Accurate glioma MRI segmentation is pivotal for diagnosis, tumor characterization, grade detection, molecular subtyping, and treatment response assessment. However, existing models are often hindered by high computational costs and insufficient segmentation performance, limiting their clinical utility. To address these challenges, we propose a lightweight Mixed Vision Transformer (MViT) with an adaptive residual-enhanced network, named MViT-ARNet, for 3D glioma MRI segmentation. MViT-ARNet employs an encoder-decoder architecture. The encoder leverages a lightweight framework incorporating MViT modules, which effectively capture both local details and longrange dependencies. Crucially, it fuses features extracted by convolutional layers with those learned through spatial reduction attention in the Transformer, significantly reducing model parameters and enhancing segmentation accuracy. Within the decoder, we introduce an adaptive residual-enhanced upsampling module, which progressively reconstructs high-resolution segmentation maps by adaptively integrating skip features from the encoder. Evaluated on a mixed multimodal MRI dataset, the model demonstrates notable efficiency, requiring only 4.41 M parameters and 22.27 GFLOPs. It achieves robust segmentation performance, with Dice scores of 0.7632 for the enhancing tumor, 0.8229 for the tumor core, and 0.9022 for the whole tumor on the test set. These results demonstrate the potential of MViT-ARNet for accurate glioma delineation and its viability for real-world clinical applications.

Accurate MRI-based glioma segmentation and IDH genotyping are critical for surgical planning, prognostic assessment, and therapeutic decision-making. Existing approaches typically address glioma segmentation and IDH genotyping independently, requiring separate models for each task and thereby limiting the potential benefits from shared representations. To overcome these limitations, we present MTL-SAM3D, an interactive multi-task framework that simultaneously outputs wholetumor segmentation together with IDH genotype predictions while remaining effective on unseen datasets without any targetdomain fine-tuning. The design freezes the SAM3D image encoder and inserts LoRA adapters only in the query and value projections, so the number of trainable parameters stays small. A 3D prompt encoder converts point and mask prompts into embeddings. Subsequently, the mask decoder generates voxellevel tumor masks, after which the resulting image features along with the predicted masks are aggregated by the IDH classifier to produce case-level genotype. Trained on the UCSF-PDGM dataset, MTL-SAM3D achieves segmentation performance with Dice score of 0.91, HD95 of 4.66 mm, and achieves IDH genotyping performance with AUC of 98.34 %, accuracy of 95.96 % on the UCSF-PDGM test set. When the same weights are evaluated on BraTS 2020 without any fine-tuning, the model still delivers Dice of 0.82 and AUC of 88.04 %, surpassing state-of-the-art baselines in this cross-domain setting.

Deep clustering has exhibited remarkable performance; however, the over confidence problem, i.e., the estimated confidence for a sample belonging to a particular cluster greatly exceeds its actual prediction accuracy, has been over looked in prior research. To tackle this critical issue, we pioneer the development of a calibrated deep clustering framework. Specifically, we propose a novel dual head (calibration head and clustering head) deep clustering model that can effectively calibrate the estimated confidence and the actual accuracy. The calibration head adjusts the overconfident predictions of the clustering head, generating prediction confidence that matches the model learning status. Then, the clustering head dynamically selects reliable high-confidence samples estimated by the calibration head for pseudo-label self-training. Additionally, we introduce an effective network initialization strategy that enhances both training speed and network robustness. The effectiveness of the proposed calibration approach and initialization strategy are both endorsed with solid theoretical guarantees. Extensive experiments demonstrate the proposed calibrated deep clustering model not only surpasses the state-of-the-art deep clustering methods by 5× on average in terms of expected calibration error, but also significantly outperforms them in terms of clustering accuracy. The code is available at https://github.com/ChengJianH/CDC.

