Uncertainty quantification via localized gradients for deep learning-based medical image assessments.

Objective.Deep learning models that aid in medical image assessment tasks must be both accurate and reliable to be deployed within clinical settings. While deep learning models have been shown to be highly accurate across a variety of tasks, measures that indicate the reliability of these models are less established. Increasingly, uncertainty quantification (UQ) methods are being introduced to inform users on the reliability of model outputs. However, most existing methods cannot be augmented to previously validated models because they are not post hoc, and they change a model's output. In this work, we overcome these limitations by introducing a novel post hoc UQ method, termedLocal Gradients UQ, and demonstrate its utility for deep learning-based metastatic disease delineation.Approach.This method leverages a trained model's localized gradient space to assess sensitivities to trained model parameters. We compared the Local Gradients UQ method to non-gradient measures defined using model probability outputs. The performance of each uncertainty measure was assessed in four clinically relevant experiments: (1) response to artificially degraded image quality, (2) comparison between matched high- and low-quality clinical images, (3) false positive (FP) filtering, and (4) correspondence with physician-rated disease likelihood.Main results.(1) Response to artificially degraded image quality was enhanced by the Local Gradients UQ method, where the median percent difference between matching lesions in non-degraded and most degraded images was consistently higher for the Local Gradients uncertainty measure than the non-gradient uncertainty measures (e.g. 62.35% vs. 2.16% for additive Gaussian noise). (2) The Local Gradients UQ measure responded better to high- and low-quality clinical images (p< 0.05 vsp> 0.1 for both non-gradient uncertainty measures). (3) FP filtering performance was enhanced by the Local Gradients UQ method when compared to the non-gradient methods, increasing the area under the receiver operating characteristic curve (ROC AUC) by 20.1% and decreasing the false positive rate by 26%. (4) The Local Gradients UQ method also showed more favorable correspondence with physician-rated likelihood for malignant lesions by increasing ROC AUC for correspondence with physician-rated disease likelihood by 16.2%.Significance. In summary, this work introduces and validates a novel gradient-based UQ method for deep learning-based medical image assessments to enhance user trust when using deployed clinical models.

Deep survival analysis for interpretable time-varying prediction of preeclampsia risk.

OBJECTIVE: Survival analysis is widely utilized in healthcare to predict the timing of disease onset. Traditional methods of survival analysis are usually based on Cox Proportional Hazards model and assume proportional risk for all subjects. However, this assumption is rarely true for most diseases, as the underlying factors have complex, non-linear, and time-varying relationships. This concern is especially relevant for pregnancy, where the risk for pregnancy-related complications, such as preeclampsia, varies across gestation. Recently, deep learning survival models have shown promise in addressing the limitations of classical models, as the novel models allow for non-proportional risk handling, capturing nonlinear relationships, and navigating complex temporal dynamics. METHODS: We present a methodology to model the temporal risk of preeclampsia during pregnancy and investigate the associated clinical risk factors. We utilized a retrospective dataset including 66,425 pregnant individuals who delivered in two tertiary care centers from 2015 to 2023. We modeled the preeclampsia risk by modifying DeepHit, a deep survival model, which leverages neural network architecture to capture time-varying relationships between covariates in pregnancy. We applied time series k-means clustering to DeepHit's normalized output and investigated interpretability using Shapley values. RESULTS: We demonstrate that DeepHit can effectively handle high-dimensional data and evolving risk hazards over time with performance similar to the Cox Proportional Hazards model, achieving an area under the curve (AUC) of 0.78 for both models. The deep survival model outperformed traditional methodology by identifying time-varied risk trajectories for preeclampsia, providing insights for early and individualized intervention. K-means clustering resulted in patients delineating into low-risk, early-onset, and late-onset preeclampsia groups-notably, each of those has distinct risk factors. CONCLUSION: This work demonstrates a novel application of deep survival analysis in time-varying prediction of preeclampsia risk. Our results highlight the advantage of deep survival models compared to Cox Proportional Hazards models in providing personalized risk trajectory and demonstrating the potential of deep survival models to generate interpretable and meaningful clinical applications in medicine.

PTFSpot: deep co-learning on transcription factors and their binding regions attains impeccable universality in plants.

