Faisal Mahmood, Daniel Borders, Richard J. Chen, Jordan A. Sweer, S. Tilley, N. Nishioka, J. Stayman, N. Durr
Colorectal cancer is the second leading cause of cancer deaths in the United States and causes over 50,000 deaths annually. The standard of care for colorectal cancer detection and prevention is an optical colonoscopy and polypectomy. However, over 20% of the polyps are typically missed during a standard colonoscopy procedure and 60% of colorectal cancer cases are attributed to these missed polyps. Surface topography plays a vital role in identification and characterization of lesions, but topographic features often appear subtle to a conventional endoscope. Chromoendoscopy can highlight topographic features of the mucosa and has shown to improve lesion detection rate, but requires dedicated training and increases procedure time. Photometric stereo endoscopy captures this topography but is qualitative due to unknown working distances from each point of mucosa to the endoscope. In this work, we use deep learning to estimate a depth map from an endoscope camera with four alternating light sources. Since endoscopy videos with ground truth depth maps are challenging to attain, we generated synthetic data using graphical rendering from an anatomically realistic 3D colon model and a forward model of a virtual endoscope with alternating light sources. We propose an encoder-decoder style deep network, where the encoder is split into four branches of sub-encoder networks that simultaneously extract features from each of the four sources and fuse these feature maps as the network goes deeper. This is complemented by skip connections, which maintain spatial consistency when the features are decoded. We demonstrate that, when compared to monocular depth estimation, this setup can reduce the average NRMS error for depth estimation in a silicone colon phantom by 38% and in a pig colon by 31%.
{"title":"Robust photometric stereo endoscopy via deep learning trained on synthetic data (Conference Presentation)","authors":"Faisal Mahmood, Daniel Borders, Richard J. Chen, Jordan A. Sweer, S. Tilley, N. Nishioka, J. Stayman, N. Durr","doi":"10.1117/12.2509878","DOIUrl":"https://doi.org/10.1117/12.2509878","url":null,"abstract":"Colorectal cancer is the second leading cause of cancer deaths in the United States and causes over 50,000 deaths annually. The standard of care for colorectal cancer detection and prevention is an optical colonoscopy and polypectomy. However, over 20% of the polyps are typically missed during a standard colonoscopy procedure and 60% of colorectal cancer cases are attributed to these missed polyps. Surface topography plays a vital role in identification and characterization of lesions, but topographic features often appear subtle to a conventional endoscope. Chromoendoscopy can highlight topographic features of the mucosa and has shown to improve lesion detection rate, but requires dedicated training and increases procedure time. Photometric stereo endoscopy captures this topography but is qualitative due to unknown working distances from each point of mucosa to the endoscope. In this work, we use deep learning to estimate a depth map from an endoscope camera with four alternating light sources. Since endoscopy videos with ground truth depth maps are challenging to attain, we generated synthetic data using graphical rendering from an anatomically realistic 3D colon model and a forward model of a virtual endoscope with alternating light sources. We propose an encoder-decoder style deep network, where the encoder is split into four branches of sub-encoder networks that simultaneously extract features from each of the four sources and fuse these feature maps as the network goes deeper. This is complemented by skip connections, which maintain spatial consistency when the features are decoded. We demonstrate that, when compared to monocular depth estimation, this setup can reduce the average NRMS error for depth estimation in a silicone colon phantom by 38% and in a pig colon by 31%.","PeriodicalId":309073,"journal":{"name":"Multimodal Biomedical Imaging XIV","volume":"94 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117207573","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Kavon Karrobi, A. Pilvar, Anup Tank, Kshama A. Doshi, D. Waxman, D. Roblyer
{"title":"Towards in vivo preclinical monitoring of multiscale vascular structure-function relationships in resistant breast cancers with an integrated diffuse and nonlinear imaging system (Conference Presentation)","authors":"Kavon Karrobi, A. Pilvar, Anup Tank, Kshama A. Doshi, D. Waxman, D. Roblyer","doi":"10.1117/12.2508872","DOIUrl":"https://doi.org/10.1117/12.2508872","url":null,"abstract":"","PeriodicalId":309073,"journal":{"name":"Multimodal Biomedical Imaging XIV","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134629268","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
R. Maltais-Tariant, C. Boudoux, N. Uribe-Patarroyo
{"title":"Toward co-localized OCT surveillance of laser therapy using real-time speckle decorrelation (Conference Presentation)","authors":"R. Maltais-Tariant, C. Boudoux, N. Uribe-Patarroyo","doi":"10.1117/12.2510413","DOIUrl":"https://doi.org/10.1117/12.2510413","url":null,"abstract":"","PeriodicalId":309073,"journal":{"name":"Multimodal Biomedical Imaging XIV","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130715852","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Émile Beaulieu, Audrey Laurence, M. Latour, R. Albadine, D. Trudel, F. Leblond
