Ruining Deng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Ruining Deng has a total of 153 co-authors. Here are the top co-authors by number of shared publications:
- Yuankai Huo
Vanderbilt University
20 shared publications - Haichun Yang
Vanderbilt University Medical Center 37232 Nashville TN USA
11 shared publications - Tianyuan Yao
Vanderbilt University, Department of Computer Science, Nashville, Tennessee, United States
10 shared publications - Can Cui
Vanderbilt University, Department of Computer Science, Nashville, Tennessee, United States
9 shared publications - Quan Liu
Vanderbilt University, Department of Computer Science, Nashville, Tennessee, United States
7 shared publications - Agnes B. Fogo
Department of Pathology, Microbiology and Immunology, Vanderbilt University Medical Center, Nashville, Tennessee
6 shared publications
Latest Publications
- 31 July 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 114, 104234
KPIs 2024 challenge: Advancing glomerular segmentation from patch- to slide-level
- 31 July 2026
- Published by Elsevier in Medical Image Analysis
- Vol. 114, 104234
KPIs 2024 challenge: Advancing glomerular segmentation from patch- to slide-level
- Open Access
- 24 November 2025
- Published by Springer Nature in Communications Medicine
- Vol. 5 (1) , 495
Evaluating cell AI foundation models in kidney pathology with human-in-the-loop enrichment
Background Large-scale artificial intelligence foundation models have emerged as promising tools for addressing healthcare challenges, including digital pathology. While many have been developed for complex tasks such as disease diagnosis and tissue quantification using extensive and diverse datasets, their readiness for seemingly simpler tasks, such as nuclei segmentation within a single organ (for example, the kidney), remains unclear. This study answers two questions: How good are current cell foundation models? and How can we improve them? Methods We curated a multi-center, multi-disease, and multi-species dataset sampled from 2542 kidney whole slide images. Three state-of-the-art cell foundation models—Cellpose, StarDist, and CellViT—were evaluated. To enhance performance, we developed a human-in-the-loop strategy that distilled multi-model predictions, improving data quality while reducing reliance on pixel-level annotation. Fine-tuning was performed using the enriched datasets, and segmentation performance was quantitatively assessed. Results Here we show that cell nuclei segmentation in kidney pathology still requires improvement with more organ-targeted foundation models. Among the evaluated models, CellViT achieves the highest baseline performance, with an F1 score of 0.78. Fine-tuning with enriched data improves all three models, with StarDist achieving the highest F1 score of 0.82. The combination of the foundation model-generated pseudo-labels and a subset of pathologist-corrected “hard” patches yields consistent performance gains across all models. Conclusions This study establishes a benchmark for the development and deployment of cell AI foundation models tailored to real-world data. The proposed framework, which leverages foundation models with reduced expert annotation, supports more efficient workflows in clinical pathology.
- 24 November 2025
- Published by Springer Nature in Communications Medicine
- Vol. 5 (1) , 495
Open AccessEvaluating cell AI foundation models in kidney pathology with human-in-the-loop enrichment
Background Large-scale artificial intelligence foundation models have emerged as promising tools for addressing healthcare challenges, including digital pathology. While many have been developed for complex tasks such as disease diagnosis and tissue quantification using extensive and diverse datasets, their readiness for seemingly simpler tasks, such as nuclei segmentation within a single organ (for example, the kidney), remains unclear. This study answers two questions: How good are current cell foundation models? and How can we improve them? Methods We curated a multi-center, multi-disease, and multi-species dataset sampled from 2542 kidney whole slide images. Three state-of-the-art cell foundation models—Cellpose, StarDist, and CellViT—were evaluated. To enhance performance, we developed a human-in-the-loop strategy that distilled multi-model predictions, improving data quality while reducing reliance on pixel-level annotation. Fine-tuning was performed using the enriched datasets, and segmentation performance was quantitatively assessed. Results Here we show that cell nuclei segmentation in kidney pathology still requires improvement with more organ-targeted foundation models. Among the evaluated models, CellViT achieves the highest baseline performance, with an F1 score of 0.78. Fine-tuning with enriched data improves all three models, with StarDist achieving the highest F1 score of 0.82. The combination of the foundation model-generated pseudo-labels and a subset of pathologist-corrected “hard” patches yields consistent performance gains across all models. Conclusions This study establishes a benchmark for the development and deployment of cell AI foundation models tailored to real-world data. The proposed framework, which leverages foundation models with reduced expert annotation, supports more efficient workflows in clinical pathology.
