Phytopathology. 2024 Jun 3. doi: 10.1094/PHYTO-02-24-0056-KC. Online ahead of print.

ABSTRACT

Net blotch disease caused by Drechslera teres is a major fungal disease that affects barley (Hordeum vulgare) plants and can result in significant crop losses. In this study, we developed a deep-learning model to quantify net blotch disease symptoms on different days post-infection on seedling leaves using Cascade R-CNN (Region-Based Convolutional Neural Networks) and U-Net (a convolutional neural network) architectures. We used a dataset of barley leaf images with annotations of net blotch disease to train and evaluate the model. The model achieved an accuracy of 95% for cascade R-CNN in net blotch disease detection and a Jaccard index score of 0.99, indicating high accuracy in disease quantification and location. The combination of Cascade R-CNN and U-Net architectures improved the detection of small and irregularly shaped lesions in the images at 4-days post infection, leading to better disease quantification. To validate the model developed, we compared the results obtained by automated measurement with a classical method (necrosis diameter measurement) and a pathogen detection by real-time PCR. The proposed deep learning model could be used in automated systems for disease quantification and to screen the efficacy of potential biocontrol agents to protect against disease.

PMID:38831567 | DOI:10.1094/PHYTO-02-24-0056-KC

BMJ Health Care Inform. 2024 Jun 3;31(1):e100942. doi: 10.1136/bmjhci-2023-100942.

ABSTRACT

BACKGROUND: Current approaches for initial coronary artery disease (CAD) assessment rely on pretest probability (PTP) based on risk factors and presentations, with limited performance. Infrared thermography (IRT), a non-contact technology that detects surface temperature, has shown potential in assessing atherosclerosis-related conditions, particularly when measured from body regions such as faces. We aim to assess the feasibility of using facial IRT temperature information with machine learning for the prediction of CAD.

METHODS: Individuals referred for invasive coronary angiography or coronary CT angiography (CCTA) were enrolled. Facial IRT images captured before confirmatory CAD examinations were used to develop and validate a deep-learning IRT image model for detecting CAD. We compared the performance of the IRT image model with the guideline-recommended PTP model on the area under the curve (AUC). In addition, interpretable IRT tabular features were extracted from IRT images to further validate the predictive value of IRT information.

RESULTS: A total of 460 eligible participants (mean (SD) age, 58.4 (10.4) years; 126 (27.4%) female) were included. The IRT image model demonstrated outstanding performance (AUC 0.804, 95% CI 0.785 to 0.823) compared with the PTP models (AUC 0.713, 95% CI 0.691 to 0.734). A consistent level of superior performance (AUC 0.796, 95% CI 0.782 to 0.811), achieved with comprehensive interpretable IRT features, further validated the predictive value of IRT information. Notably, even with only traditional temperature features, a satisfactory performance (AUC 0.786, 95% CI 0.769 to 0.803) was still upheld.

CONCLUSION: In this prospective study, we demonstrated the feasibility of using non-contact facial IRT information for CAD prediction.

PMID:38830766 | DOI:10.1136/bmjhci-2023-100942

Mod Pathol. 2024 Jun 1:100531. doi: 10.1016/j.modpat.2024.100531. Online ahead of print.

ABSTRACT

Histopathological assessment of esophageal biopsies is a key part in the management of patients with Barrett's esophagus (BE) but prone to observer variability and reliable diagnostic methods are needed. Artificial intelligence (AI) is emerging as a powerful tool for aided diagnosis but often relies on abstract test and validation sets while real world behavior is unknown. In this study, we developed a two-stage AI system for histopathological assessment of BE-related dysplasia using deep learning to enhance the efficiency and accuracy of the pathology workflow. The AI system was developed and trained on 290 whole slide images (WSIs) that were annotated at glandular and tissue level. The system was designed to identify individual glands, grade dysplasia, and assign a WSI-level diagnosis. The proposed method is evaluated by comparing the performance of our AI system to a large international and heterogeneous group of 55 GI pathologists assessing 55 digitized biopsies spanning the complete spectrum of BE-related dysplasia. The AI system correctly graded 76.4% of the WSIs, surpassing the performance of 53 out of the 55 participating pathologists. Furthermore, the ROC analysis showed that the system's ability to predict the absence (non-dysplastic BE) versus the presence of any dysplasia with an AUC of 0.94 and a sensitivity of 0.92 at a specificity of 0.94. These findings demonstrate that this AI system has the potential to assist pathologists in assessment of BE-related dysplasia. The system's outputs could provide a reliable and consistent secondary diagnosis in challenging cases or be used for triaging low-risk non-dysplastic biopsies, thereby reducing the workload of pathologists, and increasing throughput.

PMID:38830407 | DOI:10.1016/j.modpat.2024.100531

Med Image Anal. 2024 May 22;96:103212. doi: 10.1016/j.media.2024.103212. Online ahead of print.

