Deep Learning Fusion Research Articles

A Fusion Deep Learning Model of ResNet and Vision Transformer for 3D CT Images

A Fusion Deep Learning Model of ResNet and Vision Transformer for 3D CT Images

COVID-19 classification based on a deep learning and machine learning fusion technique using chest CT images

Coronavirus disease (COVID-19), impacted by SARS-CoV-2, is one of the greatest challenges of the twenty-first century. COVID-19 broke out in the world over the last 2 years and has caused many injuries and killed persons. Computer-aided diagnosis has become a necessary tool to prevent the spreading of this virus. Detecting COVID-19 at an early stage is essential to reduce the mortality risk of patients. Researchers seek to find rapid solutions based on techniques of Machine Learning and Deep Learning. In this paper, we introduced a hybrid model for COVID-19 detection based on machine learning and deep learning models. We used 10 different deep CNN network models to extract features from CT images. We extract features from different layers in each network and find the optimum layer that gives the best-extracted features for each CNN network. Then, for classifying these features, we used five different classifiers based on machine learning. The dataset consists of 2481 CT images divided into COVID-19 and non-COVID-19 categories. Three folds are extracted with a different size between testing and training. Through experiments, we define the best layer for all used CNN networks, the best network, and the best-used classifier. The measured performance shows the superiority of the proposed system over the literature with a highest accuracy of 99.39%. Our models are tested with the three folds that gained maximum average accuracy. The result is 98.69%.

CheXNet and feature pyramid network: a fusion deep learning architecture for multilabel chest X-Ray clinical diagnoses classification.

The existing multilabel X-Ray image learning tasks generally contain much information on pathology co-occurrence and interdependency, which is very important for clinical diagnosis. However, the challenging part of this subject is to accurately diagnose multiple diseases that occurred in a single X-Ray image since multiple levels of features are generated in the images, and create different features as in single label detection. Various works were developed to address this challenge with proposed deep learning architectures to improve classification performance and enrich diagnosis results with multi-probability disease detection. The objective is to create an accurate result and a faster inference system to support a quick diagnosis in the medical system. To contribute to this state-of-the-art, we designed a fusion architecture, CheXNet and Feature Pyramid Network (FPN), to classify and discriminate multiple thoracic diseases from chest X-Rays. This concept enables the model to extract while creating a pyramid of feature maps with different spatial resolutions that capture low-level and high-level semantic information to encounter multiple features. The model's effectiveness is evaluated using the NIH ChestXray14 dataset, with the Area Under Curve (AUC) and accuracy metrics used to compare the results against other cutting-edge approaches. The overall results demonstrate that our method outperforms other approaches and has become promising for multilabel disease classification in chest X-Rays, with potential applications in clinical practice. The result demonstrated that we achieved an average AUC of 0.846 and an accuracy of 0.914. Further, our proposed architecture diagnoses images in 0.013s, faster than the latest approaches.

CT-based deep learning model for predicting hospital discharge outcome in spontaneous intracerebral hemorrhage.

To predict the functional outcome of patients with intracerebral hemorrhage (ICH) using deep learning models based on computed tomography (CT) images. A retrospective, bi-center study of ICH patients was conducted. Firstly, a custom 3D convolutional model was built for predicting the functional outcome of ICH patients based on CT scans from randomly selected ICH patients in H training dataset collected from H hospital. Secondly, clinical data and radiological features were collected at admission and the Extreme Gradient Boosting (XGBoost) algorithm was used to establish a second model, named the XGBoost model. Finally, the Convolution model and XGBoost model were fused to build the third "Fusion model." Favorable outcome was defined as modified Rankin Scale score of 0-3 at discharge. The prognostic predictive accuracy of the three models was evaluated using an H test dataset and an external Y dataset, and compared with the performance of ICH score and ICH grading scale (ICH-GS). A total of 604 patients with ICH were included in this study, of which 450 patients were in the H training dataset, 50 patients in the H test dataset, and 104 patients in the Y dataset. In the Y dataset, the areas under the curve (AUCs) of the Convolution model, XGBoost model, and Fusion model were 0.829, 0.871, and 0.905, respectively. The Fusion model prognostic performance exceeded that of ICH score and ICH-GS (p = 0.043 and p = 0.045, respectively). Deep learning models have good accuracy for predicting functional outcome of patients with spontaneous intracerebral hemorrhage. The proposed deep learning Fusion model may assist clinicians in predicting functional outcome and developing treatment strategies, thereby improving the survival and quality of life of patients with spontaneous intracerebral hemorrhage. • Integrating clinical presentations, CT images, and radiological features to establish deep learning model for functional outcome prediction of patients with intracerebral hemorrhage. • Deep learning applied to CT images provides great help in prognosing functional outcome of intracerebral hemorrhage patients. • The developed deep learning model performs better than clinical prognostic scores in predicting functional outcome of patients with intracerebral hemorrhage.

