Ali Bashashati
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Ali Bashashati has a total of 247 co-authors. Here are the top co-authors by number of shared publications:
- Hossein Farahani
University of British Columbia
16 shared publications Hossein Farahani
University of British Columbia
10 shared publicationsAmirali Darbandsari
Department of Electrical and Computer Engineering, University of British Columbia, BC, Vancouver, Canada
9 shared publications- Steven J. M. Jones
Canada’s Michael Smith Genome Sciences Centre
8 shared publications Allen Zhang
Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
6 shared publicationsDavid Farnell
Department of Pathology and Laboratory Medicine, University of British Columbia, Vancouver, Canada
6 shared publications
Latest Publications
- 24 September 2026
- Published by Springer Nature
- p. 354-364
Knowledge-Enhanced Representation Learning with Retrieval-Augmented Multimodal Fusion for Survival Prediction
- 24 September 2026
- Published by Springer Nature
- p. 354-364
Knowledge-Enhanced Representation Learning with Retrieval-Augmented Multimodal Fusion for Survival Prediction
- 18 September 2026
- Published by Springer Nature
- p. 175-193
SurvMILKD: A Weakly Supervised Survival Analysis Framework for Multi-teacher Knowledge Distillation Using Pathology Foundation Models
- 18 September 2026
- Published by Springer Nature
- p. 175-193
SurvMILKD: A Weakly Supervised Survival Analysis Framework for Multi-teacher Knowledge Distillation Using Pathology Foundation Models
- 12 September 2026
- Published by Springer Nature
- p. 526-543
MedGSSR: Generalizable Medical Image Super-Resolution 3D Reconstruction via Hierarchical Feed-Forward Gaussian Splatting
- 12 September 2026
- Published by Springer Nature
- p. 526-543
MedGSSR: Generalizable Medical Image Super-Resolution 3D Reconstruction via Hierarchical Feed-Forward Gaussian Splatting
- 1 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 30 (3) , 2203-2216
KAFSTExp: Kernel Adaptive Filtering With Nyström Approximation for Predicting Spatial Gene Expression From Histology Images
Spatial transcriptomics (ST), known as an expensive medical examination, plays an important role in analyzing the spatial heterogeneity of tumors. When considering the correlation between tissue morphological patterns and gene profiles, predicting corresponding gene expression from pathology image obtained from affordable biopsies is regarded as an instantaneous and cost-effective alternative. However, accurately modeling the complex and nonlinear relationship between histological features and gene expression remains challenging. Existing deep learning models often struggle to generalize on limited ST datasets due to their large and overparameterized architectures. The primary advantage of kernel adaptive filtering (KAF) lies in its ability to transform a challenging nonlinear problem arising in the original space into a linear regression problem in the higher-dimensional feature space via kernel methods. Therefore, this paper proposes a framework called KAFSTExp, which utilizes the state-ofthe-art pathology foundation model UNI to encode image feature vectors, and then introduces the kernel least mean square algorithm with Nystrom approximation to predict the ¨ normalized transcript counts of specific genes. Extensive experiments show that KAFSTExp significantly improves prediction accuracy while reducing computational cost and training time. KAFSTExp demonstrates consistent performance gains across multiple ST datasets, achieving relative improvements in PCC ranging from 1.24% to 94.23%, with an average increase of 19.80% over the best-performing non-KAF methods. External validation and further clinical analysis confirm the generalization performance and clinical application value of the proposed KAFSTExp.
- 1 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 30 (3) , 2203-2216
KAFSTExp: Kernel Adaptive Filtering With Nyström Approximation for Predicting Spatial Gene Expression From Histology Images
Spatial transcriptomics (ST), known as an expensive medical examination, plays an important role in analyzing the spatial heterogeneity of tumors. When considering the correlation between tissue morphological patterns and gene profiles, predicting corresponding gene expression from pathology image obtained from affordable biopsies is regarded as an instantaneous and cost-effective alternative. However, accurately modeling the complex and nonlinear relationship between histological features and gene expression remains challenging. Existing deep learning models often struggle to generalize on limited ST datasets due to their large and overparameterized architectures. The primary advantage of kernel adaptive filtering (KAF) lies in its ability to transform a challenging nonlinear problem arising in the original space into a linear regression problem in the higher-dimensional feature space via kernel methods. Therefore, this paper proposes a framework called KAFSTExp, which utilizes the state-ofthe-art pathology foundation model UNI to encode image feature vectors, and then introduces the kernel least mean square algorithm with Nystrom approximation to predict the ¨ normalized transcript counts of specific genes. Extensive experiments show that KAFSTExp significantly improves prediction accuracy while reducing computational cost and training time. KAFSTExp demonstrates consistent performance gains across multiple ST datasets, achieving relative improvements in PCC ranging from 1.24% to 94.23%, with an average increase of 19.80% over the best-performing non-KAF methods. External validation and further clinical analysis confirm the generalization performance and clinical application value of the proposed KAFSTExp.
