Browsing by Author "Shin, Dongsuk"
Now showing 1 - 6 of 6
Results Per Page
Sort Options
Item Automated frame selection process for high-resolution microendoscopy(SPIE, 2015) Ishijima, Ayumu; Schwarz, Richard A.; Shin, Dongsuk; Mondrik, Sharon; Vigneswaran, Nadarajah; Gillenwater, Ann M.; Anandasabapathy, Sharmila; Richards-Kortum, RebeccaWe developed an automated frame selection algorithm for high-resolution microendoscopy video sequences. The algorithm rapidly selects a representative frame with minimal motion artifact from a short video sequence, enabling fully automated image analysis at the point-of-care. The algorithm was evaluated by quantitative comparison of diagnostically relevant image features and diagnostic classification results obtained using automated frame selection versus manual frame selection. A data set consisting of video sequences collected in vivo from 100 oral sites and 167 esophageal sites was used in the analysis. The area under the receiver operating characteristic curve was 0.78 (automated selection) versus 0.82 (manual selection) for oral sites, and 0.93 (automated selection) versus 0.92 (manual selection) for esophageal sites. The implementation of fully automated high-resolution microendoscopy at the point-of-care has the potential to reduce the number of biopsies needed for accurate diagnosis of precancer and cancer in low-resource settings where there may be limited infrastructure and personnel for standard histologic analysis.Item Low-Cost High-Resolution Microendoscopy for the Detection of Esophageal Squamous Cell Neoplasia: An International Trial(Elsevier, 2015) Protano, Marion-Anna; Xu, Hong; Wang, Guiqi; Polydorides, Alexandros D.; Dawsey, Sanford M.; Cui, Junsheng; Xue, Liyan; Zhang, Fan; Quang, Timothy; Pierce, Mark C.; Shin, Dongsuk; Schwarz, Richard A.; Bhutani, Manoop S.; Lee, Michelle; Parikh, Neil; Hur, Chin; Xu, Weiran; Moshier, Erin; Godbold, James; Mitcham, Josephine; Hudson, Courtney; Richards-Kortum, Rebecca R.; Anandasabapathy, SharmilaBackground & Aims: Esophageal squamous cell neoplasia has a high mortality rate as a result of late detection. In high-risk regions such as China, screening is performed by Lugol’s chromoendoscopy (LCE). LCE has low specificity, resulting in unnecessary tissue biopsy with a subsequent increase in procedure cost and risk. The purpose of this study was to evaluate the accuracy of a novel, low-cost, high-resolution microendoscope (HRME) as an adjunct to LCE. Methods: In this prospective trial, 147 consecutive high-risk patients were enrolled from 2 US and 2 Chinese tertiary centers. Three expert and 4 novice endoscopists performed white-light endoscopy followed by LCE and HRME. All optical images were compared with the gold standard of histopathology. Results: By using a per-biopsy analysis, the sensitivity of LCE vs LCE + HRME was 96% vs 91% (P = .0832), specificity was 48% vs 88% (P < .001), positive predictive value was 22% vs 45% (P < .0001), negative predictive value was 98% vs 98% (P = .3551), and overall accuracy was 57% vs 90% (P < .001), respectively. By using a per-patient analysis, the sensitivity of LCE vs LCE + HRME was 100% vs 95% (P = .16), specificity was 29% vs 79% (P < .001), positive predictive value was 32% vs 60%, 100% vs 98%, and accuracy was 47% vs 83% (P < .001). With the use of HRME, 136 biopsies (60%; 95% confidence interval, 53%–66%) could have been spared, and 55 patients (48%; 95% confidence interval, 38%–57%) could have been spared any biopsy. Conclusions: In this trial, HRME improved the accuracy of LCE for esophageal squamous cell neoplasia screening and surveillance. HRME may be a cost-effective optical biopsy adjunct to LCE, potentially reducing unnecessary biopsies and facilitating real-time decision making in globally underserved regions. ClinicalTrials.gov, NCT 01384708.Item Micro-anatomical quantitative optical imaging: toward automated assessment of breast tissues(BioMed Central, 2015) Dobbs, Jessica L.; Mueller, Jenna L.; Krishnamurthy, Savitri; Shin, Dongsuk; Kuerer, Henry; Yang, Wei; Ramanujam, Nirmala; Richards-Kortum, RebeccaAbstract Introduction Pathologists