The Experts below are selected from a list of 1629 Experts worldwide ranked by ideXlab platform
James M. Rehg - One of the best experts on this subject based on the ideXlab platform.
-
automated Foveola localization in retinal 3d oct images using structural support vector machine prediction
Medical Image Computing and Computer-Assisted Intervention, 2012Co-Authors: Yu-ying Liu, Hiroshi Ishikawa, Mei Chen, Gadi Wollstein, Joel S. Schuman, James M. RehgAbstract:We develop an automated method to determine the Foveola location in macular 3D-OCT images in either healthy or pathological conditions. Structural Support Vector Machine (S-SVM) is trained to directly predict the location of the Foveola, such that the score at the ground truth position is higher than that at any other position by a margin scaling with the associated localization loss. This S-SVM formulation directly minimizes the empirical risk of localization error, and makes efficient use of all available training data. It deals with the localization problem in a more principled way compared to the conventional binary classifier learning that uses zero-one loss and random sampling of negative examples. A total of 170 scans were collected for the experiment. Our method localized 95.1% of testing scans within the anatomical area of the Foveola. Our experimental results show that the proposed method can effectively identify the location of the Foveola, facilitating diagnosis around this important landmark.
-
MICCAI (1) - Automated Foveola localization in retinal 3D-OCT images using structural support vector machine prediction
Medical Image Computing and Computer-Assisted Intervention – MICCAI 2012, 2012Co-Authors: Yu-ying Liu, Hiroshi Ishikawa, Mei Chen, Gadi Wollstein, Joel S. Schuman, James M. RehgAbstract:We develop an automated method to determine the Foveola location in macular 3D-OCT images in either healthy or pathological conditions. Structural Support Vector Machine (S-SVM) is trained to directly predict the location of the Foveola, such that the score at the ground truth position is higher than that at any other position by a margin scaling with the associated localization loss. This S-SVM formulation directly minimizes the empirical risk of localization error, and makes efficient use of all available training data. It deals with the localization problem in a more principled way compared to the conventional binary classifier learning that uses zero-one loss and random sampling of negative examples. A total of 170 scans were collected for the experiment. Our method localized 95.1% of testing scans within the anatomical area of the Foveola. Our experimental results show that the proposed method can effectively identify the location of the Foveola, facilitating diagnosis around this important landmark.
Muh Shy Chen - One of the best experts on this subject based on the ideXlab platform.
-
Reconstructing Foveola by Foveolar Internal Limiting Membrane Non-Peeling and Tissue Repositioning for Lamellar Hole-Related Epiretinal Proliferation.
Scientific Reports, 2019Co-Authors: Muh Shy ChenAbstract:Differences in the pathogenesis and clinical characteristics between lamellar macular hole (LMH) with and without LMH-associated epiretinal proliferation (LHEP) can have surgical implications. This study investigated the effects of treating LHEP by Foveolar internal limiting membrane (ILM) non-peeling and epiretinal proliferative (EP) tissue repositioning on visual acuity and Foveolar architecture. Consecutive patients with LHEP treated at our institution were enrolled. The eyes were divided into a conventional total ILM peeling group (group 1, n = 11) and a Foveolar ILM non-peeling group (group 2, n = 22). In group 2, a doughnut-shaped ILM was peeled, leaving a 400-μm-diameter ILM without elevated margin over the Foveola after EP tissue repositioning. The EP tissue was elevated, trimmed, and inverted into the LMH. Postoperatively, the LMH was sealed in all eyes in group 2, with significantly better best-corrected visual acuity (−0.26 vs −0.10 logMAR; p = 0.002). A smaller retinal defect (p = 0.003), a more restored ellipsoid zone (p = 0.002), and a more smooth foveal depression (p
-
reconstructing Foveola by Foveolar internal limiting membrane non peeling and tissue repositioning for lamellar hole related epiretinal proliferation
Scientific Reports, 2019Co-Authors: Muh Shy ChenAbstract:Differences in the pathogenesis and clinical characteristics between lamellar macular hole (LMH) with and without LMH-associated epiretinal proliferation (LHEP) can have surgical implications. This study investigated the effects of treating LHEP by Foveolar internal limiting membrane (ILM) non-peeling and epiretinal proliferative (EP) tissue repositioning on visual acuity and Foveolar architecture. Consecutive patients with LHEP treated at our institution were enrolled. The eyes were divided into a conventional total ILM peeling group (group 1, n = 11) and a Foveolar ILM non-peeling group (group 2, n = 22). In group 2, a doughnut-shaped ILM was peeled, leaving a 400-μm-diameter ILM without elevated margin over the Foveola after EP tissue repositioning. The EP tissue was elevated, trimmed, and inverted into the LMH. Postoperatively, the LMH was sealed in all eyes in group 2, with significantly better best-corrected visual acuity (−0.26 vs −0.10 logMAR; p = 0.002). A smaller retinal defect (p = 0.003), a more restored ellipsoid zone (p = 0.002), and a more smooth foveal depression (p < 0.001) were achieved in group 2. Foveolar ILM non-peeling and EP tissue repositioning sealed the LMH, released the tangential traction, and achieved better visual acuity. The presumed Foveolar architecture may be reconstructed surgically. LMH with LHEP could have a combined degenerative and tractional mechanism.
