Automated Muscle Segmentation from Clinical CT Using Bayesian U-Net for Personalized Musculoskeletal Modeling
Open accessThe DOI matches its Crossref record; 43 of the 45 references it cites were found in a registry or an index.
Abstract
We propose a method for automatic segmentation of individual muscles from a clinical CT. The method uses Bayesian convolutional neural networks with the U-Net architecture, using Monte Carlo dropout that infers an uncertainty metric in addition to the segmentation label. We evaluated the performance of the proposed method using two data sets: 20 fully annotated CTs of the hip and thigh regions and 18 partially annotated CTs that are publicly available from The Cancer Imaging Archive (TCIA) database. The experiments showed a Dice coefficient (DC) of 0.891±0.016 (mean±std) and an average symmetric surface distance (ASD) of 0.994±0.230 mm over 19 muscles in the set of 20 CTs. These results were statistically significant improvements compared to the state-of-the-art hierarchical multi-atlas method which resulted in 0.845 ± 0.031 DC and 1.556 ± 0.444 mm ASD. We evaluated validity of the uncertainty metric in the multi-class organ segmentation problem and demonstrated a correlation between the pixels with high uncertainty and the segmentation failure. One application of the uncertainty metric in active-learning is demonstrated, and the proposed query pixel selection method considerably reduced the manual annotation cost for expanding the training data set. The proposed method allows an accurate patient-specific analysis of individual muscle shapes in a clinical routine. This would open up various applications including personalization of biomechanical simulation and quantitative evaluation of muscle atrophy.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Xue al. [685] | et | 2019 | BNN and U-Net | Classification | Gigapixel phase im- ages | 1 | MC dropout and ensembles”
From A review of uncertainty quantification in deep learning: Techniques, applications and challenges (2021) · cites this paper as [685]
Citations
Cites
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 4×
“the weights in the same way as in [9], and then trained the networks using adaptive moment estimation (Adam) [16] for 1×105 iterations at the learning rate of 0.”
MethodInfluentialFully convolutional networks for semantic segmentation
Long, J., Shelhamer, E., & Darrell, T. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298965 · Cited by 38,834
(2014) · Cited by 33,419
Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification
He, K., Zhang, X., Ren, S., & Sun, J. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.123 · Cited by 19,741 · cited 1×
“We initialized the weights in the same way as in [9], and then trained the networks using adaptive moment estimation (Adam) [16] for 1 × 10 5 iterations at the learning rate of 0.0001.”
MethodInfluentialMeasures of the Amount of Ecologic Association Between Species
Dice, L. R. (1945) · Ecology · doi:10.2307/1932409 · Cited by 11,874 · cited 1×
“We used the Dice coefficient (DC) [5] and the average symmetric surface distance (ASD) [31] as the error metrics.”
MethodDropout as a bayesian approximation: Representing model uncertainty in deep learning
Gal, Y., & Ghahramani, Z. (2015) · arXiv · doi:10.48550/arxiv.1506.02142 · Cited by 6,471 · cited 6×
“Below, we briefly summarize the theoretical background described in [7], formulate the specific metric that we employed in this paper, and propose a new structure-wise”
BackgroundMethodInfluential(2009) · Minds at UW (University of Wisconsin) · Cited by 4,791
What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
Kendall, A., & Gal, Y. (2017) · arXiv · doi:10.48550/arxiv.1703.04977 · Cited by 3,088 · cited 1×
“It measures the degree of difference of each test sample from the training data set, originated from the deficiency of training data, namely epistemic uncertainty [15].”
BackgroundSeth, A., Hicks, J. L., Uchida, T. K., Habib, A., Dembia, C. L., Dunne, J. J., Ong, C. F., DeMers, M. S., Rajagopal, A., Millard, M., et al. (2018) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1006223 · Cited by 1,322 · cited 1×
“The computational simulation of human motion using musculoskeletal modeling has been performed in a number of studies to investigate musculo-tendon forces and joint contact forces, which cannot be easily achieved by physical measurements [25], [27], [29].”
