Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning

Background and study aims Storage of full-length endoscopic procedures is becoming increasingly popular. To facilitate large-scale machine learning (ML) focused on clinical outcomes, these videos must be merged with the patient-level data in the electronic health record (EHR). Our aim was to present...

Full description

Bibliographic Details
Main Authors: Rajesh N. Keswani, Daniel Byrd, Florencia Garcia Vicente, J. Alex Heller, Matthew Klug, Nikhilesh R. Mazumder, Jordan Wood, Anthony D. Yang, Mozziyar Etemadi
Format: Article
Language:English
Published: Georg Thieme Verlag KG 2021-02-01
Series:Endoscopy International Open
Online Access:http://www.thieme-connect.de/DOI/DOI?10.1055/a-1326-1289
id doaj-68318bf5da744320baca9248b9388f1a
record_format Article
spelling doaj-68318bf5da744320baca9248b9388f1a2021-02-04T00:50:00ZengGeorg Thieme Verlag KGEndoscopy International Open2364-37222196-97362021-02-010902E233E23810.1055/a-1326-1289Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learningRajesh N. Keswani0Daniel Byrd1Florencia Garcia Vicente2J. Alex Heller3Matthew Klug4Nikhilesh R. Mazumder5Jordan Wood6Anthony D. Yang7Mozziyar Etemadi8Digestive Health Center, Northwestern Medicine, Chicago, Illinois, United StatesDepartment of Anesthesiology, Northwestern Medicine, Chicago, Illinois, United StatesDepartment of Anesthesiology, Northwestern Medicine, Chicago, Illinois, United StatesDepartment of Anesthesiology, Northwestern Medicine, Chicago, Illinois, United StatesDepartment of Anesthesiology, Northwestern Medicine, Chicago, Illinois, United StatesDigestive Health Center, Northwestern Medicine, Chicago, Illinois, United StatesDigestive Health Center, Northwestern Medicine, Chicago, Illinois, United StatesSurgical Outcomes and Quality Improvement Center, Department of Surgery, Northwestern University Feinberg School of Medicine, Chicago, Illinois, United StatesDepartment of Anesthesiology, Northwestern Medicine, Chicago, Illinois, United StatesBackground and study aims Storage of full-length endoscopic procedures is becoming increasingly popular. To facilitate large-scale machine learning (ML) focused on clinical outcomes, these videos must be merged with the patient-level data in the electronic health record (EHR). Our aim was to present a method of accurately linking patient-level EHR data with cloud stored colonoscopy videos. Methods This study was conducted at a single academic medical center. Most procedure videos are automatically uploaded to the cloud server but are identified only by procedure time and procedure room. We developed and then tested an algorithm to match recorded videos with corresponding exams in the EHR based upon procedure time and room and subsequently extract frames of interest. Results Among 28,611 total colonoscopies performed over the study period, 21,170 colonoscopy videos in 20,420 unique patients (54.2 % male, median age 58) were matched to EHR data. Of 100 randomly sampled videos, appropriate matching was manually confirmed in all. In total, these videos represented 489,721 minutes of colonoscopy performed by 50 endoscopists (median 214 colonoscopies per endoscopist). The most common procedure indications were polyp screening (47.3 %), surveillance (28.9 %) and inflammatory bowel disease (9.4 %). From these videos, we extracted procedure highlights (identified by image capture; mean 8.5 per colonoscopy) and surrounding frames. Conclusions We report the successful merging of a large database of endoscopy videos stored with limited identifiers to rich patient-level data in a highly accurate manner. This technique facilitates the development of ML algorithms based upon relevant patient outcomes.http://www.thieme-connect.de/DOI/DOI?10.1055/a-1326-1289
collection DOAJ
language English
format Article
sources DOAJ
author Rajesh N. Keswani
Daniel Byrd
Florencia Garcia Vicente
J. Alex Heller
Matthew Klug
