TY - GEN
T1 - Machine Learning for Rhabdomyosarcoma Whole Slide Images Sub-type Classification
AU - Yadav, Ankur
AU - Daescu, Ovidiu
AU - Leavey, Patrick
AU - Rudzinski, Erin
N1 - Funding Information:
Material for this report was provided as part of Protocol ARST18B4-Q by the Children’s Oncology Group through support from the following grants: NCTN Operations Center Grant U10CA180886 NCTN SDC Grant U10CA180899 COG Biospecimen Bank Grant U24CA196173
Publisher Copyright:
© 2023 Owner/Author.
PY - 2023/7/5
Y1 - 2023/7/5
N2 - The most frequent malignant soft tissue tumor in children is Rhabdomyosarcoma (RMS). RMS has several subtypes that differentiate treatment and patient outcomes. Because of variations in the appearance of histopathology images, manual subtype classification requires a high level of expertise and is time-consuming. While several machine-learning techniques have been developed to classify the most common tumor types in histology images, more must be understood about the automatic classification of tumor subtypes. Moreover, existing techniques for classifying whole slide image (WSI) histopathological subtypes rely on small, randomly selected image tiles or on representative tiles chosen by specialists from the considerably larger WSIs. These methods do not draw knowledge from the whole tissue region captured by a WSI, possibly missing significant tumor signatures. They also fail to account for the spatial distribution of patterns that could play a role in reliable subtype classification. This paper proposes a novel methodology that combines different methods to extract features from a whole slide image and generates a whole slide feature map (WSFM). This map and additional clinical features are then used to train various machine-learning models. We obtain 91.84% WSI tumor subtype classification accuracy on a diverse dataset. A direct advantage of our methodology is that it does not require any WSI-level annotation by pathology experts. Training and testing can be performed much faster computationally using simple machine learning algorithms instead of complex deep learning architectures.
AB - The most frequent malignant soft tissue tumor in children is Rhabdomyosarcoma (RMS). RMS has several subtypes that differentiate treatment and patient outcomes. Because of variations in the appearance of histopathology images, manual subtype classification requires a high level of expertise and is time-consuming. While several machine-learning techniques have been developed to classify the most common tumor types in histology images, more must be understood about the automatic classification of tumor subtypes. Moreover, existing techniques for classifying whole slide image (WSI) histopathological subtypes rely on small, randomly selected image tiles or on representative tiles chosen by specialists from the considerably larger WSIs. These methods do not draw knowledge from the whole tissue region captured by a WSI, possibly missing significant tumor signatures. They also fail to account for the spatial distribution of patterns that could play a role in reliable subtype classification. This paper proposes a novel methodology that combines different methods to extract features from a whole slide image and generates a whole slide feature map (WSFM). This map and additional clinical features are then used to train various machine-learning models. We obtain 91.84% WSI tumor subtype classification accuracy on a diverse dataset. A direct advantage of our methodology is that it does not require any WSI-level annotation by pathology experts. Training and testing can be performed much faster computationally using simple machine learning algorithms instead of complex deep learning architectures.
KW - Machine Learning
KW - Rhabdomyosarcoma
KW - Subtype Classification
KW - Whole Slide Feature Map
UR - http://www.scopus.com/inward/record.url?scp=85170374428&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85170374428&partnerID=8YFLogxK
U2 - 10.1145/3594806.3594865
DO - 10.1145/3594806.3594865
M3 - Conference contribution
AN - SCOPUS:85170374428
T3 - ACM International Conference Proceeding Series
SP - 192
EP - 196
BT - 16th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2023
PB - Association for Computing Machinery
T2 - 16th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2023
Y2 - 5 July 2023 through 7 July 2023
ER -