- Open Access
- 24 February 2026
- Published by Springer Nature in Journal of Clinical Monitoring and Computing
- Vol. 40 (4) , 1-23
Regional tissue perfusion index (RTPI): a new optical-based metric for quantifying regional tissue perfusion
Accurate, continuous assessment of regional tissue perfusion remains a significant clinical challenge, as most existing modalities are invasive, indirect, or impractical for routine monitoring. Near-infrared spectroscopy (NIRS) has been widely adopted to assess tissue oxygenation; however, conventional NIRS-derived indices are insufficient surrogates for true perfusion and often fail to capture rapid hemodynamic changes. This study aimed to introduce and validate the Regional Tissue Perfusion Index (RTPI), a novel NIRS-derived metric that integrates multiple features of the NIRS signal to provide continuous, non-invasive, and physiologically relevant assessment of tissue perfusion. RTPI was developed using principal component analysis (PCA) of multiple NIRS-derived parameters, including pulse amplitude ratio, derivative, and area under the curve. Its performance was evaluated in healthy volunteers during controlled ischemia–reperfusion protocols and compared with established reference standards, including laser Doppler flowmetry (LDF) and photoplethysmography (PPG). RTPI showed acceptable correlations with LDF Flux and PPG perfusion index (PI) during dynamic perfusion changes. Unlike conventional NIRS-derived oxygenation and hemodynamic indices, which often exhibited delayed responses, RTPI demonstrated immediate and significant sensitivity to both complete and partial ischemia–reperfusion episodes in the analyzed datasets. Intraclass correlation indicated stronger session-to-session reliability for RTPI than for both Flux and PI. RTPI represents a multiparametric, physiologically meaningful, and computationally efficient metric for real-time monitoring of tissue perfusion. Its ability to detect perfusion compromise independently of oxygenation indices highlights its translational potential for bedside implementation in critical care, trauma, perioperative, and vascular medicine, where improved diagnostic accuracy could significantly impact patient outcomes.
- 24 February 2026
- Published by Springer Nature in Journal of Clinical Monitoring and Computing
- Vol. 40 (4) , 1-23
Open AccessRegional tissue perfusion index (RTPI): a new optical-based metric for quantifying regional tissue perfusion
Accurate, continuous assessment of regional tissue perfusion remains a significant clinical challenge, as most existing modalities are invasive, indirect, or impractical for routine monitoring. Near-infrared spectroscopy (NIRS) has been widely adopted to assess tissue oxygenation; however, conventional NIRS-derived indices are insufficient surrogates for true perfusion and often fail to capture rapid hemodynamic changes. This study aimed to introduce and validate the Regional Tissue Perfusion Index (RTPI), a novel NIRS-derived metric that integrates multiple features of the NIRS signal to provide continuous, non-invasive, and physiologically relevant assessment of tissue perfusion. RTPI was developed using principal component analysis (PCA) of multiple NIRS-derived parameters, including pulse amplitude ratio, derivative, and area under the curve. Its performance was evaluated in healthy volunteers during controlled ischemia–reperfusion protocols and compared with established reference standards, including laser Doppler flowmetry (LDF) and photoplethysmography (PPG). RTPI showed acceptable correlations with LDF Flux and PPG perfusion index (PI) during dynamic perfusion changes. Unlike conventional NIRS-derived oxygenation and hemodynamic indices, which often exhibited delayed responses, RTPI demonstrated immediate and significant sensitivity to both complete and partial ischemia–reperfusion episodes in the analyzed datasets. Intraclass correlation indicated stronger session-to-session reliability for RTPI than for both Flux and PI. RTPI represents a multiparametric, physiologically meaningful, and computationally efficient metric for real-time monitoring of tissue perfusion. Its ability to detect perfusion compromise independently of oxygenation indices highlights its translational potential for bedside implementation in critical care, trauma, perioperative, and vascular medicine, where improved diagnostic accuracy could significantly impact patient outcomes.