Empowering fire and smoke detection in smart monitoring through deep learning fusion

Empowering fire and smoke detection in smart monitoring through deep learning fusion

MIFDELN: A multi-sensor information fusion deep ensemble learning network for diagnosing bearing faults in noisy scenarios

Owing to the harsh operating environment of rolling bearings, acquired vibration signals contain strong noise interference, which makes it challenging for conventional methods to effectively extract fault information directly from the original signal. Additionally, obtaining satisfactory diagnostic results using traditional methods based on a single sensor is often difficult because the complementary information between different sensors is ignored. To address these challenges, this paper proposes a multi-sensor information fusion deep ensemble learning network (MIFDELN) for diagnosing bearing faults. First, the collected multi-sensor signals are organically fused using the weighted fusion strategy based on the composite index, which not only avoids the selection of poor-quality sensor signals but also suppresses part of the strong noise interference in the original signal. Second, the cross-scale attention feature extraction module (CAFEM) is presented to automatically learn discriminative features from the fused signals and reduce the impact of useless feature information through cross-scale learning and attentional weight enhancement. Subsequently, a weighted topology learning module (WTLM) is introduced to further excavate the spatial structure features and strengthen the distinctness of the learned features. Finally, the softmax classifier is employed to finalize the fault identification. Two experiments were conducted to verify the effectiveness and superiority of the proposed approach. The results indicated that the proposed approach performs best in extracting discriminating features with the highest accuracy and greatest robustness compared with several representative multi-sensor fusion technologies.

An enhanced multimodal fusion deep learning neural network for lung cancer classification

Cancer remains one of the leading causes of mortality worldwide, necessitating continuous advancements in early diagnosis and treatment. Deep learning, a subset of artificial intelligence, has emerged as a powerful tool in the field of medical image analysis, revolutionizing the way cancer is detected and diagnosed. The study discusses the various modalities employed in lung cancer diagnosis, such as medical imaging (e.g., radiology and pathology), genomics, and clinical data, highlighting the unique challenges associated with each domain. The proposed Multimodal Fusion Deep Neural Network (MFDNN) architecture design effectively integrates information from different modalities (e.g., medical imaging, genomics, clinical data) to enhance lung cancer diagnostic accuracy. Furthermore, it delves into the integration of clinical data, electronic health records, and multimodal approaches to improve the accuracy and reliability of lung cancer diagnosis. Moreover, we highlight the ethical considerations surrounding the deployment of Artificial Intelligence (AI) in clinical settings and the need for robust validation and regulatory guidelines. The Multimodal Fusion Deep Neural Network (MFDNN) achieves an exceptional accuracy rate of 92.5 %, marking a significant breakthrough in the realm of medical AI. MFDNN excels in precision, with 87.4 % accuracy in predicting cancer cases, and equally impresses in recall, capturing approximately 86.4 % of actual cancerous cases. The F1-score of 86.2 further exemplifies MFDNN's ability to strike a harmonious equilibrium, ensuring both diagnostic accuracy and minimized missed diagnoses. The performance is compared with established methods like CNN, DNN, and ResNet. The results underscore MFDNN's pivotal role in revolutionizing lung cancer diagnosis, promising more accurate and timely identification of this critical condition.