currently diagnose breast lesions through histologic assessment, which requires fixation and tissue preparation. The diagnostic criteria used to classify breast lesions are qualitative and subjective, and inter-observer discordance has been shown to be a significant challenge in the diagnosis of selected breast lesions, particularly for borderline proliferative lesions. Thus, there is an opportunity to develop tools to rapidly visualize and quantitatively interpret breast tissue morphology for a variety of clinical applications. Methods Toward this end, we acquired images of freshly excised breast tissue specimens from a total of 34 patients using confocal fluorescence microscopy and proflavine as a topical stain. We developed computerized algorithms to segment and quantify nuclear and ductal parameters that characterize breast architectural features. A total of 33 parameters were evaluated and used as input to develop a decision tree model to classify benign and malignant breast tissue. Benign features were classified in tissue specimens acquired from 30 patients and malignant features were classified in specimens from 22 patients. Results The decision tree model that achieved the highest accuracy for distinguishing between benign and malignant breast features used the following parameters: standard deviation of inter-nuclear distance and number of duct lumens. The model achieved 81 % sensitivity and 93 % specificity, corresponding to an area under the curve of 0.93 and an overall accuracy of 90 %. The model classified IDC and DCIS with 92 % and 96 % accuracy, respectively. The cross-validated model achieved 75 % sensitivity and 93 % specificity and an overall accuracy of 88 %. Conclusions These results suggest that proflavine staining and confocal fluorescence microscopy combined with image analysis strategies to segment morphological features could potentially be used to quantitatively diagnose freshly obtained breast tissue at the point of care without the need for tissue preparation.Item Micro-anatomical quantitative optical imaging: toward automated assessment of breast tissues(BioMed Central, 2015) Dobbs, Jessica L.; Mueller, Jenna L.; Krishnamurthy, Savitri; Shin, Dongsuk; Kuerer, Henry; Yang, Wei; Ramanujam, Nirmala; Richards-Kortum, RebeccaIntroduction: Pathologists currently diagnose breast lesions through histologic assessment, which requires fixation and tissue preparation. The diagnostic criteria used to classify breast lesions are qualitative and subjective, and inter-observer discordance has been shown to be a significant challenge in the diagnosis of selected breast lesions, particularly for borderline proliferative lesions. Thus, there is an opportunity to develop tools to rapidly visualize and quantitatively interpret breast tissue morphology for a variety of clinical applications. Methods: Toward this end, we acquired images of freshly excised breast tissue specimens from a total of 34 patients using confocal fluorescence microscopy and proflavine as a topical stain. We developed computerized algorithms to segment and quantify nuclear and ductal parameters that characterize breast architectural features. A total of 33 parameters were evaluated and used as input to develop a decision tree model to classify benign and malignant breast tissue. Benign features were classified in tissue specimens acquired from 30 patients and malignant features were classified in specimens from 22 patients. Results: The decision tree model that achieved the highest accuracy for distinguishing between benign and malignant breast features used the following parameters: standard deviation of inter-nuclear distance and number of duct lumens. The model achieved 81ᅠ% sensitivity and 93ᅠ% specificity, corresponding to an area under the curve of 0.93 and an overall accuracy of 90ᅠ%. The model classified IDC and DCIS with 92ᅠ% and 96ᅠ% accuracy, respectively. The cross-validated model achieved 75ᅠ% sensitivity and 93ᅠ% specificity and an overall accuracy of 88ᅠ%. Conclusions: These results suggest that proflavine staining and confocal fluorescence microscopy combined with image analysis strategies