Prostate cancer is the most diagnosed form of cancer among American men and, in vast proportion, the standard of care treatment includes radical prostatectomy. Important risk factors associated with prostatectomies are the presence of post-surgery residual prostate tissue and positive cancer margins, potentially leading to recurrences. Prostate histopathology analysis following the procedure is used to determine follow-up treatment. However, only a limited fraction of the prostate margins can be sampled, which can lead to suboptimal evaluation and treatment. Here we present the development of a wide-field multimodal imaging system designed to quantify intrinsic tissue fluorescence and map scattering and absorption coefficients using spatial frequency domain imaging (SFDI). The system allows targeting of suspicious prostate regions to guide histopathology analysis, aiming to improve diagnostic accuracy and treatment planning. Tissue excitation for endogenous fluorescence is achieved with a 405 nm laser diode and, for SFDI, a digital light projector transmits structured white light used to reconstruct tissue optical properties (absorption, scattering) between 420 and 720 nm. A light transport model-based quantification algorithm then corrects the fluorescence spectra for tissue attenuation, lending a biomarker that correlates with local fluorophore concentrations. Spectral and spatial calibration of both modalities was done on optical phantoms and validation of the fluorescence quantification on biological tissue. Finally, imaging results are presented for 5 human prostates interrogated with the system, along with spatially-registered histopathology analyses. Future work involves massive data acquisition and development of artificial intelligence models for tissue classification (prostate, non-prostate; healthy, cancerous) and adaptation for intraoperative use.
{"title":"Post-prostatectomy spatial frequency domain imaging for positive margins identification using endogenous tissue fluorescence, absorption and scattering (Conference Presentation)","authors":"Émile Beaulieu, Audrey Laurence, M. Latour, R. Albadine, D. Trudel, F. Leblond","doi":"10.1117/12.2510067","DOIUrl":"https://doi.org/10.1117/12.2510067","url":null,"abstract":"Prostate cancer is the most diagnosed form of cancer among American men and, in vast proportion, the standard of care treatment includes radical prostatectomy. Important risk factors associated with prostatectomies are the presence of post-surgery residual prostate tissue and positive cancer margins, potentially leading to recurrences. Prostate histopathology analysis following the procedure is used to determine follow-up treatment. However, only a limited fraction of the prostate margins can be sampled, which can lead to suboptimal evaluation and treatment. Here we present the development of a wide-field multimodal imaging system designed to quantify intrinsic tissue fluorescence and map scattering and absorption coefficients using spatial frequency domain imaging (SFDI). The system allows targeting of suspicious prostate regions to guide histopathology analysis, aiming to improve diagnostic accuracy and treatment planning. Tissue excitation for endogenous fluorescence is achieved with a 405 nm laser diode and, for SFDI, a digital light projector transmits structured white light used to reconstruct tissue optical properties (absorption, scattering) between 420 and 720 nm. A light transport model-based quantification algorithm then corrects the fluorescence spectra for tissue attenuation, lending a biomarker that correlates with local fluorophore concentrations. Spectral and spatial calibration of both modalities was done on optical phantoms and validation of the fluorescence quantification on biological tissue. Finally, imaging results are presented for 5 human prostates interrogated with the system, along with spatially-registered histopathology analyses. Future work involves massive data acquisition and development of artificial intelligence models for tissue classification (prostate, non-prostate; healthy, cancerous) and adaptation for intraoperative use.","PeriodicalId":309073,"journal":{"name":"Multimodal Biomedical Imaging XIV","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133032890","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Aditi Sahu, O. Yélamos, N. Iftimia, M. Cordova, C. Alessi-Fox, M. Gill, G. Maguluri, S. Dusza, Cristian Navarrete, S. González, A. Rossi, A. Marghoob, M. Rajadhyaksha, C. Chen
{"title":"Combined reflectance confocal microscopy-optical coherence tomography for detection and deep margin assessment of basal cell carcinomas: a clinical study (Conference Presentation)","authors":"Aditi Sahu, O. Yélamos, N. Iftimia, M. Cordova, C. Alessi-Fox, M. Gill, G. Maguluri, S. Dusza, Cristian Navarrete, S. González, A. Rossi, A. Marghoob, M. Rajadhyaksha, C. Chen","doi":"10.1117/12.2510953","DOIUrl":"https://doi.org/10.1117/12.2510953","url":null,"abstract":"","PeriodicalId":309073,"journal":{"name":"Multimodal Biomedical Imaging XIV","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116057656","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：