- Open Access
A Deep Learning Framework for Classification of Neuroendocrine Neoplasm Whole Slide Images
Background/Objectives: Neuroendocrine neoplasms (NENs) are uncommon neoplasms. Grading informs the prognosis and treatment decision of NENs and is determined by cell proliferation, which is measured by mitotic count and Ki-67 index. These measurements present challenges for pathologists as they suffer inter- and intra-observer variability and are cumbersome to quantify. To address these challenges, we developed a machine learning pipeline for identifying tumor areas, proliferating cells, and grading NENs. Methods: Our study includes 385 samples of gastroenteropancreatic NENs from across British Columbia with two stains (247 H&E and 138 Ki-67 images). Labels for these cases are at the patient-level, and there are 186 patients. We systematically investigated three settings for our study: H&E, H&E with Ki-67, and pathologist-reviewed and corrected cases. Results: Our H&E framework achieved a three-fold balanced accuracy of 77.5% in NEN grading. The H&E with Ki-67 framework yields a performance improvement to 83.0% on grading. We provide survival and multivariate analysis with a c-index of 0.65. Grade 1 NENs misclassified by the model were reviewed by a pathologist to assess reasons. Analysis of our AI-graded NENs for the subset of pathologist-assessed G1s demonstrated a significant (p-value = 0.007) survival difference amongst samples the algorithm assigned to a higher grade (n = 20; median survival 4.22 years) compared to concordant G1 samples (n = 60; median survival 10.13 years). Conclusions: Our model identifies NEN grades with high accuracy and identified some grade 1 tumors as prognostically unique, suggesting potential improvements to standard grading. Further studies are needed to determine if this discordant group is a different clinical entity.
Open AccessA Deep Learning Framework for Classification of Neuroendocrine Neoplasm Whole Slide Images
Background/Objectives: Neuroendocrine neoplasms (NENs) are uncommon neoplasms. Grading informs the prognosis and treatment decision of NENs and is determined by cell proliferation, which is measured by mitotic count and Ki-67 index. These measurements present challenges for pathologists as they suffer inter- and intra-observer variability and are cumbersome to quantify. To address these challenges, we developed a machine learning pipeline for identifying tumor areas, proliferating cells, and grading NENs. Methods: Our study includes 385 samples of gastroenteropancreatic NENs from across British Columbia with two stains (247 H&E and 138 Ki-67 images). Labels for these cases are at the patient-level, and there are 186 patients. We systematically investigated three settings for our study: H&E, H&E with Ki-67, and pathologist-reviewed and corrected cases. Results: Our H&E framework achieved a three-fold balanced accuracy of 77.5% in NEN grading. The H&E with Ki-67 framework yields a performance improvement to 83.0% on grading. We provide survival and multivariate analysis with a c-index of 0.65. Grade 1 NENs misclassified by the model were reviewed by a pathologist to assess reasons. Analysis of our AI-graded NENs for the subset of pathologist-assessed G1s demonstrated a significant (p-value = 0.007) survival difference amongst samples the algorithm assigned to a higher grade (n = 20; median survival 4.22 years) compared to concordant G1 samples (n = 60; median survival 10.13 years). Conclusions: Our model identifies NEN grades with high accuracy and identified some grade 1 tumors as prognostically unique, suggesting potential improvements to standard grading. Further studies are needed to determine if this discordant group is a different clinical entity.
- Open Access
- 17 February 2025
- Published by Springer Nature in Discover Oncology
- Vol. 16 (1) , 1-15
Benchmarking histopathology foundation models for ovarian cancer bevacizumab treatment response prediction from whole slide images
Purpose Bevacizumab is a widely studied targeted therapeutic drug used in conjunction with standard chemotherapy for the treatment of recurrent ovarian cancer. While its administration has been shown to increase progression-free survival (PFS) in patients with advanced-stage ovarian cancer, the lack of identifiable biomarkers for predicting patient response has been a major roadblock in its effective adoption towards personalized medicine. Methods In this work, we leverage the latest histopathology foundation models trained on large-scale whole slide image (WSI) datasets to extract ovarian tumor tissue features for predicting bevacizumab response from WSIs. Results Our extensive experiments across a combination of different histopathology foundation models and multiple instance learning (MIL) strategies demonstrate the capability of these large models in predicting bevacizumab response in ovarian cancer patients with the best models achieving a patient-level balanced accuracy score close to 70%. Furthermore, these models can effectively stratify high- and low-risk patients (p < 0.05) during the first year of bevacizumab treatment. Conclusion This work highlights the utility of histopathology foundation models to predict response to bevacizumab treatment from WSIs. The high-attention regions of the WSIs highlighted by these models not only aid the model explainability but also serve as promising imaging biomarkers for treatment prognosis.