- Open Access
- 8 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 18 (19391404) , 26983-26999
DeepAndes: A Self-Supervised Vision Foundation Model for Multispectral Remote Sensing Imagery of the Andes
By mapping sites at large scales using remotely sensed data, archaeologists can generate unique insights into long-term demographic trends, inter-regional social networks, and human adaptations in the past. Remote sensing surveys complement field-based approaches, and their reach can be especially great when combined with deep learning and computer vision techniques. However, conventional supervised deep learning methods face challenges in annotating fine-grained archaeological features at scale. In addition, while recent vision foundation models have shown remarkable success in learning large-scale remote sensing data with minimal annotations, most off-the-shelf solutions are designed for RGB images rather than multi-spectral satellite imagery, such as the 8-band data used in our study. In this paper, we introduce DeepAndes, a transformer-based vision foundation model trained on three million multi-spectral satellite images, specifically tailored for Andean archaeology. DeepAndes incorporates a customized DINOv2 self-supervised learning algorithm optimized for 8-band multi-spectral imagery, marking the first foundation model designed explicitly for the Andes region. We evaluate its image understanding performance through imbalanced image classification, image instance retrieval, and pixel-level semantic segmentation tasks. Our experiments show that DeepAndes achieves superior F1 scores, mean average precision, and Dice scores in few-shot learning scenarios, significantly outperforming models trained from scratch or pre-trained on smaller datasets. This underscores the effectiveness of large-scale self-supervised pre-training in archaeological remote sensing. Codes will be available on https://github.com/geopacha/DeepAndes.
- 8 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
- Vol. 18 (19391404) , 26983-26999
Open AccessDeepAndes: A Self-Supervised Vision Foundation Model for Multispectral Remote Sensing Imagery of the Andes
By mapping sites at large scales using remotely sensed data, archaeologists can generate unique insights into long-term demographic trends, inter-regional social networks, and human adaptations in the past. Remote sensing surveys complement field-based approaches, and their reach can be especially great when combined with deep learning and computer vision techniques. However, conventional supervised deep learning methods face challenges in annotating fine-grained archaeological features at scale. In addition, while recent vision foundation models have shown remarkable success in learning large-scale remote sensing data with minimal annotations, most off-the-shelf solutions are designed for RGB images rather than multi-spectral satellite imagery, such as the 8-band data used in our study. In this paper, we introduce DeepAndes, a transformer-based vision foundation model trained on three million multi-spectral satellite images, specifically tailored for Andean archaeology. DeepAndes incorporates a customized DINOv2 self-supervised learning algorithm optimized for 8-band multi-spectral imagery, marking the first foundation model designed explicitly for the Andes region. We evaluate its image understanding performance through imbalanced image classification, image instance retrieval, and pixel-level semantic segmentation tasks. Our experiments show that DeepAndes achieves superior F1 scores, mean average precision, and Dice scores in few-shot learning scenarios, significantly outperforming models trained from scratch or pre-trained on smaller datasets. This underscores the effectiveness of large-scale self-supervised pre-training in archaeological remote sensing. Codes will be available on https://github.com/geopacha/DeepAndes.
- Open Access
- 1 October 2025
- Published by Machine Learning for Biomedical Imaging in Machine Learning for Biomedical Imaging
- Vol. 3 (October 20) , 545-558
Circle Representation for Medical Instance Object Segmentation
Recently, circle representation has been introduced for medical imaging, designed specifically to enhance the detection of instance objects that are spherically shaped (e.g., cells, glomeruli, and nuclei). Given its outstanding effectiveness in instance detection, it is compelling to consider the application of circle representation for segmenting instance medical objects. In this study, we introduce CircleSnake, a simple end-to-end segmentation approach that utilizes circle contour deformation for segmenting ball-shaped medical objects at the instance level. The innovation of CircleSnake lies in these three areas: (1) It substitutes the complex bounding box-to-octagon contour transformation with a more consistent and rotation-invariant bounding circle-to-circle contour adaptation. This adaptation specifically targets ball-shaped medical objects. (2) The circle representation employed in CircleSnake significantly reduces the degrees of freedom to two, compared to eight in the octagon representation. This reduction enhances both the robustness of the segmentation performance and the rotational consistency of the method. (3) CircleSnake is the first end-to-end deep instance segmentation pipeline to incorporate circle representation, encompassing consistent circle detection, circle contour proposal, and circular convolution in a unified framework. This integration is achieved through the novel application of circular convolution within the context of circle detection and instance segmentation. In practical applications, such as the detection of glomeruli, nuclei, and eosinophils in pathological images, CircleSnake has demonstrated superior performance and greater rotation invariance when compared to benchmarks. The code has been made publicly available at: https://github.com/hrlblab/CircleSnake.