YWnet: A convolutional block attention-based fusion deep learning method for complex underwater small target detection

Underwater target detection holds a noteworthy role in the field of marine exploration. However, it is difficult to extract useful feature information from blurred images with complex backgrounds, resulting in suboptimal and unsatisfactory target detection in conventional models. Among them, YOLOv5 leverages the advantages of fast detection performs better in detecting underwater samples. Nevertheless, YOLOv5 still faces difficulties including missed and incorrect detections due to the underwater environment's small scale of objects, dense distribution of organisms, and occlusion. To address these challenges, we propose a novel YoLoWaternet (YWnet) model that builds upon the YOLOv5 framework for complex underwater species detection with three main innovations: 1) A convolutional block attention module (CBAM) is introduced to enhance feature extraction for blurry images in the initial stages of the network and a new feature fusion network called the CRFPN is created to transfer important information and detect underwater targets. 2) A novel feature extraction module is presented, namely, the skip residual C3 module (SRC3), by effectively merging information from various scales to minimize the loss of original data during transmission. 3) Regression and classification algorithms are separated using the decoupled head to improve the effectiveness of detection and the EIoU loss function is employed to accelerate the convergence speed. Finally, the experimental results demonstrate that YWnet achieves remarkable accuracies of 73.2% mAp and 39.3% mAp50–95 on the underwater dataset, surpassing YOLOv5 by 2.3% and 2.4%, respectively. Furthermore, the proposed fusion model outperforms nine state-of-the-art baseline models on the undersea dataset and has generalization capabilities in other datasets.

A fusion deep learning framework based on breast cancer grade prediction

In breast cancer grading, the subtle differences between HE-stained pathological images and the insufficient number of data samples lead to grading inefficiency. With its rapid development, deep learning technology has been widely used for automatic breast cancer grading based on pathological images. In this paper, we propose an integrated breast cancer grading framework based on a fusion deep learning model, which uses three different convolutional neural networks as submodels to extract feature information at different levels from pathological images. Then, the output features of each submodel are learned by the fusion network based on stacking to generate the final decision results. To validate the effectiveness and reliability of our proposed model, we perform dichotomous and multiclassification experiments on the Invasive Ductal Carcinoma (IDC) pathological image dataset and a generated dataset and compare its performance with those of the state-of-the-art models. The classification accuracy of the proposed fusion network is 93.8%, the recall is 93.5%, and the F1 score is 93.8%, which outperforms the state-of-the-art methods.

Multi-Dimensional Fusion Deep Learning for Side Channel Analysis

The rapid advancement of deep learning has significantly heightened the threats posed by Side-Channel Attacks (SCAs) to information security, transforming their effectiveness to a degree several orders of magnitude superior to conventional signal processing techniques. However, the majority of existing Deep-Learning Side-Channel Attacks (DLSCAs) primarily focus on the classification accuracy of the trained model at the attack stage, often assuming that adversaries have unlimited computational and time resources during the profiling stage. This might result in an inflated assessment of the trained model’s fitting capability in a real attack scenario. In this paper, we present a novel DLSCA model, called a Multi-Dimensional Fusion Convolutional Residual Dendrite (MD_CResDD) network, to enhance and speed up the feature extraction process by incorporating a multi-scale feature fusion mechanism. By testing the proposed model on two software implementations of AES-128, we show that it is feasible to improve the profiling speed by at least 34% compared to other existing deep-learning models for DLSCAs and meanwhile achieved a certain level of improvement (8.4% and 0.8% for two implementations) in the attack accuracy. Furthermore, we also investigate how different fusion approaches, fusion times, and residual blocks can affect the attack efficiency on the same two datasets.

Deep learning method for magnetic resonance imaging fluid-attenuated inversion recovery image synthesis

Magnetic resonance imaging(MRI) can obtain multi-modal images with different contrast, which provides rich information for clinical diagnosis. However, some contrast images are not scanned or the quality of the acquired images cannot meet the diagnostic requirements due to the difficulty of patient's cooperation or the limitation of scanning conditions. Image synthesis techniques have become a method to compensate for such image deficiencies. In recent years, deep learning has been widely used in the field of MRI synthesis. In this paper, a synthesis network based on multi-modal fusion is proposed, which firstly uses a feature encoder to encode the features of multiple unimodal images separately, and then fuses the features of different modal images through a feature fusion module, and finally generates the target modal image. The similarity measure between the target image and the predicted image in the network is improved by introducing a dynamic weighted combined loss function based on the spatial domain and K-space domain. After experimental validation and quantitative comparison, the multi-modal fusion deep learning network proposed in this paper can effectively synthesize high-quality MRI fluid-attenuated inversion recovery (FLAIR) images. In summary, the method proposed in this paper can reduce MRI scanning time of the patient, as well as solve the clinical problem of missing FLAIR images or image quality that is difficult to meet diagnostic requirements.