Unlike animals, variability in transcription factors (TFs) and their binding regions (TFBRs) across the plants species is a major problem that most of the existing TFBR finding software fail to tackle, rendering them hardly of any use. This limitation has resulted into underdevelopment of plant regulatory research and rampant use of Arabidopsis-like model species, generating misleading results. Here, we report a revolutionary transformers-based deep-learning approach, PTFSpot, which learns from TF structures and their binding regions' co-variability to bring a universal TF-DNA interaction model to detect TFBR with complete freedom from TF and species-specific models' limitations. During a series of extensive benchmarking studies over multiple experimentally validated data, it not only outperformed the existing software by >30% lead but also delivered consistently >90% accuracy even for those species and TF families that were never encountered during the model-building process. PTFSpot makes it possible now to accurately annotate TFBRs across any plant genome even in the total lack of any TF information, completely free from the bottlenecks of species and TF-specific models.

Combined deep learning and radiomics in pretreatment radiation esophagitis prediction for patients with esophageal cancer underwent volumetric modulated arc therapy.

PURPOSE: To develop a combined radiomics and deep learning (DL) model in predicting radiation esophagitis (RE) of a grade ≥ 2 for patients with esophageal cancer (EC) underwent volumetric modulated arc therapy (VMAT) based on computed tomography (CT) and radiation dose (RD) distribution images. MATERIALS AND METHODS: A total of 273 EC patients underwent VMAT were retrospectively reviewed and enrolled from two centers and divided into training (n = 152), internal validation (n = 66), and external validation (n = 55) cohorts, respectively. Radiomic and dosiomic features along with DL features using convolutional neural networks were extracted and screened from CT and RD images to predict RE. The performance of these models was evaluated and compared using the area under curve (AUC) of the receiver operating characteristic curves (ROC). RESULTS: There were 5 and 10 radiomic and dosiomic features were screened, respectively. XGBoost achieved a best AUC of 0.703, 0.694 and 0.801, 0.729 with radiomic and dosiomic features in the internal and external validation cohorts, respectively. ResNet34 achieved a best prediction AUC of 0.642, 0.657 and 0.762, 0.737 for radiomics based DL model (DLR) and RD based DL model (DLD) in the internal and external validation cohorts, respectively. Combined model of DLD + Dosiomics + clinical factors achieved a best AUC of 0.913, 0.821 and 0.805 in the training, internal, and external validation cohorts, respectively. CONCLUSION: Although the dose was not responsible for the prediction accuracy, the combination of various feature extraction methods was a factor in improving the RE prediction accuracy. Combining DLD with dosiomic features was promising in the pretreatment prediction of RE for EC patients underwent VMAT.

Differential diagnostic value of radiomics models in benign versus malignant vertebral compression fractures: A systematic review and meta-analysis.

PURPOSE: Early diagnosis of benign and malignant vertebral compression fractures by analyzing imaging data is crucial to guide treatment and assess prognosis, and the development of radiomics made it an alternative option to biopsy examination. This systematic review and meta-analysis was conducted with the purpose of quantifying the diagnostic efficacy of radiomics models in distinguishing between benign and malignant vertebral compression fractures. METHODS: Searching on PubMed, Embase, Web of Science and Cochrane Library was conducted to identify eligible studies published before September 23, 2023. After evaluating for methodological quality and risk of bias using the Radiomics Quality Score (RQS) and the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2), we selected studies providing confusion matrix results to be included in random-effects meta-analysis. RESULTS: A total of sixteen articles, involving 1,519 vertebrae with pathological-diagnosed tumor infiltration, were included in our meta-analysis. The combined sensitivity and specificity of the top-performing models were 0.92 (95 % CI: 0.87-0.96) and 0.93 (95 % CI: 0.88-0.96), respectively. Their AUC was 0.97 (95 % CI: 0.96-0.99). By contrast, radiologists' combined sensitivity was 0.90 (95 %CI: 0.75-0.97) and specificity was 0.92 (95 %CI: 0.67-0.98). The AUC was 0.96 (95 %CI: 0.94-0.97). Subsequent subgroup analysis and sensitivity test suggested that part of the heterogeneity might be explained by differences in imaging modality, segmentation, deep learning and cross-validation. CONCLUSION: We found remarkable diagnosis potential in correctly distinguishing vertebral compression fractures in complex clinical contexts. However, the published radiomics models still have a great heterogeneity, and more large-scale clinical trials are essential to validate their generalizability.

