false
Home
AOSSM 2023 Annual Meeting Recordings no CME
Q & A: Human vs. Machine: De-bugging Artificial I ...
Q & A: Human vs. Machine: De-bugging Artificial Intelligence and Machine Learning in Sports Medicine
Back to course
[Please upgrade your browser to play this video content]
Video Transcription
Video Summary
In this video, the speakers address questions from the audience about the limitations of their image models and the need for external validation. Dr. Liu explains that the resolution requirement for segmentation is not very high and that the biggest limitation comes from the rotation and quality of lateral knee x-rays. They manually curated the images and eliminated those with poor quality. On the other hand, Dr. Ingbertson discusses the need for high-quality and large amounts of data for deep learning and machine learning. They also mention using an open access database for external validation. Lastly, they mention a logistic regression study that provides patients with a probability of healing, and the machine learning algorithm can provide a level of certainty and confidence interval for healing chances. The video then moves on to the next speaker.
Asset Caption
Christopher Camp, MD; Yining Lu, MD; Thomas Johnstone, BS; Ryan Ingebritsen, BS; Danny Goel
Keywords
limitations
external validation
image models
deep learning
machine learning
×
Please select your language
1
English