Novel Solution for Using Neural Networks for Kidney Boundary Extraction in 2D Ultrasound Data.

Background and Objective: Kidney ultrasound (US) imaging is a significant imaging modality for evaluating kidney health and is essential for diagnosis, treatment, surgical intervention planning, and follow-up assessments. Kidney US image segmentation consists of extracting useful objects or regions from the total image, which helps determine tissue organization and improve diagnosis. Thus, obtaining accurate kidney segmentation data is an important first step for precisely diagnosing kidney diseases. However, manual delineation of the kidney in US images is complex and tedious in clinical practice. To overcome these challenges, we developed a novel automatic method for US kidney segmentation. Methods: Our method comprises two cascaded steps for US kidney segmentation. The first step utilizes a coarse segmentation procedure based on a deep fusion learning network to roughly segment each input US kidney image. The second step utilizes a refinement procedure to fine-tune the result of the first step by combining an automatic searching polygon tracking method with a machine learning network. In the machine learning network, a suitable and explainable mathematical formula for kidney contours is denoted by basic parameters. Results: Our method is assessed using 1380 trans-abdominal US kidney images obtained from 115 patients. Based on comprehensive comparisons of different noise levels, our method achieves accurate and robust results for kidney segmentation. We use ablation experiments to assess the significance of each component of the method. Compared with state-of-the-art methods, the evaluation metrics of our method are significantly higher. The Dice similarity coefficient (DSC) of our method is 94.6 ± 3.4%, which is higher than those of recent deep learning and hybrid algorithms (89.4 ± 7.1% and 93.7 ± 3.8%, respectively). Conclusions: We develop a coarse-to-refined architecture for the accurate segmentation of US kidney images. It is important to precisely extract kidney contour features because segmentation errors can cause under-dosing of the target or over-dosing of neighboring normal tissues during US-guided brachytherapy. Hence, our method can be used to increase the rigor of kidney US segmentation.

SADFusion: A multi-scale infrared and visible image fusion method based on salient-aware and domain-specific

Image fusion intends to generate an informative image with maximally possible features and details of the source images. However, the existing deep-learning fusion methods are rarely considering the discrepancy of visible and infrared (IVIF) modalities, besides losing control of balance in preserving the texture details and thermal target, owing to the identical structure and parameters of multi-modality feature extractors’ designing and weak domain guidance of network training. In this paper, a novel multi-scale fusion network based on the salient-aware and domain-specific methods is proposed, termed as SADFusion. To boost the multi-modality feature extracting performance of our method, the proposed network adopts dual encoding with short connection and multi-scale structure. Moreover, the domain-specific framework and corresponding training strategy are designed to achieve the identical encoding for different image modalities. Moreover, multi-scale attention fusion modules (MSFAF modules) are proposed to effectively fuse the extracted complementary features at every scale. Finally, we construct the specific salient-aware loss to guide the model to trade off the preserving necessary information, by utilizing salient modality features as pixel-to-pixel intensity and gradient maps. Experiments based on the public datasets demonstrate the superiority of our method over the state-of-art fusion methods, which particularly highlights the targets and retains the effective information.

Enhancing Visual Feedback Control through Early Fusion Deep Learning.

A visual servoing system is a type of control system used in robotics that employs visual feedback to guide the movement of a robot or a camera to achieve a desired task. This problem is addressed using deep models that receive a visual representation of the current and desired scene, to compute the control input. The focus is on early fusion, which consists of using additional information integrated into the neural input array. In this context, we discuss how ready-to-use information can be directly obtained from the current and desired scenes, to facilitate the learning process. Inspired by some of the most effective traditional visual servoing techniques, we introduce early fusion based on image moments and provide an extensive analysis of approaches based on image moments, region-based segmentation, and feature points. These techniques are applied stand-alone or in combination, to allow obtaining maps with different levels of detail. The role of the extra maps is experimentally investigated for scenes with different layouts. The results show that early fusion facilitates a more accurate approximation of the linear and angular camera velocities, in order to control the movement of a 6-degree-of-freedom robot from a current configuration to a desired one. The best results were obtained for the extra maps providing details of low and medium levels.

