Summary
A recent discussion focuses on the intricacies of medical data analysis, particularly using the MIMIC database, which contains extensive information from over 60,000 ICU admissions at Beth Israel Deaconess Medical Center. The conversation highlights the challenges of interpreting physiological data, such as heart rates, which can be misleading due to inconsistencies between different information systems used by the hospital. For instance, a notable spike in heart rates was observed in the old CareVue system, attributed to the inclusion of neonatal data that was not captured in the newer MetaVision system.
The discussion also delves into the importance of understanding the context and limitations of medical data. It emphasizes the need for careful analysis to avoid drawing incorrect conclusions, as illustrated by the peculiar case of patients recorded as being 300 years old due to data entry errors related to age confidentiality laws. Furthermore, the conversation outlines the various types of data available in electronic health records, including demographics, vital signs, medications, lab tests, and clinical notes, all of which are essential for comprehensive patient care and research.
The episode underscores the significance of harmonizing medical data standards across different systems to improve interoperability and data accuracy. It concludes with a reminder that thorough knowledge of data is crucial for effective analysis, as misinterpretations can lead to erroneous insights in healthcare research and practice.
Show Notes
MIT 6.S897 Machine Learning for Healthcare, Spring 2019Instructor: Peter SzolovitsView the complete course: https://ocw.mit.edu/6-S897S19YouTube Playlist: https://www.youtube.com/playlist?list=PLUl4u3cNGP60B0PQXVQyGNdCyCTDU1Q5jUploaded 22-10-20Prof. Szolovits gives a deep dive into clinical data, the types of data, and how the data can be standardized. MIMIC-III data is used to illustrate these points. He discusses the difficulties of medicine harmonization and interoperability.License: Creative Commons BY-NC-SAMore information at https://ocw.mit.edu/termsMore courses at https://ocw.mit.edu Learn more about your ad choices. Visit megaphone.fm/adchoices Learn more about your ad choices. Visit megaphone.fm/adchoices
More from MIT OpenCourseWare: Free Courses from the institute
MIT 21M.383 Computational Music Theory and Analysis Spring 2023Instructor: Michael Scott Asato CuthbertView the complete course: https://ocw.mit.edu/courses/21m-383-computational-music-theory-and-analysis-spring-2023/YouTube Playlist: https://www.youtube.com/playlist?list=PLUl4u3cNGP62vSB2sI0W8lQFKsmS2-A6RUploaded 15-10-25This session covers questions about Problem Set 4, an in-class analysis of Bach and Haydn's works, close vs. distant listening, Problem Set 5 guidelines, and additional music21 tools.License: Creative Commons BY-NC-SAMore information at https://ocw.mit.edu/termsMore courses at https://ocw.mit.eduSupport OCW at http://ow.ly/a1If50zVRlWe encourage constructive comments and discussion on OCW’s YouTube and other social media channels. Personal attacks, hate speech, trolling, and inappropriate comments are not allowed and may be removed. More details at https://ocw.mit.edu/comments. Learn more about your ad choices. Visit megaphone.fm/adchoices Learn more about your ad choices. Visit megaphone.fm/adchoices
MIT STS.042J / 8.225J Einstein, Oppenheimer, Feynman: Physics in the 20th Century, Fall 2020Instructor: David KaiserUploaded 22-08-23View the complete course: https://ocw.mit.edu/courses/sts-042-einstein-oppenheimer-feynman-physics-in-the-20th-century-fall-2020YouTube Playlist: https://www.youtube.com/playlist?list=PLUl4u3cNGP63bAfjGas3TuA4ZCPUtN6XfProf. Kaiser discusses early reactions to relativity, Minkowski and spacetime, and Cambridge Wranglers and relativity.License: Creative Commons BY-NC-SAMore information at https://ocw.mit.edu/termsMore courses at https://ocw.mit.eduSupport OCW at http://ow.ly/a1If50zVRlQWe encourage constructive comments and discussion on OCW’s YouTube and other social media channels. Personal attacks, hate speech, trolling, and inappropriate comments are not allowed and may be removed. More details at https://ocw.mit.edu/comments. Learn more about your ad choices. Visit megaphone.fm/adchoices Learn more about your ad choices. Visit megaphone.fm/adchoices
MIT 6.S897 Machine Learning for Healthcare, Spring 2019Instructor: Adam YalaView the complete course: https://ocw.mit.edu/6-S897S19YouTube Playlist: https://www.youtube.com/playlist?list=PLUl4u3cNGP60B0PQXVQyGNdCyCTDU1Q5jUploaded 22-10-20Dr. Yala discusses deep learning models for mammogram interpretation and triaging. While many aspects of cancer detection based on mammograms are similar to those of natural image classification (ImageNet), there are some key differences.License: Creative Commons BY-NC-SAMore information at https://ocw.mit.edu/termsMore courses at https://ocw.mit.edu Learn more about your ad choices. Visit megaphone.fm/adchoices Learn more about your ad choices. Visit megaphone.fm/adchoices
MIT 21M.383 Computational Music Theory and Analysis Spring 2023Instructor: Michael Scott Asato CuthbertView the complete course: https://ocw.mit.edu/courses/21m-383-computational-music-theory-and-analysis-spring-2023/YouTube Playlist: https://www.youtube.com/playlist?list=PLUl4u3cNGP62vSB2sI0W8lQFKsmS2-A6RUploaded 15-10-25The instructor presents a brief synopsis of topics covered by guest speakers in the past few class sessions.License: Creative Commons BY-NC-SAMore information at https://ocw.mit.edu/termsMore courses at https://ocw.mit.eduSupport OCW at http://ow.ly/a1If50zVRlWe encourage constructive comments and discussion on OCW’s YouTube and other social media channels. Personal attacks, hate speech, trolling, and inappropriate comments are not allowed and may be removed. More details at https://ocw.mit.edu/comments. Learn more about your ad choices. Visit megaphone.fm/adchoices Learn more about your ad choices. Visit megaphone.fm/adchoices
MIT 6.S897 Machine Learning for Healthcare, Spring 2019Instructor: Peter SzolovitsView the complete course: https://ocw.mit.edu/6-S897S19YouTube Playlist: https://www.youtube.com/playlist?list=PLUl4u3cNGP60B0PQXVQyGNdCyCTDU1Q5jUploaded 22-10-20Prof. Szolovits discusses ways of improving medical care and narrowing performance distribution through decision support from local data and change-point detection to monitor rule firings. The Clinical Pathways COWpath study is used as an example.License: Creative Commons BY-NC-SAMore information at https://ocw.mit.edu/termsMore courses at https://ocw.mit.edu Learn more about your ad choices. Visit megaphone.fm/adchoices Learn more about your ad choices. Visit megaphone.fm/adchoices