In partial label learning (PLL), every sample is associated with a candidate label set comprising the ground-truth label and several noisy labels. The conventional PLL assumes the noisy labels are randomly generated (instance-independent), while in practical scenarios, the noisy labels are always instance-dependent and are highly related to the sample features, leading to the instance-dependent partial label learning (IDPLL) problem. Instance-dependent noisy label is a double-edged sword. On one side, it may promote model training as the noisy labels can depict the sample to some extent. On the other side, it brings high label ambiguity as the noisy labels are quite undistinguishable from the ground-truth label. To leverage the nuances of IDPLL effectively, for the first time we create class-wise embeddings for each sample, which allow us to explore the relationship of instance-dependent noisy labels, i.e., the class-wise embeddings in the candidate label set should have high similarity, while the class-wise embeddings between the candidate label set and the non-candidate label set should have high dissimilarity. Moreover, to reduce the high label ambiguity, we introduce the concept of class prototypes containing global feature information to disambiguate the candidate label set. Extensive experimental comparisons with twelve methods on six benchmark data sets, including four fine-grained data sets, demonstrate the effectiveness of the proposed method. The code implementation is publicly available at https://github.com/Yangfc-ML/CEL.

Isocitrate dehydrogenase (IDH) is a key molecular feature for gliomas, and the prediction of IDH is also an important task for computer-aided diagnosis using magnetic resonance imaging (MRI). To address this changllenge, we introduce a multi-modal MRI-based characteristics inspired network for IDH Genotyping (M3CI-Net), which pay more attention to the different characteristics information of different MRI modalities T1, T2, T1ce, Flair. In M3CI-Net, a pre-fusion module with multi-channel attention mechanism is used to fuse T1ce and Flair modalities and capture as much as possible luminance and contrast information, and the edge information is obtained from T2 modality by using edge detection module. Finally, the feature information between modalities are fused and input into a CNN-Transformer based encoder structure to extract shared spatial and global information from multi-modal MRI, and the information of multiple scales frome encoder are input into the linear layer for IDH genotype classification after pooling, meanwhile, the CNN based decoder with skip-connection for glioma segmentation works for assisting IDH genotyping. Then, we proposed images’ pre-fusion loss, segmentation loss, IDH genotyping loss, and use uncertainty weight training method to balance the weights of these loss. we evaluate our proposed method on Brats2020, and achieve an acceracy of 0.88, an AUC of 0.94, a specificity of 0.92, a sensitivity of 0.84 in IDH genotyping, which is superior to the state-of-the-art methods.

Accurately predicting survival of esophageal cancer is essential for clinical precision treatment. However, the existing region of interest (ROI) based methods not only require prior medical knowledge to complete the delineation of tumor, but may also lead to excessive sensitivity of the model towards ROI. To address these challenges, we design a fully automated CT-guided learning that combines a CNN-Transformer size aware U-Net and a ranked survival prediction network together to automatically predict the survival of patients with esophageal cancer. Specifically, we first incorporate the Transformer with shifted windowing multi-head self-attention mechanism into the base of the encoder in the U-Net to capture the long-range dependency in the 3D CT images. Then, to alleviate the imbalance between the ROI and the background in CT images, we design a size-aware coefficient for the segmentation loss. Finally, we design a ranked pair sorting loss to learn more fully the ranked information hidden in esophageal cancer patients. To validate the effectiveness of our method, we conduct extensive experiments on a dataset containing 759 esophageal cancer samples. The experimental results demonstrate that our proposed method can still achieve the best performance in survival prediction without ROI ground truth.

Ki67 is a promising molecular biomarker for the diagnosis of lung adenocarcinoma. However, previous methods to determine Ki67 status often require tumor tissue sampling, which is invasive for patients. This study proposes a multi-view signature fusion network (MVSF), combining deep learning encoded (DLE) signatures, handcrafted radiomics (HCR) signatures, and clinical information to noninvasively predict Ki67 status. Multi-view signatures are combined through a tensor fusion network to obtain potentially high-dimensional signatures. Finally, a cooperative game theory-based approach is applied to quantitatively interpret the contribution of signatures to decision-making. The proposed MVSF is evaluated on a retrospectively collected dataset of 661 patients. Experimental results show that the MVSF achieves encouraging performance, with an area under the receiver operating characteristic curve of 0.80 and an accuracy of 0.78, outperforming several state-of-the-art Ki67 status prediction methods, which implies that our proposed method could provide potential support for Ki67 status prediction.