to segment morphological features could potentially be used to quantitatively diagnose freshly obtained breast tissue at the point of care without the need for tissue preparation.Item Quantitative Analysis of High-Resolution Microendoscopic Images for Diagnosis of Esophageal Squamous Cell Carcinomaᅠ(Elsevier, 2015) Shin, Dongsuk; Protano, Marion-Anna; Polydorides, Alexandros D.; Dawsey, Sanford M.; Pierce, Mark C.; Kim, Michelle Kang; Schwarz, Richard A.; Quang, Timothy; Parikh, Neil; Bhutani, Manoop S.; Zhang, Fan; Wang, Guiqi; Xue, Liyan; Wang, Xueshan; Xu, Hong; Anandasabapathy, Sharmila; Richards-Kortum, Rebecca R.Background & Aims: High-resolution microendoscopy is an optical imaging technique with the potential to improve the accuracy of endoscopic screening for esophageal squamous neoplasia. Although these microscopic images can be interpreted readily by trained personnel, quantitative image analysis software could facilitate the use of this technology in low-resource settings. In this study, we developed and evaluated quantitative image analysis criteria for the evaluation of neoplastic and non-neoplastic squamous esophageal mucosa. Methods: We performed an image analysis of 177 patients undergoing standard upper endoscopy for screening or surveillance of esophageal squamous neoplasia, using high-resolution microendoscopy, at 2 hospitals in China and at 1 hospital in the United States from May 2010 to October 2012. Biopsy specimens were collected from imaged sites (n = 375), and a consensus diagnosis was provided by 2 expert gastrointestinal pathologists and used as the standard. Results: Quantitative information from the high-resolution images was used to develop an algorithm to identify high-grade squamous dysplasia or invasive squamous cell cancer, based on histopathology findings. Optimal performance was obtained using the mean nuclear area as the basis for classification, resulting in sensitivities and specificities of 93% and 92% in the training set, 87% and 97% in the test set, and 84% and 95% in an independent validation set, respectively. Conclusions: High-resolution microendoscopy with quantitative image analysis can aid in the identification of esophageal squamous neoplasia. Use of software-based image guides may overcome issues of training and expertise in low-resource settings, allowing for widespread use of these optical biopsy technologies.Item Quantitative analysis of high-resolution microendoscopic images for diagnosis of neoplasia in patients with Barrett’s esophagus(Elsevier, 2016) Shin, Dongsuk; Lee, Michelle H.; Polydorides, Alexandros D.; Pierce, Mark C.; Vila, Peter M.; Parikh, Neil D.; Rosen, Daniel G.; Anandasabapathy, Sharmila; Richards-Kortum, Rebecca R.Background and Aims: Previous studies show that microendoscopic images can be interpreted visually to identify the presence of neoplasia in patients with Barrett’s esophagus (BE), but this approach is subjective and requires clinical expertise. This study describes an approach for quantitative image analysis of microendoscopic images to identify neoplastic lesions in patients with BE. Methods: Images were acquired from 230 sites from 58 patients by using a fiberoptic high-resolution microendoscope during standard endoscopic procedures. Images were analyzed by a fully automated image processing algorithm, which automatically selected a region of interest and calculated quantitative image features. Image features were used to develop an algorithm to identify the presence of neoplasia; results were compared with a histopathology diagnosis. Results: A sequential classification algorithm that used image features related to glandular and cellular morphology resulted in a sensitivity of 84% and a specificity of 85%. Applying the algorithm to an independent validation set resulted in a sensitivity of 88% and a specificity of 85%. Conclusions: This pilot study demonstrates that automated analysis of microendoscopic images can provide an objective, quantitative framework to assist clinicians in evaluating esophageal lesions from patients with BE. (Clinical trial registration number: NCT01384227 and NCT02018367.)