A deep-learning-based surrogate model for Monte-Carlo simulations of the linear energy transfer in primary brain tumor patients treated with proton-beam radiotherapy.

OBJECTIVE: This study explores the use of neural networks (NNs) as surrogate models for Monte-Carlo (MC) simulations in predicting the dose-averaged linear energy transfer (LETd) of protons in proton-beam therapy based on the planned dose distribution and patient anatomy in the form of computed tomography (CT) images. As LETdis associated with variability in the relative biological effectiveness (RBE) of protons, we also evaluate the implications of using NN predictions for normal tissue complication probability (NTCP) models within a variable-RBE context. Approach: The predictive performance of three-dimensional NN architectures was evaluated using five-fold cross-validation on a cohort of brain tumor patients (n=151). The best-performing model was identified and externally validated on patients from a different center (n=107). LETdpredictions were compared to MC-simulated results in clinically relevant regions of interest. We assessed the impact on NTCP models by leveraging LETdpredictions to derive RBE-weighted doses, using the Wedenberg RBE model. Main results: We found NNs based solely on the planned dose profile, i.e. without additional usage of CT images, can approximate MC-based LETddistributions. Root mean squared errors (RMSE) for the median LETdwithin the brain, brainstem, CTV, chiasm, lacrimal glands (ipsilateral/contralateral) and optic nerves (ipsilateral/contralateral) were 0.36, 0.87, 0.31, 0.73, 0.68, 1.04, 0.69 and 1.24~keV/µm, respectively. Although model predictions showed statistically significant differences from MC outputs, these did not result in substantial changes in NTCP predictions, with RMSEs of at most 3.2 percentage points. Significance: The ability of NNs to predict LETdbased solely on planned dose profiles suggests a viable alternative to the compute-intensive MC simulations in a variable-RBE setting. This is particularly useful in scenarios where MC simulation data are unavailable, facilitating resource-constrained proton therapy treatment planning, retrospective patient data analysis and further investigations on the variability of proton RBE.

Efficient determination of Born-effective charges, LO-TO splitting, and Raman tensors of solids with a real-space atom-centered deep learning approach.

We introduce a deep neural network (DNN) framework called the Real-space Atomic Decomposition NETwork (RADNET), which is capable of making accurate predictions of polarization and of electronic dielectric permittivity tensors in solids. This framework builds on previous, atom-centered approaches while utilizing deep convolutional neural networks. We report excellent accuracies on direct predictions for two prototypical examples: GaAs and BN. We then use automatic differentiation to calculate the Born-effective charges, longitudinal optical-transverse optical (LO-TO) splitting frequencies, and Raman tensors of these materials. We compute the Raman spectra, and find agreement with ab initio results. Lastly, we explore ways to generalize polarization predictions while taking into account periodic boundary conditions and symmetries.

Artificial intelligence in cardiac surgery: A systematic review.

BACKGROUND: Artificial intelligence (AI) has emerged as a tool to potentially increase the efficiency and efficacy of cardiovascular care and improve clinical outcomes. This study aims to provide an overview of applications of AI in cardiac surgery. METHODS: A systematic literature search on AI applications in cardiac surgery from inception to February 2024 was conducted. Articles were then filtered based on the inclusion and exclusion criteria and the risk of bias was assessed. Key findings were then summarized. RESULTS: A total of 81 studies were found that reported on AI applications in cardiac surgery. There is a rapid rise in studies since 2020. The most popular machine learning technique was random forest (n = 48), followed by support vector machine (n = 33), logistic regression (n = 32), and eXtreme Gradient Boosting (n = 31). Most of the studies were on adult patients, conducted in China, and involved procedures such as valvular surgery (24.7%), heart transplant (9.4%), coronary revascularization (11.8%), congenital heart disease surgery (3.5%), and aortic dissection repair (2.4%). Regarding evaluation outcomes, 35 studies examined the performance, 26 studies examined clinician outcomes, and 20 studies examined patient outcomes. CONCLUSION: AI was mainly used to predict complications following cardiac surgeries and improve clinicians' decision-making by providing better preoperative risk assessment, stratification, and prognostication. While the application of AI in cardiac surgery has greatly progressed in the last decade, further studies need to be conducted to verify accuracy and ensure safety before use in clinical practice.