-
Long-term outcome of Foveolar internal limiting membrane nonpeeling for myopic traction maculopathy.
Retina, 2014Co-Authors: Chung-may Yang, Jen Shang Huang, Chang-hao Yang, Po-ting Yeh, Ta-ching Chen, Muh Shy ChenAbstract:PURPOSE To investigate the long-term results of a novel technique to preserve the Foveolar cone without peeling off the Foveolar internal limiting membrane (ILM) during myopic traction maculopathy surgery. METHODS Nineteen patients (19 eyes) were retrospectively studied and divided into 2 groups by the extent of ILM peeled and followed for more than 3 years. Group 1: Foveolar ILM nonpeeling group (FN) (12 eyes) and Group 2: total peeling of foveal ILM group (TP) (7 eyes). A donut-shaped ILM was peeled off, leaving a 400-μm diameter ILM over Foveola with a sharp margin in FN group. RESULTS Macular hole was developed in 2 of the 7 eyes (28.6%) in the TP group and none in the FN group. Long-term central fovea thickness thinning and decrease of vision were found in the TP group, but not in the FN group (P < 0.05). Inner segment/outer segment line recovered in 75% of the 12 eyes in the FN group, but in only 14.3% of the 7 eyes in the TP group. CONCLUSION Preservation of the Foveolar cone by Foveola nonpeeling surgery correlates with better anatomical and visual results than total peel, prevents long-term Foveolar retinal thinning, and successfully saves the fovea from macular hole formation.
Yu-ying Liu - One of the best experts on this subject based on the ideXlab platform.
-
automated Foveola localization in retinal 3d oct images using structural support vector machine prediction
Medical Image Computing and Computer-Assisted Intervention, 2012Co-Authors: Yu-ying Liu, Hiroshi Ishikawa, Mei Chen, Gadi Wollstein, Joel S. Schuman, James M. RehgAbstract:We develop an automated method to determine the Foveola location in macular 3D-OCT images in either healthy or pathological conditions. Structural Support Vector Machine (S-SVM) is trained to directly predict the location of the Foveola, such that the score at the ground truth position is higher than that at any other position by a margin scaling with the associated localization loss. This S-SVM formulation directly minimizes the empirical risk of localization error, and makes efficient use of all available training data. It deals with the localization problem in a more principled way compared to the conventional binary classifier learning that uses zero-one loss and random sampling of negative examples. A total of 170 scans were collected for the experiment. Our method localized 95.1% of testing scans within the anatomical area of the Foveola. Our experimental results show that the proposed method can effectively identify the location of the Foveola, facilitating diagnosis around this important landmark.
-
MICCAI (1) - Automated Foveola localization in retinal 3D-OCT images using structural support vector machine prediction
Medical Image Computing and Computer-Assisted Intervention – MICCAI 2012, 2012Co-Authors: Yu-ying Liu, Hiroshi Ishikawa, Mei Chen, Gadi Wollstein, Joel S. Schuman, James M. RehgAbstract:We develop an automated method to determine the Foveola location in macular 3D-OCT images in either healthy or pathological conditions. Structural Support Vector Machine (S-SVM) is trained to directly predict the location of the Foveola, such that the score at the ground truth position is higher than that at any other position by a margin scaling with the associated localization loss. This S-SVM formulation directly minimizes the empirical risk of localization error, and makes efficient use of all available training data. It deals with the localization problem in a more principled way compared to the conventional binary classifier learning that uses zero-one loss and random sampling of negative examples. A total of 170 scans were collected for the experiment. Our method localized 95.1% of testing scans within the anatomical area of the Foveola. Our experimental results show that the proposed method can effectively identify the location of the Foveola, facilitating diagnosis around this important landmark.
