Multimodal Deep Learning Methods on Image and Textual Data to Predict Radiotherapy Structure Names
Physicians often label anatomical structure sets in Digital Imaging and Communications in Medicine (DICOM) images with nonstandard random names. Hence, the standardization of these names for the Organs at Risk (OARs), Planning Target Volumes (PTVs), and ‘Other’ organs is a vital problem. This paper...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-06-01
|
Series: | BioMedInformatics |
Subjects: | |
Online Access: | https://www.mdpi.com/2673-7426/3/3/34 |
_version_ | 1797581087579308032 |
---|---|
author | Priyankar Bose Pratip Rana William C. Sleeman Sriram Srinivasan Rishabh Kapoor Jatinder Palta Preetam Ghosh |
author_facet | Priyankar Bose Pratip Rana William C. Sleeman Sriram Srinivasan Rishabh Kapoor Jatinder Palta Preetam Ghosh |
author_sort | Priyankar Bose |
collection | DOAJ |
description | Physicians often label anatomical structure sets in Digital Imaging and Communications in Medicine (DICOM) images with nonstandard random names. Hence, the standardization of these names for the Organs at Risk (OARs), Planning Target Volumes (PTVs), and ‘Other’ organs is a vital problem. This paper presents novel deep learning methods on structure sets by integrating multimodal data compiled from the radiotherapy centers of the US Veterans Health Administration (VHA) and Virginia Commonwealth University (VCU). These de-identified data comprise 16,290 prostate structures. Our method integrates the multimodal textual and imaging data with Convolutional Neural Network (CNN)-based deep learning approaches such as CNN, Visual Geometry Group (VGG) network, and Residual Network (ResNet) and shows improved results in prostate radiotherapy structure name standardization. Evaluation with macro-averaged F1 score shows that our model with single-modal textual data usually performs better than previous studies. The models perform well on textual data alone, while the addition of imaging data shows that deep neural networks achieve better performance using information present in other modalities. Additionally, using masked images and masked doses along with text leads to an overall performance improvement with the CNN-based architectures than using all the modalities together. Undersampling the majority class leads to further performance enhancement. The VGG network on the masked image-dose data combined with CNNs on the text data performs the best and presents the state-of-the-art in this domain. |
first_indexed | 2024-03-10T23:00:10Z |
format | Article |
id | doaj.art-4c5cd532d66047d1beef2232096f4d3d |
institution | Directory Open Access Journal |
issn | 2673-7426 |
language | English |
last_indexed | 2024-03-10T23:00:10Z |
publishDate | 2023-06-01 |
publisher | MDPI AG |
record_format | Article |
series | BioMedInformatics |
spelling | doaj.art-4c5cd532d66047d1beef2232096f4d3d2023-11-19T09:43:29ZengMDPI AGBioMedInformatics2673-74262023-06-013349351310.3390/biomedinformatics3030034Multimodal Deep Learning Methods on Image and Textual Data to Predict Radiotherapy Structure NamesPriyankar Bose0Pratip Rana1William C. Sleeman2Sriram Srinivasan3Rishabh Kapoor4Jatinder Palta5Preetam Ghosh6Department of Computer Science, Virginia Commonwealth University, Richmond, VA 23284, USADepartment of Computer Science, Virginia Commonwealth University, Richmond, VA 23284, USADepartment of Computer Science, Virginia Commonwealth University, Richmond, VA 23284, USADepartment of Radiation Oncology, Virginia Commonwealth University, Richmond, VA 23284, USADepartment of Radiation Oncology, Virginia Commonwealth University, Richmond, VA 23284, USADepartment of Radiation Oncology, Virginia Commonwealth University, Richmond, VA 23284, USADepartment of Computer Science, Virginia Commonwealth University, Richmond, VA 23284, USAPhysicians often label anatomical structure sets in Digital Imaging and Communications in Medicine (DICOM) images with nonstandard random names. Hence, the standardization of these names for the Organs at Risk (OARs), Planning Target Volumes (PTVs), and ‘Other’ organs is a vital problem. This paper presents novel deep learning methods on structure sets by integrating multimodal data compiled from the radiotherapy centers of the US Veterans Health Administration (VHA) and Virginia Commonwealth University (VCU). These de-identified data comprise 16,290 prostate structures. Our method integrates the multimodal textual and imaging data with Convolutional Neural Network (CNN)-based deep learning approaches such as CNN, Visual Geometry Group (VGG) network, and Residual Network (ResNet) and shows improved results in prostate radiotherapy structure name standardization. Evaluation with macro-averaged F1 score shows that our model with single-modal textual data usually performs better than previous studies. The models perform well on textual data alone, while the addition of imaging data shows that deep neural networks achieve better performance using information present in other modalities. Additionally, using masked images and masked doses along with text leads to an overall performance improvement with the CNN-based architectures than using all the modalities together. Undersampling the majority class leads to further performance enhancement. The VGG network on the masked image-dose data combined with CNNs on the text data performs the best and presents the state-of-the-art in this domain.https://www.mdpi.com/2673-7426/3/3/34multimodal data integrationradiotherapy structure namesradiation oncologydeep learningTG-263 names |
spellingShingle | Priyankar Bose Pratip Rana William C. Sleeman Sriram Srinivasan Rishabh Kapoor Jatinder Palta Preetam Ghosh Multimodal Deep Learning Methods on Image and Textual Data to Predict Radiotherapy Structure Names BioMedInformatics multimodal data integration radiotherapy structure names radiation oncology deep learning TG-263 names |
title | Multimodal Deep Learning Methods on Image and Textual Data to Predict Radiotherapy Structure Names |
title_full | Multimodal Deep Learning Methods on Image and Textual Data to Predict Radiotherapy Structure Names |
title_fullStr | Multimodal Deep Learning Methods on Image and Textual Data to Predict Radiotherapy Structure Names |
title_full_unstemmed | Multimodal Deep Learning Methods on Image and Textual Data to Predict Radiotherapy Structure Names |
title_short | Multimodal Deep Learning Methods on Image and Textual Data to Predict Radiotherapy Structure Names |
title_sort | multimodal deep learning methods on image and textual data to predict radiotherapy structure names |
topic | multimodal data integration radiotherapy structure names radiation oncology deep learning TG-263 names |
url | https://www.mdpi.com/2673-7426/3/3/34 |
work_keys_str_mv | AT priyankarbose multimodaldeeplearningmethodsonimageandtextualdatatopredictradiotherapystructurenames AT pratiprana multimodaldeeplearningmethodsonimageandtextualdatatopredictradiotherapystructurenames AT williamcsleeman multimodaldeeplearningmethodsonimageandtextualdatatopredictradiotherapystructurenames AT sriramsrinivasan multimodaldeeplearningmethodsonimageandtextualdatatopredictradiotherapystructurenames AT rishabhkapoor multimodaldeeplearningmethodsonimageandtextualdatatopredictradiotherapystructurenames AT jatinderpalta multimodaldeeplearningmethodsonimageandtextualdatatopredictradiotherapystructurenames AT preetamghosh multimodaldeeplearningmethodsonimageandtextualdatatopredictradiotherapystructurenames |