Diagnostic Accuracy of Ultra-Low Dose CT Compared to Standard Dose CT for Identification of Fresh Rib Fractures by Deep Learning Algorithm.

The present study aimed to evaluate the diagnostic accuracy of ultra-low dose computed tomography (ULD-CT) compared to standard dose computed tomography (SD-CT) in discerning recent rib fractures using a deep learning algorithm detection of rib fractures (DLADRF). A total of 158 patients undergoing forensic diagnosis for rib fractures were included in this study: 50 underwent SD-CT, and 108 were assessed using ULD-CT. Junior and senior radiologists independently evaluated the images to identify and characterize the rib fractures. The sensitivity of rib fracture diagnosis by radiologists and radiologist + DLADRF was better using SD-CT than ULD-CT. However, the diagnosis sensitivity of DLADRF using ULD-CT alone was slightly more than SD-CT. Nonetheless, no substantial differences were observed in specificity, positive predictive value, and negative predictive value between SD-CT and ULD-CT by the same radiologist, radiologist + DLADRF, and DLADRF (P > 0.05). The area under the curve (AUC) of receiver operating characteristic indicated that senior radiologist + DLADRF was significantly better than senior and junior radiologists, junior radiologists + DLADRF, and DLADRF alone using SD-CT or ULD-CT (all P < 0.05). Also, junior radiologists + DLADRF was better with ULD-CT than senior and junior radiologists (P < 0.05). The AUC of the rib fracture diagnosed by senior radiologists did not differ from DLADRF using ULD-CT. Also, no significant differences were observed between junior + AI and senior and between junior and DLADRF using SD-CT. DLADRF enhanced the diagnostic performance of radiologists in detecting recent rib fractures. The diagnostic outcomes between SD-CT and ULD-CT across radiologists' experience and DLADRF did not differ significantly.

Fully and Weakly Supervised Deep Learning for Meniscal Injury Classification, and Location Based on MRI.

Meniscal injury is a common cause of knee joint pain and a precursor to knee osteoarthritis (KOA). The purpose of this study is to develop an automatic pipeline for meniscal injury classification and localization using fully and weakly supervised networks based on MRI images. In this retrospective study, data were from the osteoarthritis initiative (OAI). The MR images were reconstructed using a sagittal intermediate-weighted fat-suppressed turbo spin-echo sequence. (1) We used 130 knees from the OAI to develop the LGSA-UNet model which fuses the features of adjacent slices and adjusts the blocks in Siam to enable the central slice to obtain rich contextual information. (2) One thousand seven hundred and fifty-six knees from the OAI were included to establish segmentation and classification models. The segmentation model achieved a DICE coefficient ranging from 0.84 to 0.93. The AUC values ranged from 0.85 to 0.95 in the binary models. The accuracy for the three types of menisci (normal, tear, and maceration) ranged from 0.60 to 0.88. Furthermore, 206 knees from the orthopedic hospital were used as an external validation data set to evaluate the performance of the model. The segmentation and classification models still performed well on the external validation set. To compare the diagnostic performances between the deep learning (DL) models and radiologists, the external validation sets were sent to two radiologists. The binary classification model outperformed the diagnostic performance of the junior radiologist (0.82-0.87 versus 0.74-0.88). This study highlights the potential of DL in knee meniscus segmentation and injury classification which can help improve diagnostic efficiency.

Bifurcation detection in intravascular optical coherence tomography using vision transformer based deep learning.