Nikhilesh R. Mazumder
Jordan Wood
Anthony D. Yang
Mozziyar Etemadi
spellingShingle Rajesh N. Keswani
Daniel Byrd
Florencia Garcia Vicente
J. Alex Heller
Matthew Klug
Nikhilesh R. Mazumder
Jordan Wood
Anthony D. Yang
Mozziyar Etemadi
Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning
Endoscopy International Open
author_facet Rajesh N. Keswani
Daniel Byrd
Florencia Garcia Vicente
J. Alex Heller
Matthew Klug
Nikhilesh R. Mazumder
Jordan Wood
Anthony D. Yang
Mozziyar Etemadi
author_sort Rajesh N. Keswani
title Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning
title_short Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning
title_full Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning
title_fullStr Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning
title_full_unstemmed Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning
title_sort amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning
publisher Georg Thieme Verlag KG
series Endoscopy International Open
issn 2364-3722
2196-9736
publishDate 2021-02-01
description Background and study aims Storage of full-length endoscopic procedures is becoming increasingly popular. To facilitate large-scale machine learning (ML) focused on clinical outcomes, these videos must be merged with the patient-level data in the electronic health record (EHR). Our aim was to present a method of accurately linking patient-level EHR data with cloud stored colonoscopy videos. Methods This study was conducted at a single academic medical center. Most procedure videos are automatically uploaded to the cloud server but are identified only by procedure time and procedure room. We developed and then tested an algorithm to match recorded videos with corresponding exams in the EHR based upon procedure time and room and subsequently extract frames of interest. Results Among 28,611 total colonoscopies performed over the study period, 21,170 colonoscopy videos in 20,420 unique patients (54.2 % male, median age 58) were matched to EHR data. Of 100 randomly sampled videos, appropriate matching was manually confirmed in all. In total, these videos represented 489,721 minutes of colonoscopy performed by 50 endoscopists (median 214 colonoscopies per endoscopist). The most common procedure indications were polyp screening (47.3 %), surveillance (28.9 %) and inflammatory bowel disease (9.4 %). From these videos, we extracted procedure highlights (identified by image capture; mean 8.5 per colonoscopy) and surrounding frames. Conclusions We report the successful merging of a large database of endoscopy videos stored with limited identifiers to rich patient-level data in a highly accurate manner. This technique facilitates the development of ML algorithms based upon relevant patient outcomes.
url http://www.thieme-connect.de/DOI/DOI?10.1055/a-1326-1289
work_keys_str_mv AT rajeshnkeswani amalgamationofcloudbasedcolonoscopyvideoswithpatientlevelmetadatatofacilitatelargescalemachinelearning
AT danielbyrd amalgamationofcloudbasedcolonoscopyvideoswithpatientlevelmetadatatofacilitatelargescalemachinelearning
AT florenciagarciavicente amalgamationofcloudbasedcolonoscopyvideoswithpatientlevelmetadatatofacilitatelargescalemachinelearning
AT jalexheller amalgamationofcloudbasedcolonoscopyvideoswithpatientlevelmetadatatofacilitatelargescalemachinelearning
AT matthewklug amalgamationofcloudbasedcolonoscopyvideoswithpatientlevelmetadatatofacilitatelargescalemachinelearning
AT nikhileshrmazumder amalgamationofcloudbasedcolonoscopyvideoswithpatientlevelmetadatatofacilitatelargescalemachinelearning
AT jordanwood amalgamationofcloudbasedcolonoscopyvideoswithpatientlevelmetadatatofacilitatelargescalemachinelearning
AT anthonydyang amalgamationofcloudbasedcolonoscopyvideoswithpatientlevelmetadatatofacilitatelargescalemachinelearning
AT mozziyaretemadi amalgamationofcloudbasedcolonoscopyvideoswithpatientlevelmetadatatofacilitatelargescalemachinelearning
_version_ 1724285715497353216