The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Multi-task Learning for Melanoma Detection with Julianna Ianni - #531
Oct 28, 2021
Julianna Ianni, VP of AI Research & Development at Proscia, discusses groundbreaking advancements in using deep learning for melanoma detection. She highlights the development of a multitask classifier, improving accuracy in distinguishing low-risk from high-risk cases. Julianna also elaborates on the challenges of achieving consensus among pathologists and the complexities of training AI to handle image artifacts. As AI transforms cancer diagnostics, she shares insights into the promising future for integrating these technologies into clinical practices.
AI Snips
Chapters
Transcript
Episode notes
Julianna's Path to Machine Learning
- Julianna Ianni's interest in helping people led her to biomedical engineering and informatics.
- A negative mentorship experience inadvertently sparked her interest in machine learning.
Pathology vs. Radiology
- Pathology images, digitized slides of biopsied tissue, differ from radiology images like MRIs.
- Proscia uses deep learning to improve pathologists' efficiency and diagnostic accuracy.
Ground Truth Challenges in Melanoma Diagnosis
- Creating ground truth for melanoma diagnosis is difficult due to high pathologist disagreement.
- The study used consensus from three dermatopathologists as ground truth, discarding cases with disagreement.