Monitoring spring leaf phenology of individual trees in a temperate forest fragment with multi-scale satellite time series

Forest fragmentation has been increasingly exacerbated by deforestation, urbanization, and agricultural expansion. Monitoring the forest fragments via the lens of tree-crown scale leaf phenology is critical to understand tree species phenological responses to climate change and identify the fragment species vulnerable to environmental disturbance. Despite advances in remote sensing for phenology monitoring, detecting tree-crown scale leaf phenology in fragmented forests remains challenging. Simultaneous tracking of key spring phenological events that are crucial to ecosystem functions and climate change responses is also neglected. To address these challenges, we develop a novel tree-crown scale remote sensing phenological monitoring framework to characterize all the critical spring phenological events of individual trees of deciduous forest fragments, with Trelease Woods in Champaign, Illinois as a case study. The novel framework comprises four components: 1) generate high spatiotemporal resolution fusion imagery from multi-scale satellite time series with a hybrid deep learning fusion model; 2) calibrate PlanetScope imagery time series with fusion data using histogram matching; 3) model tree-crown scale phenology trajectory with a Beck logistic-based method; 4) detect a diversity of tree-crown scale phenological events using several phenological metric extraction methods (i.e., threshold- and curve feature-based methods). Combined with weekly in-situ phenological observations of 123 individual trees across 12 broadleaf species from 2017 to 2020, the framework effectively bridges the satellite- and field-based phenological measures for the key spring phenological events (i.e., budswell, budburst, leaf expansion, and leaf maturity events) at the tree-crown scale, particularly for large individuals (RMSE <1 week for most events). Calibration of PlanetScope imagery using multi-scale satellite fusion data in consideration of landscape fragmentation is critical for monitoring tree phenology of forest fragments. Compared to curve feature-based methods, threshold-based phenometric extraction methods demonstrate enhanced capability in detecting spring leaf phenological dynamics of individual trees. Among the phenological events, full leaf out and early leaf expansion events are retrieved with high accuracy using calibrated PlanetScope time series (RMSE from 3 to 5 days and R-squared higher than 0.8). With both intensive satellite and field phenological efforts, this novel framework is at the forefront of interpreting tree-crown scale remotely sensed phenological metrics in the context of biologically meaningful field phenological events in fragmented forest setting.

Learning-Based Multimodal Information Fusion and Behavior Recognition of Vascular Interventionists' Operating Skills.

The operating skills of vascular interventionists have an important impact on the effect of surgery. However, current research on behavior recognition and skills learning of interventionists' operating skills is limited. In this study, an innovative deep learning-based multimodal information fusion architecture is proposed for recognizing and analyzing eight common operating behaviors of interventionists. An experimental platform integrating four modal sensors is used to collect multimodal data from interventionists. The ANOVA and Manner-Whitney tests is used for relevance analysis of the data. The analysis results demonstrate that there is almost no significant difference ( p <0.001) between the actions related to the unimodal data, which cannot be used for accurate behavior recognition. Therefore, a study of the fusion architecture based on the existing machine learning classifier and the proposed deep learning fusion architecture is carried out. The research findings indicate that the proposed deep learning-based fusion architecture achieves an impressive overall accuracy of 98.5%, surpassing both the machine learning classifier (93.51%) and the unimodal data (90.05%). The deep learning-based multimodal information fusion architecture proves the feasibility of behavior recognition and skills learning of interventionist's operating skills. Furthermore, the application of deep learning-based multimodal fusion technology of surgeon's operating skills will help to improve the autonomy and intelligence of surgical robotic systems.

A Novel Robotic Path Planning Method in Grid Map Context Based on D* Lite Algorithm and Deep Learning

Mobile robot path planning has received more and more attention as an important technology in robotics. Based on the D* Lite algorithm, this paper constructs a robot path planning model in a grid map environment, and proposes a deep learning fusion algorithm for path planning under complex large maps. The D* Lite algorithm with excellent performance adopts the enhanced neural network algorithm with environment self-learning ability for local parts. The model introduces the gentle update method of the Q value in the D* Lite algorithm into the optimization target calculation, calculates the loss function and updates the network parameters, thereby solving the overestimation problem of deep reinforcement learning in the application of mobile robot path planning. In the simulation process, the idea of averaging is introduced into the e-greedy strategy, and the value function output by the previous generation parameter network is used to obtain the average result to determine the next action direction of the mobile robot. The experimental results show that in the simple environment and the complex environment, the planned path lengths in the environment are 44.21, 43.63 and 43.61[Formula: see text]m, respectively, reducing the number of collisions with obstacles during the training process of the mobile robot, and improving the superiority and effectiveness of the algorithm.