- 17 February 2025
- Published by Springer Nature in Discover Oncology
- Vol. 16 (1) , 1-15
Open AccessBenchmarking histopathology foundation models for ovarian cancer bevacizumab treatment response prediction from whole slide images
Purpose Bevacizumab is a widely studied targeted therapeutic drug used in conjunction with standard chemotherapy for the treatment of recurrent ovarian cancer. While its administration has been shown to increase progression-free survival (PFS) in patients with advanced-stage ovarian cancer, the lack of identifiable biomarkers for predicting patient response has been a major roadblock in its effective adoption towards personalized medicine. Methods In this work, we leverage the latest histopathology foundation models trained on large-scale whole slide image (WSI) datasets to extract ovarian tumor tissue features for predicting bevacizumab response from WSIs. Results Our extensive experiments across a combination of different histopathology foundation models and multiple instance learning (MIL) strategies demonstrate the capability of these large models in predicting bevacizumab response in ovarian cancer patients with the best models achieving a patient-level balanced accuracy score close to 70%. Furthermore, these models can effectively stratify high- and low-risk patients (p < 0.05) during the first year of bevacizumab treatment. Conclusion This work highlights the utility of histopathology foundation models to predict response to bevacizumab treatment from WSIs. The high-attention regions of the WSIs highlighted by these models not only aid the model explainability but also serve as promising imaging biomarkers for treatment prognosis.
- Open Access
- 31 August 2024
- Published by Elsevier in The American Journal of Pathology
- Vol. 194 (12) , 2302-2312
A Deep Learning Approach for the Identification of the Molecular Subtypes of Pancreatic Ductal Adenocarcinoma Based on Whole Slide Pathology Images
- 31 August 2024
- Published by Elsevier in The American Journal of Pathology
- Vol. 194 (12) , 2302-2312
Open AccessA Deep Learning Approach for the Identification of the Molecular Subtypes of Pancreatic Ductal Adenocarcinoma Based on Whole Slide Pathology Images
- Open Access
- 15 August 2024
- Published by Cold Spring Harbor Laboratory in Genome Research
- Vol. 34 (8) , 1196-1210
Benchmarking bulk and single-cell variant-calling approaches on Chromium scRNA-seq and scATAC-seq libraries
Single-cell sequencing methodologies such as scRNA-seq and scATAC-seq have become widespread and effective tools to interrogate tissue composition. Increasingly, variant callers are being applied to these methodologies to resolve the genetic heterogeneity of a sample, especially in the case of detecting the clonal architecture of a tumor. Typically, traditional bulk DNA variant callers are applied to the pooled reads of a single-cell library to detect candidate mutations. Recently, multiple studies have applied such callers on reads from individual cells, with some citing the ability to detect rare variants with higher sensitivity. Many studies apply these two approaches to the Chromium (10x Genomics) scRNA-seq and scATAC-seq methodologies. However, Chromium-based libraries may offer additional challenges to variant calling compared with existing single-cell methodologies, raising questions regarding the validity of variants obtained from such a workflow. To determine the merits and challenges of various variant-calling approaches on Chromium scRNA-seq and scATAC-seq libraries, we use sample libraries with matched bulk whole-genome sequencing to evaluate the performance of callers. We review caller performance, finding that bulk callers applied on pooled reads significantly outperform individual-cell approaches. We also evaluate variants unique to scRNA-seq and scATAC-seq methodologies, finding patterns of noise but also potential capture of RNA-editing events. Finally, we review the notion that variant calling at the single-cell level can detect rare somatic variants, providing empirical results that suggest resolving such variants is infeasible in single-cell Chromium libraries.