Objective. Bifurcation detection in intravascular optical coherence tomography (IVOCT) images plays a significant role in guiding optimal revascularization strategies for percutaneous coronary intervention (PCI). We propose a bifurcation detection method using vision transformer (ViT) based deep learning in IVOCT.Approach. Instead of relying on lumen segmentation, the proposed method identifies the bifurcation image using a ViT-based classification model and then estimate bifurcation ostium points by a ViT-based landmark detection model.Main results. By processing 8640 clinical images, the Accuracy and F1-score of bifurcation identification by the proposed ViT-based model are 2.54% and 16.08% higher than that of traditional non-deep learning methods, are similar to the best performance of convolutional neural networks (CNNs) based methods, respectively. The ostium distance error of the ViT-based model is 0.305 mm, which is reduced 68.5% compared with the traditional non-deep learning method and reduced 24.81% compared with the best performance of CNNs based methods. The results also show that the proposed ViT-based method achieves the highest success detection rate are 11.3% and 29.2% higher than the non-deep learning method, and 4.6% and 2.5% higher than the best performance of CNNs based methods when the distance section is 0.1 and 0.2 mm, respectively.Significance. The proposed ViT-based method enhances the performance of bifurcation detection of IVOCT images, which maintains a high correlation and consistency between the automatic detection results and the expert manual results. It is of great significance in guiding the selection of PCI treatment strategies.

An efficient learning based approach for automatic record deduplication with benchmark datasets.

With technological innovations, enterprises in the real world are managing every iota of data as it can be mined to derive business intelligence (BI). However, when data comes from multiple sources, it may result in duplicate records. As data is given paramount importance, it is also significant to eliminate duplicate entities towards data integration, performance and resource optimization. To realize reliable systems for record deduplication, late, deep learning could offer exciting provisions with a learning-based approach. Deep ER is one of the deep learning-based methods used recently for dealing with the elimination of duplicates in structured data. Using it as a reference model, in this paper, we propose a framework known as Enhanced Deep Learning-based Record Deduplication (EDL-RD) for improving performance further. Towards this end, we exploited a variant of Long Short Term Memory (LSTM) along with various attribute compositions, similarity metrics, and numerical and null value resolution. We proposed an algorithm known as Efficient Learning based Record Deduplication (ELbRD). The algorithm extends the reference model with the aforementioned enhancements. An empirical study has revealed that the proposed framework with extensions outperforms existing methods.

Detection and severity assessment of obstructive sleep apnea according to deep learning of single-lead electrocardiogram signals.

Developing a convenient detection method is important for diagnosing and treating obstructive sleep apnea. Considering availability and medical reliability, we established a deep-learning model that uses single-lead electrocardiogram signals for obstructive sleep apnea detection and severity assessment. The detection model consisted of signal preprocessing, feature extraction, time-frequency domain information fusion, and classification segments. A total of 375 patients who underwent polysomnography were included. The single-lead electrocardiogram signals obtained by polysomnography were used to train, validate and test the model. Moreover, the proposed model performance on a public dataset was compared with the findings of previous studies. In the test set, the accuracy of per-segment and per-recording detection were 82.55% and 85.33%, respectively. The accuracy values for mild, moderate and severe obstructive sleep apnea were 69.33%, 74.67% and 85.33%, respectively. In the public dataset, the accuracy of per-segment detection was 91.66%. A Bland-Altman plot revealed the consistency of true apnea-hypopnea index and predicted apnea-hypopnea index. We confirmed the feasibility of single-lead electrocardiogram signals and deep-learning model for obstructive sleep apnea detection and severity evaluation in both hospital and public datasets. The detection performance is high for patients with obstructive sleep apnea, especially those with severe obstructive sleep apnea.

Causality-inspired crop pest recognition based on Decoupled Feature Learning.

BACKGROUND: Ensuring the efficient recognition and management of crop pests is crucial for maintaining the balance in global agricultural ecosystems and ecological harmony. Deep learning-based methods have shown promise in crop pest recognition. However, prevailing methods often fail to address a critical issue: biased pest training dataset distribution stemming from the tendency to collect images primarily in certain environmental contexts, such as paddy fields. This oversight hampers recognition accuracy when encountering pest images dissimilar to training samples, highlighting the need for a novel approach to overcome this limitation. RESULTS: We introduce the Decoupled Feature Learning (DFL) framework, leveraging causal inference techniques to handle training dataset bias. DFL manipulates the training data based on classification confidence to construct different training domains and employs center triplet loss for learning class-core features. The proposed DFL framework significantly boosts existing baseline models, attaining unprecedented recognition accuracies of 95.33%, 92.59%, and 74.86% on the Li, DFSPD, and IP102 datasets, respectively. CONCLUSION: Extensive testing on three pest datasets using standard baseline models demonstrates the superiority of DFL in pest recognition. The visualization results show that DFL encourages the baseline models to capture the class-core features. The proposed DFL marks a pivotal step in mitigating the issue of data distribution bias, enhancing the reliability of deep learning in agriculture. © 2024 Society of Chemical Industry.