Joel S. Schuman - One of the best experts on this subject based on the ideXlab platform.
-
automated Foveola localization in retinal 3d oct images using structural support vector machine prediction
Medical Image Computing and Computer-Assisted Intervention, 2012Co-Authors: Yu-ying Liu, Hiroshi Ishikawa, Mei Chen, Gadi Wollstein, Joel S. Schuman, James M. RehgAbstract:We develop an automated method to determine the Foveola location in macular 3D-OCT images in either healthy or pathological conditions. Structural Support Vector Machine (S-SVM) is trained to directly predict the location of the Foveola, such that the score at the ground truth position is higher than that at any other position by a margin scaling with the associated localization loss. This S-SVM formulation directly minimizes the empirical risk of localization error, and makes efficient use of all available training data. It deals with the localization problem in a more principled way compared to the conventional binary classifier learning that uses zero-one loss and random sampling of negative examples. A total of 170 scans were collected for the experiment. Our method localized 95.1% of testing scans within the anatomical area of the Foveola. Our experimental results show that the proposed method can effectively identify the location of the Foveola, facilitating diagnosis around this important landmark.
-
MICCAI (1) - Automated Foveola localization in retinal 3D-OCT images using structural support vector machine prediction
Medical Image Computing and Computer-Assisted Intervention – MICCAI 2012, 2012Co-Authors: Yu-ying Liu, Hiroshi Ishikawa, Mei Chen, Gadi Wollstein, Joel S. Schuman, James M. RehgAbstract:We develop an automated method to determine the Foveola location in macular 3D-OCT images in either healthy or pathological conditions. Structural Support Vector Machine (S-SVM) is trained to directly predict the location of the Foveola, such that the score at the ground truth position is higher than that at any other position by a margin scaling with the associated localization loss. This S-SVM formulation directly minimizes the empirical risk of localization error, and makes efficient use of all available training data. It deals with the localization problem in a more principled way compared to the conventional binary classifier learning that uses zero-one loss and random sampling of negative examples. A total of 170 scans were collected for the experiment. Our method localized 95.1% of testing scans within the anatomical area of the Foveola. Our experimental results show that the proposed method can effectively identify the location of the Foveola, facilitating diagnosis around this important landmark.
Gadi Wollstein - One of the best experts on this subject based on the ideXlab platform.
-
automated Foveola localization in retinal 3d oct images using structural support vector machine prediction
Medical Image Computing and Computer-Assisted Intervention, 2012Co-Authors: Yu-ying Liu, Hiroshi Ishikawa, Mei Chen, Gadi Wollstein, Joel S. Schuman, James M. RehgAbstract:We develop an automated method to determine the Foveola location in macular 3D-OCT images in either healthy or pathological conditions. Structural Support Vector Machine (S-SVM) is trained to directly predict the location of the Foveola, such that the score at the ground truth position is higher than that at any other position by a margin scaling with the associated localization loss. This S-SVM formulation directly minimizes the empirical risk of localization error, and makes efficient use of all available training data. It deals with the localization problem in a more principled way compared to the conventional binary classifier learning that uses zero-one loss and random sampling of negative examples. A total of 170 scans were collected for the experiment. Our method localized 95.1% of testing scans within the anatomical area of the Foveola. Our experimental results show that the proposed method can effectively identify the location of the Foveola, facilitating diagnosis around this important landmark.
-
MICCAI (1) - Automated Foveola localization in retinal 3D-OCT images using structural support vector machine prediction
Medical Image Computing and Computer-Assisted Intervention – MICCAI 2012, 2012Co-Authors: Yu-ying Liu, Hiroshi Ishikawa, Mei Chen, Gadi Wollstein, Joel S. Schuman, James M. RehgAbstract:We develop an automated method to determine the Foveola location in macular 3D-OCT images in either healthy or pathological conditions. Structural Support Vector Machine (S-SVM) is trained to directly predict the location of the Foveola, such that the score at the ground truth position is higher than that at any other position by a margin scaling with the associated localization loss. This S-SVM formulation directly minimizes the empirical risk of localization error, and makes efficient use of all available training data. It deals with the localization problem in a more principled way compared to the conventional binary classifier learning that uses zero-one loss and random sampling of negative examples. A total of 170 scans were collected for the experiment. Our method localized 95.1% of testing scans within the anatomical area of the Foveola. Our experimental results show that the proposed method can effectively identify the location of the Foveola, facilitating diagnosis around this important landmark.
