or false positive using artificial neural networks (ANN) based on principal component analysis (PCA). The 126 features were reduced to 25 components by PCA, and were then inputted to ANN. When the method was evaluated on visible MAs using 25 retinal images from the retinopathy online challenge (ROC) database, the true positive rate was 73%, with eight false positives per image.