ABSTRACT

Deformable image registration is an essential component of medical image analysis and plays an irreplaceable role in clinical practice. In recent years, deep learning-based registration methods have demonstrated significant improvements in convenience, robustness and execution time compared to traditional algorithms. However, registering images with large displacements, such as those of the liver organ, remains underexplored and challenging. In this study, we present a novel convolutional neural network (CNN)-based unsupervised learning registration method, Cascaded Multi-scale Spatial-Channel Attention-guided Network (CMAN), which addresses the challenge of large deformation fields using a double coarse-to-fine registration approach. The main contributions of CMAN include: (i) local coarse-to-fine registration in the base network, which generates the displacement field for each resolution and progressively propagates these local deformations as auxiliary information for the final deformation field; (ii) global coarse-to-fine registration, which stacks multiple base networks for sequential warping, thereby incorporating richer multi-layer contextual details into the final deformation field; (iii) integration of the spatial-channel attention module in the decoder stage, which better highlights important features and improves the quality of feature maps. The proposed network was trained using two public datasets and evaluated on another public dataset as well as a private dataset across several experimental scenarios. We compared CMAN with four state-of-the-art CNN-based registration methods and two well-known traditional algorithms. The results show that the proposed double coarse-to-fine registration strategy outperforms other methods in most registration evaluation metrics. In conclusion, CMAN can effectively handle the large-deformation registration problem and show potential for application in clinical practice. The source code is made publicly available at https://github.com/LocPham263/CMAN.git.

PMID:38830326 | DOI:10.1016/j.media.2024.103212

BMC Med Inform Decis Mak. 2024 Jun 3;24(1):150. doi: 10.1186/s12911-024-02562-8.

NO ABSTRACT

PMID:38831373 | DOI:10.1186/s12911-024-02562-8

BMC Chem. 2024 Jun 3;18(1):108. doi: 10.1186/s13065-024-01219-x.

ABSTRACT

Determination of protein-ligand binding affinity (PLA) is a key technological tool in hit discovery and lead optimization, which is critical to the drug development process. PLA can be determined directly by experimental methods, but it is time-consuming and costly. In recent years, deep learning has been widely applied to PLA prediction, the key of which lies in the comprehensive and accurate representation of proteins and ligands. In this study, we proposed a multi-modal deep learning model based on the early fusion strategy, called DeepLIP, to improve PLA prediction by integrating multi-level information, and further used it for virtual screening of extracellular signal-regulated protein kinase 2 (ERK2), an ideal target for cancer treatment. Experimental results from model evaluation showed that DeepLIP achieved superior performance compared to state-of-the-art methods on the widely used benchmark dataset. In addition, by combining previously developed machine learning models and molecular dynamics simulation, we screened three novel hits from a drug-like natural product library. These compounds not only had favorable physicochemical properties, but also bound stably to the target protein. We believe they have the potential to serve as starting molecules for the development of ERK2 inhibitors.

PMID:38831341 | DOI:10.1186/s13065-024-01219-x

Breast Cancer Res. 2024 Jun 3;26(1):90. doi: 10.1186/s13058-024-01840-7.

ABSTRACT

BACKGROUND: Nottingham histological grade (NHG) is a well established prognostic factor in breast cancer histopathology but has a high inter-assessor variability with many tumours being classified as intermediate grade, NHG2. Here, we evaluate if DeepGrade, a previously developed model for risk stratification of resected tumour specimens, could be applied to risk-stratify tumour biopsy specimens.

METHODS: A total of 11,955,755 tiles from 1169 whole slide images of preoperative biopsies from 896 patients diagnosed with breast cancer in Stockholm, Sweden, were included. DeepGrade, a deep convolutional neural network model, was applied for the prediction of low- and high-risk tumours. It was evaluated against clinically assigned grades NHG1 and NHG3 on the biopsy specimen but also against the grades assigned to the corresponding resection specimen using area under the operating curve (AUC). The prognostic value of the DeepGrade model in the biopsy setting was evaluated using time-to-event analysis.

RESULTS: Based on preoperative biopsy images, the DeepGrade model predicted resected tumour cases of clinical grades NHG1 and NHG3 with an AUC of 0.908 (95% CI: 0.88; 0.93). Furthermore, out of the 432 resected clinically-assigned NHG2 tumours, 281 (65%) were classified as DeepGrade-low and 151 (35%) as DeepGrade-high. Using a multivariable Cox proportional hazards model the hazard ratio between DeepGrade low- and high-risk groups was estimated as 2.01 (95% CI: 1.06; 3.79).

CONCLUSIONS: DeepGrade provided prediction of tumour grades NHG1 and NHG3 on the resection specimen using only the biopsy specimen. The results demonstrate that the DeepGrade model can provide decision support to identify high-risk tumours based on preoperative biopsies, thus improving early treatment decisions.

