TY - GEN
T1 - Predicting Scores of Medical Imaging Segmentation Methods with Meta-learning
AU - van Sonsbeek, Tom
AU - Cheplygina, Veronika
PY - 2020
Y1 - 2020
N2 - Deep learning has led to state-of-the-art results for many medical imaging tasks, such as segmentation of different anatomical structures. With the increased numbers of deep learning publications and openly available code, the approach to choosing a model for a new task becomes more complicated, while time and (computational) resources are limited. A possible solution to choosing a model efficiently is meta-learning, a learning method in which prior performance of a model is used to predict the performance for new tasks. We investigate meta-learning for segmentation across ten datasets of different organs and modalities. We propose four ways to represent each dataset by meta-features: one based on statistical features of the images and three are based on deep learning features. We use support vector regression and deep neural networks to learn the relationship between the meta-features and prior model performance. On three external test datasets these methods give Dice scores within 0.10 of the true performance. These results demonstrate the potential of meta-learning in medical imaging.
AB - Deep learning has led to state-of-the-art results for many medical imaging tasks, such as segmentation of different anatomical structures. With the increased numbers of deep learning publications and openly available code, the approach to choosing a model for a new task becomes more complicated, while time and (computational) resources are limited. A possible solution to choosing a model efficiently is meta-learning, a learning method in which prior performance of a model is used to predict the performance for new tasks. We investigate meta-learning for segmentation across ten datasets of different organs and modalities. We propose four ways to represent each dataset by meta-features: one based on statistical features of the images and three are based on deep learning features. We use support vector regression and deep neural networks to learn the relationship between the meta-features and prior model performance. On three external test datasets these methods give Dice scores within 0.10 of the true performance. These results demonstrate the potential of meta-learning in medical imaging.
KW - Feature extraction
KW - Meta-learning
KW - Segmentation
UR - http://www.scopus.com/inward/record.url?scp=85092942413&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-61166-8_26
DO - 10.1007/978-3-030-61166-8_26
M3 - Conference contribution
AN - SCOPUS:85092942413
SN - 9783030611651
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 242
EP - 253
BT - Interpretable and Annotation-Efficient Learning for Medical Image Computing - 3rd International Workshop, iMIMIC 2020, 2nd International Workshop, MIL3iD 2020, and 5th International Workshop, LABELS 2020, Held in Conjunction with MICCAI 2020, Proceedings
A2 - Cardoso, Jaime
A2 - Silva, Wilson
A2 - Cruz, Ricardo
A2 - Van Nguyen, Hien
A2 - Roysam, Badri
A2 - Heller, Nicholas
A2 - Henriques Abreu, Pedro
A2 - Pereira Amorim, Jose
A2 - Isgum, Ivana
A2 - Patel, Vishal
A2 - Zhou, Kevin
A2 - Jiang, Steve
A2 - Le, Ngan
A2 - Luu, Khoa
A2 - Sznitman, Raphael
A2 - Cheplygina, Veronika
A2 - Abbasi, Samaneh
A2 - Mateus, Diana
A2 - Trucco, Emanuele
PB - Springer
T2 - LABELS 2020
Y2 - 4 October 2020 through 8 October 2020
ER -