The Society for Technology in Anesthesia is pleased to announce it's Annual Engineering Challenge to be held in conjunction with the STA 2023 Annual Meeting: Machine Learning for Characterization of Slurred Speech Following Anesthesia
The Engineering Challenge is a regular feature of the Annual Meeting, and will be held Saturday, January 14, 2023. As always, the intent of the Challenge is to educate the membership on an area of technology.
All anesthesiologists are familiar with the tendency of anesthesia to produce a period of slurred speech during recovery. While slurred speech is a frequently employed endpoint for conscious sedation with benzodiazepines, no systematic assessment of recovery from anesthetic effects has employed slurred speech as a marker. An objective measure of the time course of recovery from anesthesia utilizing machine learning approaches is the subject of the 2023 STA Engineering Challenge. As always, the intent of the Challenge is to educate the membership and to foster the entry of talent into the specialty and the Society.
Parkinson’s Disease is a progressive neurodegenerative disease of which dysarthria is a noticeable feature. Considerable effort has been invested in tracking the progression of Parkinson’s Disease with machine learning approaches. Several recent reviews are available:
Engineering Challenge Guidelines:
- The method should be able to distinguish speech recorded preoperatively to that in the PACU (i.e., slurred vs. normal speech).
- The method should work with short samples of speech (i.e., less than one minute).
- Presenters should explain the machine learning approach taken.
- It is not necessary that the demonstrated method utilize data from patients; it is acceptable to use samples collected from volunteers imbibing refreshments typically provided by the Society at the Members’ Reception. “Voiceprints” are explicitly listed in the CFR as patient identifiers. While it may be possible to obtain a waiver of confidentiality for recording short segments of voice for analysis, this may be complicated and will require investing time dealing with your IRB.
- At least one member of the team should be a trainee – student, resident, fellow, etc.
- Presentations may be in person or remote. A member of the team must contact Dr. Mandel prior to the start of the meeting to confirm intent to present.
- Should actual patient data be used in this effort, it MUST be obtained with approval of your IRB. Because this may be challenging, any entrant meeting this requirement and presenting their project will receive $500 irrespective of whether their project wins the competition.
- First Prize: $1,000
- Second Prize: $500
- Third Prize: $250
STA Engineering Challenge Contact Information:
For questions on the challenge specifications contact:
Jeff E. Mandel, MD, MS
Chair, STA Engineering Challenge