- 1 October 2025
- Published by Machine Learning for Biomedical Imaging in Machine Learning for Biomedical Imaging
- Vol. 3 (October 20) , 545-558
Open AccessCircle Representation for Medical Instance Object Segmentation
Recently, circle representation has been introduced for medical imaging, designed specifically to enhance the detection of instance objects that are spherically shaped (e.g., cells, glomeruli, and nuclei). Given its outstanding effectiveness in instance detection, it is compelling to consider the application of circle representation for segmenting instance medical objects. In this study, we introduce CircleSnake, a simple end-to-end segmentation approach that utilizes circle contour deformation for segmenting ball-shaped medical objects at the instance level. The innovation of CircleSnake lies in these three areas: (1) It substitutes the complex bounding box-to-octagon contour transformation with a more consistent and rotation-invariant bounding circle-to-circle contour adaptation. This adaptation specifically targets ball-shaped medical objects. (2) The circle representation employed in CircleSnake significantly reduces the degrees of freedom to two, compared to eight in the octagon representation. This reduction enhances both the robustness of the segmentation performance and the rotational consistency of the method. (3) CircleSnake is the first end-to-end deep instance segmentation pipeline to incorporate circle representation, encompassing consistent circle detection, circle contour proposal, and circular convolution in a unified framework. This integration is achieved through the novel application of circular convolution within the context of circle detection and instance segmentation. In practical applications, such as the detection of glomeruli, nuclei, and eosinophils in pathological images, CircleSnake has demonstrated superior performance and greater rotation invariance when compared to benchmarks. The code has been made publicly available at: https://github.com/hrlblab/CircleSnake.
- Open Access
- 1 September 2025
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 12 (05) , 057501-12
Fine-grained multiclass nuclei segmentation with molecular empowered all-in-SAM model
The field of computational pathology1 is undergoing transformative advancements by integrating computational algorithms with whole slide imaging (WSI). This integration has shown promising results with improved diagnostic precision and advancing personalized medical treatments.2 Computational pathology focuses on analyzing digital pathological images to support clinical decisions and personalized therapies.3 However, accurately segmenting and classifying cell nuclei in these images remains challenging, especially in oncology, where it significantly impacts diagnostic and therapeutic planning.
- 1 September 2025
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 12 (05) , 057501-12
Open AccessFine-grained multiclass nuclei segmentation with molecular empowered all-in-SAM model
The field of computational pathology1 is undergoing transformative advancements by integrating computational algorithms with whole slide imaging (WSI). This integration has shown promising results with improved diagnostic precision and advancing personalized medical treatments.2 Computational pathology focuses on analyzing digital pathological images to support clinical decisions and personalized therapies.3 However, accurately segmenting and classifying cell nuclei in these images remains challenging, especially in oncology, where it significantly impacts diagnostic and therapeutic planning.
- Open Access
- 28 July 2025
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 12 (06) , 061406-15
Glo-In-One-v2: holistic identification of glomerular cells, tissues, and lesions in human and mouse histopathology
Whole-slide imaging (WSI) provides high-resolution views of tissue, significantly advancing quantitative analysis in nephropathology—particularly in glomerular evaluation.1 Within renal pathology, a field known for its complexity in image interpretation, glomeruli are essential functional units for clinical assessment.2 To automate glomerular detection and segmentation, our team previously developed the Glo-In-One toolkit,3 based on convolutional neural networks (CNNs). Although effective, the original Glo-In-One was limited to whole-tuft segmentation and lacked the ability to capture sub-glomerular structures.