The accurate prediction of glioma grade before surgery is essential for treatment planning and prognosis. Since the gold standard (i.e., biopsy)for grading gliomas is both highly invasive and expensive, and there is a need for a noninvasive and accurate method. In this study, we proposed a novel radiomics-based pipeline by incorporating the intratumoral and peritumoral features extracted from preoperative mpMRI scans to accurately and noninvasively predict glioma grade. To address the unclear peritumoral boundary, we designed an algorithm to capture the peritumoral region with a specified radius. The mpMRI scans of 285 patients derived from a multi-institutional study were adopted. A total of 2153 radiomic features were calculated separately from intratumoral volumes (ITVs)and peritumoral volumes (PTVs)on mpMRI scans, and then refined using LASSO and mRMR feature ranking methods. The top-ranking radiomic features were entered into the classifiers to build radiomic signatures for predicting glioma grade. The prediction performance was evaluated with five-fold cross-validation on a patient-level split. The radiomic signatures utilizing the features of ITV and PTV both show a high accuracy in predicting glioma grade, with AUCs reaching 0.968. By incorporating the features of ITV and PTV, the AUC of IPTV radiomic signature can be increased to 0.975, which outperforms the state-of-the-art methods. Additionally, our proposed method was further demonstrated to have strong generalization performance in an external validation dataset with 65 patients. The source code of our implementation is made publicly available at https://github.com/chengjianhong/glioma_grading.git.

Accurate and rapid diagnosis of coronavirus disease 2019 (COVID-19) from chest CT scans is of great importance and urgency during the worldwide outbreak. However, radiologists have to distinguish COVID-19 pneumonia from other pneumonia in a large number of CT scans, which is tedious and inefficient. Thus, it is urgently and clinically needed to develop an efficient and accurate diagnostic tool to help radiologists to fulfill the difficult task. In this study, we proposed a deep supervised autoencoder (DSAE) framework to automatically identify COVID-19 using multi-view features extracted from CT images. To fully explore features characterizing CT images from different frequency domains, DSAE was proposed to learn the latent representation by multi-task learning. The proposal was designed to both encode valuable information from different frequency features and construct a compact class structure for separability. To achieve this, we designed a multi-task loss function, which consists of a supervised loss and a reconstruction loss. Our proposed method was evaluated on a newly collected dataset of 787 subjects including COVID-19 pneumonia patients, other pneumonia patients, and normal subjects without abnormal CT findings. Extensive experimental results demonstrated that our proposed method achieved encouraging diagnostic performance and may have potential clinical application for the diagnosis of COVID-19.

Effective fusion of multimodal magnetic resonance imaging (MRI) is of great significance to boost the accuracy of glioma grading thanks to the complementary information provided by different imaging modalities. However, how to extract the common and distinctive information from MRI to achieve complementarity is still an open problem in information fusion research. In this study, we propose a deep neural network model termed as multimodal disentangled variational autoencoder (MMD-VAE) for glioma grading based on radiomics features extracted from preoperative multimodal MRI images. Specifically, the radiomics features are quantized and extracted from the region of interest for each modality. Then, the latent representations of variational autoencoder for these features are disentangled into common and distinctive representations to obtain the shared and complementary data among modalities. Afterwards, cross-modality reconstruction loss and common-distinctive loss are designed to ensure the effectiveness of the disentangled representations. Finally, the disentangled common and distinctive representations are fused to predict the glioma grades, and SHapley Additive exPlanations (SHAP) is adopted to quantitatively interpret and analyze the contribution of the important features to grading. Experimental results on two benchmark datasets demonstrate that the proposed MMD-VAE model achieves encouraging predictive performance (AUC:0.9939) on a public dataset, and good generalization performance (AUC:0.9611) on a cross-institutional private dataset. These quantitative results and interpretations may help radiologists understand gliomas better and make better treatment decisions for improving clinical outcomes.