BackgroundFull-Body Musculoskeletal Model for Muscle-Driven Simulation of Human Gait
Rajagopal, A., Dembia, C. L., DeMers, M. S., Delp, D. D., Hicks, J. L., & Delp, S. L. (2016) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2016.2586891 · Cited by 1,053
Research works printed but not linked: 2
- hiasa, 2018, Surgical tools segmentation in laparoscopic images using convolutional neural networks with uncertainty estimation and semi-supervised learning, Proc Int Conf CARS, 14
- hiasa, 2018, Cross-modality image synthesis from unpaired data using CycleGAN, Proc Int Workshop SASHIMI, 31
Cited by
Items 101–156 of 156
Gu, Y., Otake, Y., Uemura, K., Takao, M., Soufi, M., Hiasa, Y., Talbot, H., Okada, S., Sugano, N., & Sato, Y. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-43990-2_47 · Cited by 5
Self-citationLu, J., Wang, Y., Huang, H., Chen, Z., Wang, Y., Cui, Z., Huang, J., Liang, Y., Su, Z., Chen, J., et al. (2025) · European Spine Journal · doi:10.1007/s00586-025-09541-1 · Cited by 4
Decomposition of musculoskeletal structures from radiographs using an improved CycleGAN framework
Nakanishi, N., Otake, Y., Hiasa, Y., Gu, Y., Uemura, K., Takao, M., Sugano, N., & Sato, Y. (2023) · Scientific Reports · doi:10.1038/s41598-023-35075-x · Cited by 4
Self-citationExploring Structure-Wise Uncertainty for 3D Medical Image Segmentation
Vasiliuk, A., Frolova, D., Belyaev, M., & Shirokikh, B. (2023) · Lecture notes in electrical engineering · doi:10.1007/978-981-16-6775-6_2 · Cited by 5
Ensemble learning via supervision augmentation for white matter hyperintensity segmentation
Guo, X., Ye, C., Yang, Y., Zhang, L., Li, L., Lu, S., Lv, H., Guo, C., & Ma, T. (2022) · Frontiers in Neuroscience · doi:10.3389/fnins.2022.946343 · Cited by 4
MPLDM: Multi-modal prosthetic loosening diagnostic model for total hip arthroplasty
Chen, X., Lyu, P., Han, W., Yang, L., Jin, Y., Zhang, T. G., & Shen, J. (2025) · Medical Image Analysis · doi:10.1016/j.media.2025.103921 · Cited by 2
Lindbeck, E. M., Diaz, M. T., Nichols, J. A., & Harley, J. B. (2023) · Journal of Biomechanics · doi:10.1016/j.jbiomech.2023.111834 · Cited by 4
Efficient musculoskeletal annotation using free-form deformation
Fukuda, N., Konda, S., Umehara, J., & Hirashima, M. (2024) · Scientific Reports · doi:10.1038/s41598-024-67125-3 · Cited by 3
Gu, Y., Otake, Y., Uemura, K., Takao, M., Soufi, M., Okada, S., Sugano, N., Talbot, H., & Sato, Y. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72104-5_1 · Cited by 2
“CT bone segmentation [18] and X-ray 2D-3D registration [30] were performed to produce ground truth bone 3D shapes and depth maps.”
MethodSelf-citationDynamic-budget superpixel active learning for semantic segmentation
Wang, Y., & Stavness, I. (2025) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2024.1498956 · Cited by 2
“Both models are implemented (Hiasa et al., 2019; Yu et al., 2020) with MC-Dropout (Gal and Ghahramani, 2015, 2016) to approximate a Bayesian Neural Network (BNN) for better uncertainty estimation.”
A systematic review of active learning approaches in the selection of medical images
Santos, M. C. D. S., & Marreiros, G. (2025) · Procedia Computer Science · doi:10.1016/j.procs.2025.02.186 · Cited by 5
Recent Collaborative R&D in Medical Imaging in Asia
Lin, F., Kim, J. H., Sato, Y., & Kung, S. (2020) · Journal of Signal Processing Systems · doi:10.1007/s11265-019-01514-0 · Cited by 2
“using uncertainty estimated from Bayesian U-net [34].”
MethodSelf-citationRecent Progress of Medical CT Image Processing Based on Deep Learning
Tan, Y., Qin, J., Huang, L., Tan, L., Xiang, X., Tang, H., Peng, H., & Wu, J. Y. (2021) · Communications in computer and information science · doi:10.1007/978-3-030-78615-1_37 · Cited by 2
Gomariz, Á., Egli, R., Portenier, T., Nombela‐Arrieta, C., & Göksel, O. (2021) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi48211.2021.9434158 · Cited by 2
“Different uncertainties are estimated following the frameworks proposed in [4, 5] and included within a MS-ME model f that is applied on an input image x to predict its corresponding segmentation ŷ = softmax(z), where z are the logits resulting from the model: z = f(x).”