Innovative approaches for coronary heart disease management: integrating biomedical sensors, deep learning, and stellate ganglion modulation.

Coronary heart disease (CHD) is a significant global health concern, necessitating continuous advancements in treatment modalities to improve patient outcomes. Traditional Chinese medicine (TCM) offers alternative therapeutic approaches, but integration with modern biomedical technologies remains relatively unexplored. This study aimed to assess the efficacy of a combined treatment approach for CHD, integrating traditional Chinese medicinal interventions with modern biomedical sensors and stellate ganglion modulation. The objective was to evaluate the impact of this combined treatment on symptom relief, clinical outcomes, hemorheological indicators, and inflammatory biomarkers. A randomized controlled trial was conducted on 117 CHD patients with phlegm-turbidity congestion and excessiveness type. Patients were divided into a combined treatment group (CTG) and a traditional Chinese medicinal group (CMG). The CTG group received a combination of herbal decoctions, thread-embedding therapy, and stellate ganglion modulation, while the CMG group only received traditional herbal decoctions. The CTG demonstrated superior outcomes compared to the CMG across multiple parameters. Significant reductions in TCM symptom scores, improved clinical effects, reduced angina manifestation, favorable changes in hemorheological indicators, and decreased serum inflammatory biomarkers were observed in the CTG post-intervention. The combination of traditional Chinese medicinal interventions with modern biomedical sensors and stellate ganglion modulation has shown promising results in improving symptoms, clinical outcomes, and inflammatory markers in CHD patients. This holistic approach enhances treatment efficacy and patient outcomes. Further research and advancements in sensor technology are needed to optimize this approach.

Pre-processing visual scenes for retinal prosthesis systems: A comprehensive review.

BACKGROUND: Retinal prostheses offer hope for individuals with degenerative retinal diseases by stimulating the remaining retinal cells to partially restore their vision. This review delves into the current advancements in retinal prosthesis technology, with a special emphasis on the pivotal role that image processing and machine learning techniques play in this evolution. METHODS: We provide a comprehensive analysis of the existing implantable devices and optogenetic strategies, delineating their advantages, limitations, and challenges in addressing complex visual tasks. The review extends to various image processing algorithms and deep learning architectures that have been implemented to enhance the functionality of retinal prosthetic devices. We also illustrate the testing results by demonstrating the clinical trials or using Simulated Prosthetic Vision (SPV) through phosphene simulations, which is a critical aspect of simulating visual perception for retinal prosthesis users. RESULTS: Our review highlights the significant progress in retinal prosthesis technology, particularly its capacity to augment visual perception among the visually impaired. It discusses the integration between image processing and deep learning, illustrating their impact on individual interactions and navigations within the environment through applying clinical trials and also illustrating the limitations of some techniques to be used with current devices, as some approaches only use simulation even on sighted-normal individuals or rely on qualitative analysis, where some consider realistic perception models and others do not. CONCLUSION: This interdisciplinary field holds promise for the future of retinal prostheses, with the potential to significantly enhance the quality of life for individuals with retinal prostheses. Future research directions should pivot towards optimizing phosphene simulations for SPV approaches, considering the distorted and confusing nature of phosphene perception, thereby enriching the visual perception provided by these prosthetic devices. This endeavor will not only improve navigational independence but also facilitate a more immersive interaction with the environment.

EMG-YOLO: road crack detection algorithm for edge computing devices.