- 15 August 2024
- Published by Cold Spring Harbor Laboratory in Genome Research
- Vol. 34 (8) , 1196-1210
Open AccessBenchmarking bulk and single-cell variant-calling approaches on Chromium scRNA-seq and scATAC-seq libraries
Single-cell sequencing methodologies such as scRNA-seq and scATAC-seq have become widespread and effective tools to interrogate tissue composition. Increasingly, variant callers are being applied to these methodologies to resolve the genetic heterogeneity of a sample, especially in the case of detecting the clonal architecture of a tumor. Typically, traditional bulk DNA variant callers are applied to the pooled reads of a single-cell library to detect candidate mutations. Recently, multiple studies have applied such callers on reads from individual cells, with some citing the ability to detect rare variants with higher sensitivity. Many studies apply these two approaches to the Chromium (10x Genomics) scRNA-seq and scATAC-seq methodologies. However, Chromium-based libraries may offer additional challenges to variant calling compared with existing single-cell methodologies, raising questions regarding the validity of variants obtained from such a workflow. To determine the merits and challenges of various variant-calling approaches on Chromium scRNA-seq and scATAC-seq libraries, we use sample libraries with matched bulk whole-genome sequencing to evaluate the performance of callers. We review caller performance, finding that bulk callers applied on pooled reads significantly outperform individual-cell approaches. We also evaluate variants unique to scRNA-seq and scATAC-seq methodologies, finding patterns of noise but also potential capture of RNA-editing events. Finally, we review the notion that variant calling at the single-cell level can detect rare somatic variants, providing empirical results that suggest resolving such variants is infeasible in single-cell Chromium libraries.
- Open Access
Benchmarking Histopathology Foundation Models for Ovarian Cancer Bevacizumab Treatment Response Prediction from Whole Slide Images
Bevacizumab is a widely studied targeted therapeutic drug used in conjunction with standard chemotherapy for the treatment of recurrent ovarian cancer. While its administration has shown to increase the progression-free survival (PFS) in patients with advanced stage ovarian cancer, the lack of identifiable biomarkers for predicting patient response has been a major roadblock in its effective adoption towards personalized medicine. In this work, we leverage the latest histopathology foundation models trained on large-scale whole slide image (WSI) datasets to extract ovarian tumor tissue features for predicting bevacizumab response from WSIs. Our extensive experiments across a combination of different histopathology foundation models and multiple instance learning (MIL) strategies demonstrate capability of these large models in predicting bevacizumab response in ovarian cancer patients with the best models achieving an AUC score of 0.86 and an accuracy score of 72.5%. Furthermore, our survival models are able to stratify high- and low-risk cases with statistical significance (p < 0.05) even among the patients with the aggressive subtype of high-grade serous ovarian carcinoma. This work highlights the utility of histopathology foundation models for the task of ovarian bevacizumab response prediction from WSIs. The high-attention regions of the WSIs highlighted by these models not only aid the model explainability but also serve as promising imaging biomarkers for treatment prognosis.
Open AccessBenchmarking Histopathology Foundation Models for Ovarian Cancer Bevacizumab Treatment Response Prediction from Whole Slide Images
Bevacizumab is a widely studied targeted therapeutic drug used in conjunction with standard chemotherapy for the treatment of recurrent ovarian cancer. While its administration has shown to increase the progression-free survival (PFS) in patients with advanced stage ovarian cancer, the lack of identifiable biomarkers for predicting patient response has been a major roadblock in its effective adoption towards personalized medicine. In this work, we leverage the latest histopathology foundation models trained on large-scale whole slide image (WSI) datasets to extract ovarian tumor tissue features for predicting bevacizumab response from WSIs. Our extensive experiments across a combination of different histopathology foundation models and multiple instance learning (MIL) strategies demonstrate capability of these large models in predicting bevacizumab response in ovarian cancer patients with the best models achieving an AUC score of 0.86 and an accuracy score of 72.5%. Furthermore, our survival models are able to stratify high- and low-risk cases with statistical significance (p < 0.05) even among the patients with the aggressive subtype of high-grade serous ovarian carcinoma. This work highlights the utility of histopathology foundation models for the task of ovarian bevacizumab response prediction from WSIs. The high-attention regions of the WSIs highlighted by these models not only aid the model explainability but also serve as promising imaging biomarkers for treatment prognosis.