MethodIllinger, V., Slabý, K., Suchánek, V., & Radvanský, J. (2024) · Cureus · doi:10.7759/cureus.56276 · Cited by 1
Chen, Y. J., Soufi, M., Uemura, K., Otake, Y., Takao, M., Iwakoshi, S., Tanaka, T., Sugano, N., & Sato, Y. (2024) · Advanced Biomedical Engineering · doi:10.14326/abe.13.176 · Cited by 1
Self-citationTang, W., Dai, Y., Zhang, B., Shi, Z., Chu, Y.-H., Zhuang, P., Shen, D. G., Wang, C., & Wang, H. (2025) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2025.3564783 · Cited by 1
Bone mineral density measurement in the Gruen zones using dual-energy x-ray absorptiometry
Uemura, K., Otake, Y., Tamura, K., Higuchi, R., Kono, S., Mae, H., Takashima, K., Okada, S., Sugano, N., & Hamada, H. (2025) · Bone and Joint Research · doi:10.1302/2046-3758.1410.bjr-2025-0036.r1 · Cited by 1
“30,31 Then, the surface model of the stem region was registered to the CAD model to precisely delineate the position of the stem in the postopera tive CT coordinates using iterative closest point (ICP) regis tration.”
Self-citationSotaro, K., Uemura, K., Soufi, M., Nishimura, R., Miyamoto, T., Higuchi, R., Mae, H., Takashima, K., Otake, Y., Tanaka, Y., et al. (2025) · International Journal of Computer Assisted Radiology and Surgery · doi:10.1007/s11548-025-03531-w · Cited by 1
“1) [14, 15].”
Self-citationYang, J., & Yu, X. (2026) · Neurocomputing · doi:10.1016/j.neucom.2026.133546 · Cited by 1
Asano, T., Uemura, K., Kono, S., Higuchi, R., Gu, Y., Mae, H., Tani, T., Takashima, K., Kutsuna, T., Otake, Y., et al. (2026) · Osteoporosis and Sarcopenia · doi:10.1016/j.afos.2026.05.001 · Cited by 1
“Bone and muscle structures on the unfractured side were automatically segmented from CT images using a deep learning – based system (Bayesian U-Net) that we previously developed and validated [12,13].”
Self-citationHyakuda, Y., Miyamoto, T., Ogawa, M., Kawai, A., Nishimura, H., Hara, R., Inagaki, Y., Uemura, K., Takao, M., Kono, K., et al. (2026) · BMC Musculoskeletal Disorders · doi:10.1186/s12891-026-10034-5 · Cited by 2
“We used a deep learning model based on a fully convolutional neural network with a Bayesian U-Net architecture [13, 14], which was previously shown to provide high reliability for the muscle segmentation of bones and soft tissues from hip-to-knee CT [11].”
Self-citationLi, G., Otake, Y., Soufi, M., Masuda, M., Uemura, K., Takao, M., Sugano, N., & Sato, Y. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72384-1_35 · Cited by 1
Self-citationA NOVEL ITERATIVE MATCHING SCHEME BASED ON HOMOGRAPHY METHOD FOR X-RAY IMAGE
Song, L., Zou, H., JI, Z., Xie, X., & LI, W. (2020) · Journal of Mechanics in Medicine and Biology · doi:10.1142/s0219519420500384 · Cited by 1
Extraction of lung and lesion regions from COVID-19 CT volumes using 3D fully convolutional networks
Hayashi, Y., Oda, M., Shen, C., Hashimoto, M., Otake, Y., Akashi, T., & Mori, K. (2021) · SPIE Medical Imaging: Computer-Aided Diagnosis · doi:10.1117/12.2581818 · Cited by 1
Self-citationZhao, H., Wang, Y., He, C., Yang, J., Shi, Y., & Zhu, X. (2022) · Journal of Healthcare Engineering · doi:10.1155/2022/3951112 · Cited by 1
“It shows that adding a dropout layer can prevent the model from overfitting or generalization, and it also improves the segmentation ability and robustness of the model [21].”
BackgroundNizarudeen, S., & Shunmugavel, G. R. (2022) · Lecture notes in electrical engineering · doi:10.1007/978-981-19-1111-8_43 · Cited by 1
Gomariz, Á., Portenier, T., Nombela‐Arrieta, C., & Göksel, O. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.11865 · Cited by 1
“Bayesian regression networks for calibrated detections Motivated by the advantages in calibration with the use of different Bayesian DL techniques reported for different domains (52, 57, 58, 61), we adapt our framework to take into account any uncertainties in DM regression.”
MethodKawasaki, A., Otake, Y., Uemura, K., Takao, M., Sugano, N., & Sato, Y. (2020) · EPiC series in health sciences · doi:10.29007/4h7v
“We used the Bayesian U-Net [9] for segmentation of the sacrum region, cropped the volume with its bounding box, and replaced the voxels outside the sacrum with -1000 [HU].”