The accurate prediction of isocitrate dehydrogenase (IDH) mutation and glioma segmentation are important tasks for computer-aided diagnosis using preoperative multimodal magnetic resonance imaging (MRI). The two tasks are ongoing challenges due to the significant inter-tumor and intra-tumor heterogeneity. The existing methods to address them are mostly based on single-task approaches without considering the correlation between the two tasks. In addition, the acquisition of IDH genetic labels is expensive and costly, resulting in a limited number of IDH mutation data for modeling. To comprehensively address these problems, we propose a fully automated multimodal MRI-based multi-task learning framework for simultaneous glioma segmentation and IDH genotyping. Specifically, the task correlation and heterogeneity are tackled with a hybrid CNN-Transformer encoder that consists of a convolutional neural network and a transformer to extract the shared spatial and global information learned from a decoder for glioma segmentation and a multi-scale classifier for IDH genotyping. Then, a multi-task learning loss is designed to balance the two tasks by combining the segmentation and classification loss functions with uncertain weights. Finally, an uncertainty-aware pseudo-label selection is proposed to generate IDH pseudo-labels from larger unlabeled data for improving the accuracy of IDH genotyping by using semi-supervised learning. We evaluate our method on a multi-institutional public dataset. Experimental results show that our proposed multi-task network achieves promising performance and outperforms the single-task learning counterparts and other existing state-of-the-art methods. With the introduction of unlabeled data, the semi-supervised multi-task learning framework further improves the performance of glioma segmentation and IDH genotyping. The source codes of our framework are publicly available at https://github.com/miacsu/MTTU-Net.git.

Automatic identification of adventitious respiratory sound has still been a challenging problem in recent years. To address this challenge, we propose an adventitious respiratory sound classification network (ARSC-Net), which combines residual block with channel-spatial attention for accurate classification. Specifically, we extract two types of features from adventitious respiratory sound, including Mel-Frequency Cepstral Coefficients (MFCCs) and Mel-spectrogram. The two types of features are entered into the parallel encoders paths with residual attention for extracting feature representation, and then fused into a channel-spatial attention module to adaptively focus on the important features between channel and spatial part for the classification task. Moreover, the channel-spatial attention can enhance the feature representation, in which the channel attention explores the inter-channel relationship of the spectrums, and then the inter-spatial correlation mapping is generated by the spatial attention introduced serially. We evaluate our proposed method on ICBHI 2017 database. Experimental results show that our proposed method achieves encouraging predictive performance with an accuracy of 80.0% for identifying abnormal sounds from normal sounds, and with an accuracy of 92.4% for distinguishing crackles from wheezes. In addition, our method also achieves a score of 56.76% for the four-class sound classification of adventitious sounds and outperforms several state-of-the-art methods.

Epidermal growth factor receptor (EGFR) genotyping is essential to treatment guidelines for the use of tyrosine kinase inhibitors in lung adenocarcinoma. However, accurate and noninvasive methods to detect the EGFR gene are ongoing challenges. In this study, we propose a hybrid framework, namely HC-DLR, to noninvasively predict EGFR mutation status by fusing multi-source features including low-level handcrafted radiomics (HCR) features, high-level deep learning-based radiomics (DLR) features, and demographics features. The HCR features first are selected from massive handcrafted features extracted from CT images. The DLR features are also extracted from CT images using the pre-trained 3D DenseNet. Then, multi-source feature representations are refined and fused to build an HC-DLR model for improving the predictive performance of EGFR mutations. The proposed method is evaluated on a newly collected dataset with 670 patients. Experimental results show that the HC-DLR model achieves an encouraging predictive performance with an AUC of 0.76, an accuracy of 72.47%, and an F1-score of 71.35%, which may have potential clinical value for predicting EGFR mutations in lung adenocarcinoma.

Accurate segmentation of glioma from 3D medical images is vital to numerous clinical endpoints. While manual segmentation is subjective and time-consuming, fully automated extraction is quite imperative and challenging due to the intrinsic heterogeneity of tumor structures. In this study, we propose a multi-level glioma segmentation framework, 3D Residual-Attention-Atrous U-Net (RAAU-Net), using 3D U-Net combined attention mechanism with atrous convolution. The 3D RAAU-Net can extract contextual information by combining low- and high-resolution feature maps. The attention mechanism is embedded in each skip connection layer of 3D RAAU-Net to enhance feature representations. Meanwhile, the atrous convolution is adopted in the whole network architecture to incorporate large and rich semantic information. Furthermore, we design a new training scheme to reduce false positives and enhance generalization. Eventually, our proposed segmentation method is evaluated on the validation dataset from the Multimodal Brain Tumor Image Segmentation Challenge (BraTS) 2018 and achieve a competitive result with average Dice score of 88% for the whole tumor, 79% for the tumor core and 73% for the enhancing tumor, respectively. Quantitative results and visual analysis have proven that these improvements in 3D RAAU-Net are effective and achieve a better segmentation accuracy compared with the baseline.