Mapping Three-Dimensional Tumor Heterogeneity through Deep Learning Inference of Spatial Transcriptomics from Routine Histopathology: A Proof-of-Concept Comparative Study

Zarif Azher, Gokul Srinivasan, Keluo Yao, Minh-Khang Le, Ken Lau, Harsimran Kaur, Fred Kolling, Louis Vaickus, Xiaoying Lu, Joshua Levy
Proceedings of the 4th Machine Learning for Health Symposium, PMLR 259:73-85, 2025.

Abstract

Spatial transcriptomics (ST) technologies enable the mapping of gene and protein abundance within specific tissue architectures, representing a significant advancement over conventional bulk analyses that can obscure critical prognostic markers tied to spatial contexts. Expanding these analyses to three dimensions (3D) can further uncover intricate biomolecular phenomena that may be truncated or missed in two-dimensional (2D) studies. However, the widespread application of 3D ST profiling is limited by high costs and logistical challenges. Deep learning-based inference of ST data from routine histopathological staining offers a cost-effective alternative, allowing for the exploration of histologically associated biological pathways in 3D and enhancing our ability to detect structures linked to tumor progression. In this proof-of-concept study, we employed deep learning models to infer ST data from routine histopathology for 10 colorectal cancer patients, with 10 serial sections analyzed per patient. Our downstream analyses revealed several key instances where 3D approaches provided enhanced insights into cellular phenomena compared to traditional 2D methods. These findings lay the groundwork for future research aimed at leveraging these methods to investigate subtle 3D biomarkers associated with tumor metastasis and recurrence.

Cite this Paper

BibTeX
@InProceedings{pmlr-v259-azher25a, title = {Mapping Three-Dimensional Tumor Heterogeneity through Deep Learning Inference of Spatial Transcriptomics from Routine Histopathology: A Proof-of-Concept Comparative Study}, author = {Azher, Zarif and Srinivasan, Gokul and Yao, Keluo and Le, Minh-Khang and Lau, Ken and Kaur, Harsimran and Kolling, Fred and Vaickus, Louis and Lu, Xiaoying and Levy, Joshua}, booktitle = {Proceedings of the 4th Machine Learning for Health Symposium}, pages = {73--85}, year = {2025}, editor = {Hegselmann, Stefan and Zhou, Helen and Healey, Elizabeth and Chang, Trenton and Ellington, Caleb and Mhasawade, Vishwali and Tonekaboni, Sana and Argaw, Peniel and Zhang, Haoran}, volume = {259}, series = {Proceedings of Machine Learning Research}, month = {15--16 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v259/main/assets/azher25a/azher25a.pdf}, url = {https://proceedings.mlr.press/v259/azher25a.html}, abstract = {Spatial transcriptomics (ST) technologies enable the mapping of gene and protein abundance within specific tissue architectures, representing a significant advancement over conventional bulk analyses that can obscure critical prognostic markers tied to spatial contexts. Expanding these analyses to three dimensions (3D) can further uncover intricate biomolecular phenomena that may be truncated or missed in two-dimensional (2D) studies. However, the widespread application of 3D ST profiling is limited by high costs and logistical challenges. Deep learning-based inference of ST data from routine histopathological staining offers a cost-effective alternative, allowing for the exploration of histologically associated biological pathways in 3D and enhancing our ability to detect structures linked to tumor progression. In this proof-of-concept study, we employed deep learning models to infer ST data from routine histopathology for 10 colorectal cancer patients, with 10 serial sections analyzed per patient. Our downstream analyses revealed several key instances where 3D approaches provided enhanced insights into cellular phenomena compared to traditional 2D methods. These findings lay the groundwork for future research aimed at leveraging these methods to investigate subtle 3D biomarkers associated with tumor metastasis and recurrence.} }
Endnote
%0 Conference Paper %T Mapping Three-Dimensional Tumor Heterogeneity through Deep Learning Inference of Spatial Transcriptomics from Routine Histopathology: A Proof-of-Concept Comparative Study %A Zarif Azher %A Gokul Srinivasan %A Keluo Yao %A Minh-Khang Le %A Ken Lau %A Harsimran Kaur %A Fred Kolling %A Louis Vaickus %A Xiaoying Lu %A Joshua Levy %B Proceedings of the 4th Machine Learning for Health Symposium %C Proceedings of Machine Learning Research %D 2025 %E Stefan Hegselmann %E Helen Zhou %E Elizabeth Healey %E Trenton Chang %E Caleb Ellington %E Vishwali Mhasawade %E Sana Tonekaboni %E Peniel Argaw %E Haoran Zhang %F pmlr-v259-azher25a %I PMLR %P 73--85 %U https://proceedings.mlr.press/v259/azher25a.html %V 259 %X Spatial transcriptomics (ST) technologies enable the mapping of gene and protein abundance within specific tissue architectures, representing a significant advancement over conventional bulk analyses that can obscure critical prognostic markers tied to spatial contexts. Expanding these analyses to three dimensions (3D) can further uncover intricate biomolecular phenomena that may be truncated or missed in two-dimensional (2D) studies. However, the widespread application of 3D ST profiling is limited by high costs and logistical challenges. Deep learning-based inference of ST data from routine histopathological staining offers a cost-effective alternative, allowing for the exploration of histologically associated biological pathways in 3D and enhancing our ability to detect structures linked to tumor progression. In this proof-of-concept study, we employed deep learning models to infer ST data from routine histopathology for 10 colorectal cancer patients, with 10 serial sections analyzed per patient. Our downstream analyses revealed several key instances where 3D approaches provided enhanced insights into cellular phenomena compared to traditional 2D methods. These findings lay the groundwork for future research aimed at leveraging these methods to investigate subtle 3D biomarkers associated with tumor metastasis and recurrence.
APA
Azher, Z., Srinivasan, G., Yao, K., Le, M., Lau, K., Kaur, H., Kolling, F., Vaickus, L., Lu, X. & Levy, J.. (2025). Mapping Three-Dimensional Tumor Heterogeneity through Deep Learning Inference of Spatial Transcriptomics from Routine Histopathology: A Proof-of-Concept Comparative Study. Proceedings of the 4th Machine Learning for Health Symposium, in Proceedings of Machine Learning Research 259:73-85 Available from https://proceedings.mlr.press/v259/azher25a.html.

Related Material