MethodSelf-citationFundamental Technologies for Integration and Pathology in MCA
Sato, Y., & Otake, Y. (2021) · Springer eBooks · doi:10.1007/978-981-16-4325-5_5
Self-citationTakao, M. (2020) · Journal of Japan Society of Computer Aided Surgery · doi:10.5759/jscas.22.172
Self-citationNie, X., Liu, X., Yang, H., Shi, F., Gu, W., Hou, X., Wei, Y., Lu, Q., Bai, H., Chen, J., et al. (2022) · Neural Computing and Applications · doi:10.1007/s00521-022-08094-4
“Deeplabv3? [50], U-Net [51], U-Net?? [52], OC-Net [53], PSP-Net [54], SF-Net [55] and Inf-Net [56] Neural Computing and Applications (2023) 35:22101–22114 22109”
InfluentialTakao, M. (2022) · Journal of Japan Society of Computer Aided Surgery · doi:10.5759/jscas.24.187
Self-citationDevelopment of a System to Measure the Bone Mineral Density of the Proximal Femur from CT Images
Uemura, K., Otake, Y., Takao, M., Sato, Y., Okada, S., & Sugano, N. (2023) · Journal of Japan Society of Computer Aided Surgery · doi:10.5759/jscas.25.22
Self-citationYüce, M. (2024) · Türk Ortopedi ve Travmatoloji Birliği Derneği · doi:10.5578/totbid.dergisi.2024.05
Can thigh circumference predict thigh muscle volume? -Analysis using deep learning-
Kono, S., Uemura, K., Soufi, M., Iwasa, M., Takashima, K., Hamada, H., Otake, Y., Takao, M., Sato, Y., Okada, S., et al. (2024) · EPiC series in health sciences · doi:10.29007/wlr7
Self-citationGu, Y., Otake, Y., Uemura, K., Takao, M., Soufi, M., Okada, S., Sugano, N., Talbot, H., & Sato, Y. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73281-2_13
Self-citationFrom out-of-distribution detection to quality control
Lambert, B., Forbes, F., & Dojat, M. (2025) · Elsevier eBooks · doi:10.1016/b978-0-44-323761-4.00014-6 · Cited by 2
RETRACTED: Federated learning-based semantic segmentation framework for sustainable development
Godavarthi, D., Jose, D., Mohanty, S. N., Medani, M., Kallel, M., Abdullaev, S. S., & Khan, M. I. (2025) · Egyptian Informatics Journal · doi:10.1016/j.eij.2025.100702 · Cited by 4
RetractedSato, Y., Otake, Y., Kagiyama, Y., Uemura, K., Takao, M., & Sugano, N. (2025) · Springer eBooks · doi:10.1007/978-981-96-6635-5_10
Self-citationYang, J., Yu, X., Qiu, P., Marcus, D. S., & Sotiras, A. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04978-0_1 · Cited by 1
Estimating Bone Mineral Density and Muscle Mass from EOS Low Dose X-Ray Imaging System
Suehara, K., Gu, Y., Otake, Y., Uemura, K., Okamoto, M., Tokunaga, K., Talbot, H., & Sato, Y. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05141-7_17
Self-citationIwasa, M., Takao, M., Otake, Y., Uemura, K., Mazen, S., Hamada, H., Sato, Y., Okada, S., & Sugano, N. (2025) · Journal of Japan Society of Computer Aided Surgery · doi:10.5759/jscas.27.5
Self-citationUncertainty-aware multi-class brain tumor segmentation using Bayesian U-Net variants
Pal, R., Kumar, S., & Bhatnagar, G. (2026) · Biomedical Physics & Engineering Express · doi:10.1088/2057-1976/ae5ca9
“In [15, 30] produce a pixel-level uncertainty map using variational inference.”
Quantitative whole-body MRI in FSHD: assessing systemic involvement and future directions
Friedman, S. D., Leung, D., Riem, L., DuCharme, O., Costanzo, K. E., Weatherley, E., McCrady, A., Faino, A., Wang, L., Shaw, D., et al. (2026) · Neuromuscular Disorders · doi:10.1016/j.nmd.2026.106459 · Cited by 1
FMD-AL: Cold-Start Active Learning based on Foundation Model
Ai, H., Zhou, Y., Peng, Z., Chen, Z., Guo, S., Xu, H., Yu, Z., & Li, J. (2026) · International Conference on Multimedia Retrieval · doi:10.1145/3805622.3810662
“Active Learning (AL) has emerged as a crucial paradigm to alleviate the annotation burden in medical image segmentation [5, 11, 38, 39].”