- 28 July 2025
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 12 (06) , 061406-15
Open AccessGlo-In-One-v2: holistic identification of glomerular cells, tissues, and lesions in human and mouse histopathology
Whole-slide imaging (WSI) provides high-resolution views of tissue, significantly advancing quantitative analysis in nephropathology—particularly in glomerular evaluation.1 Within renal pathology, a field known for its complexity in image interpretation, glomeruli are essential functional units for clinical assessment.2 To automate glomerular detection and segmentation, our team previously developed the Glo-In-One toolkit,3 based on convolutional neural networks (CNNs). Although effective, the original Glo-In-One was limited to whole-tuft segmentation and lacked the ability to capture sub-glomerular structures.
- Open Access
- 1 July 2025
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 12 (04) , 1-14
ZeroReg3D: a zero-shot registration pipeline for 3D consecutive histopathology image reconstruction
PurposeHistological analysis plays a crucial role in understanding tissue structure and pathology. Although recent advancements in registration methods have improved 2D histological analysis, they often struggle to preserve critical 3D spatial relationships, limiting their utility in both clinical and research applications. Specifically, constructing accurate 3D models from 2D slices remains challenging due to tissue deformation, sectioning artifacts, variability in imaging techniques, and inconsistent illumination. Deep learning-based registration methods have demonstrated improved performance but suffer from limited generalizability and require large-scale training data. In contrast, non-deep-learning approaches offer better generalizability but often compromise on accuracy.ApproachWe introduce ZeroReg3D, a zero-shot registration pipeline that integrates zero-shot deep learning-based keypoint matching and non-deep-learning registration techniques to effectively mitigate deformation and sectioning artifacts without requiring extensive training data.ResultsComprehensive evaluations demonstrate that our pairwise 2D image registration method improves registration accuracy by ∼10% over baseline methods, outperforming existing strategies in both accuracy and robustness. High-fidelity 3D reconstructions further validate the effectiveness of our approach, establishing ZeroReg3D as a reliable framework for precise 3D reconstruction from consecutive 2D histological images.ConclusionsWe introduced ZeroReg3D, a zero-shot registration pipeline tailored for accurate 3D reconstruction from serial histological sections. By combining zero-shot deep learning-based keypoint matching with optimization-based affine and non-rigid registration techniques, ZeroReg3D effectively addresses critical challenges such as tissue deformation, sectioning artifacts, staining variability, and inconsistent illumination without requiring retraining or fine-tuning.
- 1 July 2025
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 12 (04) , 1-14
Open AccessZeroReg3D: a zero-shot registration pipeline for 3D consecutive histopathology image reconstruction
PurposeHistological analysis plays a crucial role in understanding tissue structure and pathology. Although recent advancements in registration methods have improved 2D histological analysis, they often struggle to preserve critical 3D spatial relationships, limiting their utility in both clinical and research applications. Specifically, constructing accurate 3D models from 2D slices remains challenging due to tissue deformation, sectioning artifacts, variability in imaging techniques, and inconsistent illumination. Deep learning-based registration methods have demonstrated improved performance but suffer from limited generalizability and require large-scale training data. In contrast, non-deep-learning approaches offer better generalizability but often compromise on accuracy.ApproachWe introduce ZeroReg3D, a zero-shot registration pipeline that integrates zero-shot deep learning-based keypoint matching and non-deep-learning registration techniques to effectively mitigate deformation and sectioning artifacts without requiring extensive training data.ResultsComprehensive evaluations demonstrate that our pairwise 2D image registration method improves registration accuracy by ∼10% over baseline methods, outperforming existing strategies in both accuracy and robustness. High-fidelity 3D reconstructions further validate the effectiveness of our approach, establishing ZeroReg3D as a reliable framework for precise 3D reconstruction from consecutive 2D histological images.ConclusionsWe introduced ZeroReg3D, a zero-shot registration pipeline tailored for accurate 3D reconstruction from serial histological sections. By combining zero-shot deep learning-based keypoint matching with optimization-based affine and non-rigid registration techniques, ZeroReg3D effectively addresses critical challenges such as tissue deformation, sectioning artifacts, staining variability, and inconsistent illumination without requiring retraining or fine-tuning.
- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Pulse
- Vol. 16 (2) , 45-48
Why Not for Society?
- 9 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Pulse
- Vol. 16 (2) , 45-48
Why Not for Society?