Lightweight deep learning model incorporating an attention mechanism and feature fusion for automatic classification of gastric lesions in gastroscopic images.

Accurate diagnosis of various lesions in the formation stage of gastric cancer is an important problem for doctors. Automatic diagnosis tools based on deep learning can help doctors improve the accuracy of gastric lesion diagnosis. Most of the existing deep learning-based methods have been used to detect a limited number of lesions in the formation stage of gastric cancer, and the classification accuracy needs to be improved. To this end, this study proposed an attention mechanism feature fusion deep learning model with only 14 million (M) parameters. Based on that model, the automatic classification of a wide range of lesions covering the stage of gastric cancer formation was investigated, including non-neoplasm(including gastritis and intestinal metaplasia), low-grade intraepithelial neoplasia, and early gastric cancer (including high-grade intraepithelial neoplasia and early gastric cancer). 4455 magnification endoscopy with narrow-band imaging(ME-NBI) images from 1188 patients were collected to train and test the proposed method. The results of the test dataset showed that compared with the advanced gastric lesions classification method with the best performance (overall accuracy = 94.3%, parameters = 23.9 M), the proposed method achieved both higher overall accuracy and a relatively lightweight model (overall accuracy =95.6%, parameter = 14 M). The accuracy, sensitivity, and specificity of low-grade intraepithelial neoplasia were 94.5%, 93.0%, and 96.5%, respectively, achieving state-of-the-art classification performance. In conclusion, our method has demonstrated its potential in diagnosing various lesions at the stage of gastric cancer formation.

Deployment of the deep learning fusion method to emotional semantic evaluation of natural language

The emotional semantic evaluation of natural language plays a crucial role in sentiment analysis. Deep learning methods have shown great potential in capturing the complex relationships between words and emotions. This paper proposes a deep learning fusion method for deploying emotional semantic evaluation. The technique combines multiple deep learning architectures to capture local and global contextual information, including Bidirectional Gated Recurrent Units (GRU), Long Short-Term Memory (LSTM) networks, and Self-Attention mechanisms. Pretrained GloVe word embedding’s utilized to enhance word representation. A novel fusion layer combines the outputs of individual models; employing self-attention means to assign weights dynamically. This allows the model to weigh the importance of different representations in the final prediction. Benchmark movie review (MR) for sentiment analysis and emotion classification tasks are used to evaluate the proposed method. Experimental results demonstrate superior performance compared to individual deep learning models and traditional feature-based approaches. The proposed fusion method effectively captures the nuances of emotional semantics in natural language, leading to more accurate and nuanced evaluations.

Deep learning model fusion improves lung tumor segmentation accuracy across variable training-to-test dataset ratios.

This study aimed to investigate the robustness of a deep learning (DL) fusion model for low training-to-test ratio (TTR) datasets in the segmentation of gross tumor volumes (GTVs) in three-dimensional planning computed tomography (CT) images for lung cancer stereotactic body radiotherapy (SBRT). A total of 192 patients with lung cancer (solid tumor, 118; part-solid tumor, 53; ground-glass opacity, 21) who underwent SBRT were included in this study. Regions of interest in the GTVs were cropped based on GTV centroids from planning CT images. Three DL models, 3D U-Net, V-Net, and dense V-Net, were trained to segment the GTV regions. Nine fusion models were constructed with logical AND, logical OR, and voting of the two or three outputs of the three DL models. TTR was defined as the ratio of the number of cases in a training dataset to that in a test dataset. The Dice similarity coefficients (DSCs) and Hausdorff distance (HD) of the 12 models were assessed with TTRs of 1.00 (training data: validation data: test data = 40:20:40), 0.791 (35:20:45), 0.531 (31:10:59), 0.291 (20:10:70), and 0.116 (10:5:85). The voting fusion model achieved the highest DSCs of 0.829 to 0.798 for all TTRs among the 12 models, whereas the other models showed DSCs of 0.818 to 0.804 for a TTR of 1.00 and 0.788 to 0.742 for a TTR of 0.116, and an HD of 5.40 ± 3.00 to 6.07 ± 3.26mm better than any single DL models. The findings suggest that the proposed voting fusion model is a robust approach for low TTR datasets in segmenting GTVs in planning CT images of lung cancer SBRT.