Kinoshita, T. (2026) · Journal of Japan Society of Computer Aided Surgery · doi:10.5759/jscas.28.12
Side-to-side differences in hip bone mineral density and associated morphological factors
Uemura, K., Otake, Y., Soufi, M., Kono, S., Higuchi, R., Asano, T., Mae, H., Tani, T., Takashima, K., Sato, Y., et al. (2026) · Osteoporosis International · doi:10.1007/s00198-026-08141-z
“Then, nine landmarks were identified in the proximal femur using another model: the head center, four points at the head-neck junction, the neck center, the tip of the lesser trochanter, and two points on the femoral shaft located 2 and 5 cm distal to the lesser trochanter [12, 13] (Fig.”
Self-citationBi, Q., Wang, X., Tan, J., Dai, H., Wu, Y., Song, Y., Ai, C., Zhang, H., & Wu, L. (2026) · Abdominal Radiology · doi:10.1007/s00261-026-05716-6
Murakami, Y., Nagatani, Y., Masaki, Y., Otake, Y., Sato, Y., Ishimoto, S., Wakamiya, M., & Watanabe, Y. (2026) · Research Square · doi:10.21203/rs.3.rs-10641800/v1
Self-citationGourine, S. A., Soufi, M., Otake, Y., Masaki, Y., Murakami, Y., Nagatani, Y., Watanabe, Y., Uemura, K., Kono, K., Takao, M., et al. (2026) · International Journal of Computer Assisted Radiology and Surgery · doi:10.1007/s11548-026-03785-y
Self-citationSchöller, J. H. J., Gu, Y., Uemura, K., Takao, M., Okada, S., Sugano, N., Sato, Y., & Otake, Y. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-38072-2_31
Self-citationMorita, J., Uemura, K., Kono, S., Gu, Y., Nishioka, Y., Asano, T., Higuchi, R., Mae, H., Iwasa, M., Tani, T., et al. (2026) · Journal of Bone and Mineral Metabolism · doi:10.1007/s00774-026-01766-4
Self-citationTerada, C., Yokoo, S., Kondo, J., Otake, Y., & Uemura, K. (2026) · European Journal of Orthopaedic Surgery & Traumatology · doi:10.1007/s00590-026-04970-0
“CT-based muscle quantification was performed using a previously developed deep-learning model [20, 21].”
Self-citationAdvances in 3D Slicer: Bridging AI, Cloud Computing, and Interventional Technologies
Soufi, M., Yoshinobu, S. A. T. O., Nobuhiko, H. A. T. A., Kikinis, Z., & Kikinis, R. (2026) · Advanced Biomedical Engineering · doi:10.14326/abe.250416
Self-citation
Metadata
- Authors
- Hiasa, Y., Otake, Y., Takao, M., Ogawa, T., Sugano, N., & Sato, Y.
- Affiliations
- Nara Institute of Science and Technology (JP)
- The University of Osaka (JP)
- Venue
- IEEE Transactions on Medical Imaging
- Published
- 2020-04
- Volume
- 39
- Issue
- 4
- Pages
- 1030-1040
- ISSN
- 0278-0062 (ISSN-L)
- 1558-254X
- OpenAlex
- W2972629588
- Cited by
- 157
- References
- 45
- Citations by year
- 2026: 16 · 2025: 27 · 2024: 26 · 2023: 24 · 2022: 26 · 2021: 22 · 2020: 16
- FWCI
- 7.16
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Transactions on Medical Imaging (publishedVersion)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Ministry of Education, Culture, Sports, Science and Technology (19H01176, 26108004)
- Japan Society for the Promotion of Science (20407)
- AMED/ETH Strategic Japanese-Swiss Cooperative Program
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Medical Image Segmentation Techniques
- Physical Sciences › Engineering › Biomedical Engineering › Medical Imaging and Analysis
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Keywords
- muscle segmentation, Monte Carlo dropout, uncertainty estimation, musculoskeletal modeling, active learning (machine learning), muscle atrophy, biomechanical simulation, multi-atlas segmentation
- MeSH terms
- Deep Learning, Bayes Theorem, Female, Hip, Humans, Image Processing, Computer-Assisted, Male, Thigh, Tomography, X-Ray Computed, Neural Networks, Computer, Muscle, Skeletal, Arthroplasty, Replacement, Hip, Patient-Specific Modeling
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.