- Open Access
- 1 November 2024
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 11 (06) , 1-17
Data-driven nucleus subclassification on colon hematoxylin and eosin using style-transferred digital pathology
PurposeCells are building blocks for human physiology; consequently, understanding the way cells communicate, co-locate, and interrelate is essential to furthering our understanding of how the body functions in both health and disease. Hematoxylin and eosin (H&E) is the standard stain used in histological analysis of tissues in both clinical and research settings. Although H&E is ubiquitous and reveals tissue microanatomy, the classification and mapping of cell subtypes often require the use of specialized stains. The recent CoNIC Challenge focused on artificial intelligence classification of six types of cells on colon H&E but was unable to classify epithelial subtypes (progenitor, enteroendocrine, goblet), lymphocyte subtypes (B, helper T, cytotoxic T), and connective subtypes (fibroblasts). We propose to use inter-modality learning to label previously un-labelable cell types on H&E.ApproachWe took advantage of the cell classification information inherent in multiplexed immunofluorescence (MxIF) histology to create cell-level annotations for 14 subclasses. Then, we performed style transfer on the MxIF to synthesize realistic virtual H&E. We assessed the efficacy of a supervised learning scheme using the virtual H&E and 14 subclass labels. We evaluated our model on virtual H&E and real H&E.ResultsOn virtual H&E, we were able to classify helper T cells and epithelial progenitors with positive predictive values of 0.34±0.15 (prevalence 0.03±0.01) and 0.47±0.1 (prevalence 0.07±0.02), respectively, when using ground truth centroid information. On real H&E, we needed to compute bounded metrics instead of direct metrics because our fine-grained virtual H&E predicted classes had to be matched to the closest available parent classes in the coarser labels from the real H&E dataset. For the real H&E, we could classify bounded metrics for the helper T cells and epithelial progenitors with upper bound positive predictive values of 0.43±0.03 (parent class prevalence 0.21) and 0.94±0.02 (parent class prevalence 0.49) when using ground truth centroid information.ConclusionsThis is the first work to provide cell type classification for helper T and epithelial progenitor nuclei on H&E.
- 1 November 2024
- Published by SPIE-Intl Soc Optical Eng in Journal of Medical Imaging
- Vol. 11 (06) , 1-17
Open AccessData-driven nucleus subclassification on colon hematoxylin and eosin using style-transferred digital pathology
PurposeCells are building blocks for human physiology; consequently, understanding the way cells communicate, co-locate, and interrelate is essential to furthering our understanding of how the body functions in both health and disease. Hematoxylin and eosin (H&E) is the standard stain used in histological analysis of tissues in both clinical and research settings. Although H&E is ubiquitous and reveals tissue microanatomy, the classification and mapping of cell subtypes often require the use of specialized stains. The recent CoNIC Challenge focused on artificial intelligence classification of six types of cells on colon H&E but was unable to classify epithelial subtypes (progenitor, enteroendocrine, goblet), lymphocyte subtypes (B, helper T, cytotoxic T), and connective subtypes (fibroblasts). We propose to use inter-modality learning to label previously un-labelable cell types on H&E.ApproachWe took advantage of the cell classification information inherent in multiplexed immunofluorescence (MxIF) histology to create cell-level annotations for 14 subclasses. Then, we performed style transfer on the MxIF to synthesize realistic virtual H&E. We assessed the efficacy of a supervised learning scheme using the virtual H&E and 14 subclass labels. We evaluated our model on virtual H&E and real H&E.ResultsOn virtual H&E, we were able to classify helper T cells and epithelial progenitors with positive predictive values of 0.34±0.15 (prevalence 0.03±0.01) and 0.47±0.1 (prevalence 0.07±0.02), respectively, when using ground truth centroid information. On real H&E, we needed to compute bounded metrics instead of direct metrics because our fine-grained virtual H&E predicted classes had to be matched to the closest available parent classes in the coarser labels from the real H&E dataset. For the real H&E, we could classify bounded metrics for the helper T cells and epithelial progenitors with upper bound positive predictive values of 0.43±0.03 (parent class prevalence 0.21) and 0.94±0.02 (parent class prevalence 0.49) when using ground truth centroid information.ConclusionsThis is the first work to provide cell type classification for helper T and epithelial progenitor nuclei on H&E.