Introduction: Road cracks significantly shorten the service life of roads. Manual detection methods are inefficient and costly. The YOLOv5 model has made some progress in road crack detection. However, issues arise when deployed on edge computing devices. The main problem is that edge computing devices are directly connected to sensors. This results in the collection of noisy, poor-quality data. This problem adds computational burden to the model, potentially impacting its accuracy. To address these issues, this paper proposes a novel road crack detection algorithm named EMG-YOLO. Methods: First, an Efficient Decoupled Header is introduced in YOLOv5 to optimize the head structure. This approach separates the classification task from the localization task. Each task can then focus on learning its most relevant features. This significantly reduces the model's computational resources and time. It also achieves faster convergence rates. Second, the IOU loss function in the model is upgraded to the MPDIOU loss function. This function works by minimizing the top-left and bottom-right point distances between the predicted bounding box and the actual labeled bounding box. The MPDIOU loss function addresses the complex computation and high computational burden of the current YOLOv5 model. Finally, the GCC3 module replaces the traditional convolution. It performs global context modeling with the input feature map to obtain global context information. This enhances the model's detection capabilities on edge computing devices. Results: Experimental results show that the improved model has better performance in all parameter indicators compared to current mainstream algorithms. The EMG-YOLO model improves the accuracy of the YOLOv5 model by 2.7%. The mAP (0.5) and mAP (0.9) are improved by 2.9% and 0.9%, respectively. The new algorithm also outperforms the YOLOv5 model in complex environments on edge computing devices. Discussion: The EMG-YOLO algorithm proposed in this paper effectively addresses the issues of poor data quality and high computational burden on edge computing devices. This is achieved through optimizing the model head structure, upgrading the loss function, and introducing global context modeling. Experimental results demonstrate significant improvements in both accuracy and efficiency, especially in complex environments. Future research can further optimize this algorithm and explore more lightweight and efficient object detection models for edge computing devices.

CTHNet: a network for wheat ear counting with local-global features fusion based on hybrid architecture.

Accurate wheat ear counting is one of the key indicators for wheat phenotyping. Convolutional neural network (CNN) algorithms for counting wheat have evolved into sophisticated tools, however because of the limitations of sensory fields, CNN is unable to simulate global context information, which has an impact on counting performance. In this study, we present a hybrid attention network (CTHNet) for wheat ear counting from RGB images that combines local features and global context information. On the one hand, to extract multi-scale local features, a convolutional neural network is built using the Cross Stage Partial framework. On the other hand, to acquire better global context information, tokenized image patches from convolutional neural network feature maps are encoded as input sequences using Pyramid Pooling Transformer. Then, the feature fusion module merges the local features with the global context information to significantly enhance the feature representation. The Global Wheat Head Detection Dataset and Wheat Ear Detection Dataset are used to assess the proposed model. There were 3.40 and 5.21 average absolute errors, respectively. The performance of the proposed model was significantly better than previous studies.

Anatomy-aware and acquisition-agnostic joint registration with SynthMorph.

Affine image registration is a cornerstone of medical-image analysis. While classical algorithms can achieve excellent accuracy, they solve a time-consuming optimization for every image pair. Deep-learning (DL) methods learn a function that maps an image pair to an output transform. Evaluating the function is fast, but capturing large transforms can be challenging, and networks tend to struggle if a test-image characteristic shifts from the training domain, such as the resolution. Most affine methods are agnostic to the anatomy the user wishes to align, meaning the registration will be inaccurate if algorithms consider all structures in the image. We address these shortcomings with SynthMorph, a fast, symmetric, diffeomorphic, and easy-to-use DL tool for joint affine-deformable registration of any brain image without preprocessing. First, we leverage a strategy that trains networks with widely varying images synthesized from label maps, yielding robust performance across acquisition specifics unseen at training. Second, we optimize the spatial overlap of select anatomical labels. This enables networks to distinguish anatomy of interest from irrelevant structures, removing the need for preprocessing that excludes content which would impinge on anatomy-specific registration. Third, we combine the affine model with a deformable hypernetwork that lets users choose the optimal deformation-field regularity for their specific data, at registration time, in a fraction of the time required by classical methods. This framework is applicable to learning anatomy-aware, acquisition-agnostic registration of any anatomy with any architecture, as long as label maps are available for training. We analyze how competing architectures learn affine transforms and compare state-of-the-art registration tools across an extremely diverse set of neuroimaging data, aiming to truly capture the behavior of methods in the real world. SynthMorph demonstrates high accuracy and is available at https://w3id.org/synthmorph, as a single complete end-to-end solution for registration of brain magnetic resonance imaging (MRI) data.