Meng Zhou (Simon)

Meng Zhou (Simon)

ML Engineer/Scientist, MSCS @ University of Toronto & SickKids Research Institute

Publications

ClinicalFMamba: Advancing Clinical Assessment using Mamba-based Multimodal Neuroimaging Fusion

International Conference on Machine Learning in Medical Imaging, held in conjunction with MICCAI, 2025

In this work, we propose ClinicalFMamba, a novel end-to-end CNN-Mamba hybrid architecture that synergistically combines local and global feature modeling for 2D and 3D images. We further design a tri-plane scanning strategy for effectively learning volumetric dependencies in 3D images. Comprehensive evaluations on three datasets demonstrate the superior fusion performance across multiple quantitative metrics while achieving real-time fusion. We further validate the clinical utility of our approach on downstream 2D/3D brain tumor classification tasks, achieving superior performance over baseline methods. Our method establishes a new paradigm for efficient multimodal medical image fusion suitable for real-time clinical deployment.

Download here

ClinicalFMamba: Mamba-based Multimodal Medical Image Fusion for Enhanced Clinical Diagnosis

International Conference on Machine Learning NewInML Workshop, 2025

In this work, we propose ClinicalFMamba, a novel end-to-end CNN-Mamba hybrid architecture that synergistically combines local and global feature modeling. Our approach introduces: Dilated Gated Convolution Blocks for hierarchical multiscale feature extraction, and a latent Mamba module that efficiently captures long-range spatial dependencies between feature regions and enabling cross-modal fusion in latent space. Comprehensive evaluations on three datasets demonstrate the superior fusion performance across multiple quantitative metrics while achieving real-time fusion. Notably, we validate the clinical utility of our approach on the downstream brain tumor classification, achieving up to 7% improvements on the AUC score. Our method establishes a new paradigm for efficient multimodal medical image fusion suitable for real-time clinical deployment.

Download here

Edge-Enhanced Dilated Residual Attention Network for Multimodal Medical Image Fusion

IEEE International Conference on Bioinformatics and Biomedicine, 2024

In this paper, we propose a novel CNN-based architecture with a Dilated Residual Attention Network Module for effective multiscale feature extraction, coupled with a gradient operator to enhance edge detail learning. To ensure fast and efficient fusion, we present a parameter-free fusion strategy based on the weighted nuclear norm of softmax, which requires no additional computations during training or inference. Extensive experiments on MRI-CT, MRI-SPECT, and a downstream brain tumor classification task demonstrate that our approach outperforms various baseline methods in terms of visual quality, texture preservation, and fusion speed, making it a possible practical solution for real-world clinical applications.

Download here

Generating 3D Brain Tumor Regions in MRI using Vector-Quantization Generative Adversarial Networks

Computers in Biology and Medicine, 2024

In this work, we present a novel framework that uses vector-quantization GAN and a transformer incorporating masked token modeling to generate high-resolution and diverse 3D brain tumor ROIs that can be directly used as augmented data for the classification of brain tumor ROI. We apply our method to two imbalanced datasets where we augment the minority class: (1) the Multimodal Brain Tumor Segmentation Challenge (BraTS) 2019 dataset to generate new low-grade glioma (LGG) ROIs to balance with high-grade glioma (HGG) class; (2) the internal pediatric LGG (pLGG) dataset tumor ROIs with BRAF V600E Mutation genetic marker to balance with BRAF Fusion genetic marker class. We show that the proposed method outperforms various baseline models in both qualitative and quantitative measurements.

Download here

Conditional Generation of 3D Brain Tumor Regions via VQGAN and Temporal-Agnostic Masked Transformer

International Conference on Medical Imaging with Deep Learning, 2024

In this paper, we present a class-conditioned ROI generation framework that combines a conditional vector-quantization GAN and a class-conditioned masked Transformer to generate high-resolution and diverse 3D brain tumor ROIs. We also propose a temporal-agnostic masking strategy to effectively learn relationships between semantic tokens in the latent space. Our experiments demonstrate that the proposed method can generate high-quality 3D MRIs of brain tumor regions for both low- and high-grade glioma (LGG/HGG) in the BraTS 2019 dataset.

Download here

An Attention-based Multi-Scale Feature Learning Network for Multimodal Medical Image Fusion

Arxiv, 2022

In this paper, we introduce a novel Dilated Residual Attention Network for the medical image fusion task to synthesize complementary information from multimodal medical images (CT and MRI) into a single image. Our network is capable to extract multi-scale deep semantic features. Furthermore, we propose a novel fixed fusion strategy termed Softmax-based weighted strategy based on the Softmax weights and matrix nuclear norm. Extensive experiments show our proposed network and fusion strategy exceed the state-of-the-art performance compared with reference image fusion methods on four commonly used fusion metrics.

Download here

Truncated LinUCB for Stochastic Linear Bandits

Under Review, Electronic Journal of Statistics , 2022

We consider contextual bandits with a finite number of arms, where the contexts are independent and identically distributed d-dimensional random vectors, and the expected rewards are linear in both the arm parameters and contexts. We propose a truncated version of LinUCB and termed “Tr-LinUCB”, which follows LinUCB up to a truncation time S and performs pure exploitation afterwards. The Tr-LinUCB algorithm is shown to achieve O(dlog(T)) regret if S=Cdlog(T) for a sufficiently large constant C, and a matching lower bound is established, which shows the rate optimality of Tr-LinUCB in both d and T under a low dimensional regime.

Download here

Domain Transfer Through Image-to-Image Translation in Prostate Cancer Detection

Abstract, 20th Annual Symposium of the Imaging Network of Ontario (ImNO), 2022

In this paper, we have presented a novel approach for unpaired image-to-image translation of prostate mp-MRI for classifying clinically significant PCa, to be applied in data-constrained settings. First, we introduce domain transfer, a novel pipeline to translate unpaired 3.0T multi-parametric prostate MRIs to 1.5T, to increase the number of training data. Second, we estimate the uncertainty of our models through an evidential deep learning approach; and leverage the dataset filtering technique during the training process. Furthermore, we introduce a simple, yet efficient Evidential Focal Loss that incorporates the focal loss with evidential uncertainty to train our model. Experiments have shown the superior performance of the proposed approach.

Download here