PMID:38831336 | DOI:10.1186/s13058-024-01840-7

Sci Rep. 2024 Jun 3;14(1):12666. doi: 10.1038/s41598-024-63419-8.

ABSTRACT

In the paper, a new evolutionary technique called Linear Matrix Genetic Programming (LMGP) is proposed. It is a matrix extension of Linear Genetic Programming and its application is data-driven black-box control-oriented modeling in conditions of limited access to training data. In LMGP, the model is in the form of an evolutionarily-shaped program which is a sequence of matrix operations. Since the program has a hidden state, running it for a sequence of input data has a similar effect to using well-known recurrent neural networks such as Long Short-Term Memory (LSTM) or Gated Recurrent Unit (GRU). To verify the effectiveness of the LMGP, it was compared with different types of neural networks. The task of all the compared techniques was to reproduce the behavior of a nonlinear model of an underwater vehicle. The results of the comparative tests are reported in the paper and they show that the LMGP can quickly find an effective and very simple solution to the given problem. Moreover, a detailed comparison of models, generated by LMGP and LSTM/GRU, revealed that the former are up to four times more accurate than the latter in reproducing vehicle behavior.

PMID:38831089 | DOI:10.1038/s41598-024-63419-8

Commun Biol. 2024 Jun 3;7(1):679. doi: 10.1038/s42003-024-06332-0.

ABSTRACT

Proteins and nucleic-acids are essential components of living organisms that interact in critical cellular processes. Accurate prediction of nucleic acid-binding residues in proteins can contribute to a better understanding of protein function. However, the discrepancy between protein sequence information and obtained structural and functional data renders most current computational models ineffective. Therefore, it is vital to design computational models based on protein sequence information to identify nucleic acid binding sites in proteins. Here, we implement an ensemble deep learning model-based nucleic-acid-binding residues on proteins identification method, called SOFB, which characterizes protein sequences by learning the semantics of biological dynamics contexts, and then develop an ensemble deep learning-based sequence network to learn feature representation and classification by explicitly modeling dynamic semantic information. Among them, the language learning model, which is constructed from natural language to biological language, captures the underlying relationships of protein sequences, and the ensemble deep learning-based sequence network consisting of different convolutional layers together with Bi-LSTM refines various features for optimal performance. Meanwhile, to address the imbalanced issue, we adopt ensemble learning to train multiple models and then incorporate them. Our experimental results on several DNA/RNA nucleic-acid-binding residue datasets demonstrate that our proposed model outperforms other state-of-the-art methods. In addition, we conduct an interpretability analysis of the identified nucleic acid binding residue sequences based on the attention weights of the language learning model, revealing novel insights into the dynamic semantic information that supports the identified nucleic acid binding residues. SOFB is available at https://github.com/Encryptional/SOFB and https://figshare.com/articles/online_resource/SOFB_figshare_rar/25499452 .

PMID:38830995 | DOI:10.1038/s42003-024-06332-0

Sci Rep. 2024 Jun 3;14(1):12699. doi: 10.1038/s41598-024-63538-2.

ABSTRACT

Medical image segmentation has made a significant contribution towards delivering affordable healthcare by facilitating the automatic identification of anatomical structures and other regions of interest. Although convolution neural networks have become prominent in the field of medical image segmentation, they suffer from certain limitations. In this study, we present a reliable framework for producing performant outcomes for the segmentation of pathological structures of 2D medical images. Our framework consists of a novel deep learning architecture, called deep multi-level attention dilated residual neural network (MADR-Net), designed to improve the performance of medical image segmentation. MADR-Net uses a U-Net encoder/decoder backbone in combination with multi-level residual blocks and atrous pyramid scene parsing pooling. To improve the segmentation results, channel-spatial attention blocks were added in the skip connection to capture both the global and local features and superseded the bottleneck layer with an ASPP block. Furthermore, we introduce a hybrid loss function that has an excellent convergence property and enhances the performance of the medical image segmentation task. We extensively validated the proposed MADR-Net on four typical yet challenging medical image segmentation tasks: (1) Left ventricle, left atrium, and myocardial wall segmentation from Echocardiogram images in the CAMUS dataset, (2) Skin cancer segmentation from dermoscopy images in ISIC 2017 dataset, (3) Electron microscopy in FIB-SEM dataset, and (4) Fluid attenuated inversion recovery abnormality from MR images in LGG segmentation dataset. The proposed algorithm yielded significant results when compared to state-of-the-art architectures such as U-Net, Residual U-Net, and Attention U-Net. The proposed MADR-Net consistently outperformed the classical U-Net by 5.43%, 3.43%, and 3.92% relative improvement in terms of dice coefficient, respectively, for electron microscopy, dermoscopy, and MRI. The experimental results demonstrate superior performance on single and multi-class datasets and that the proposed MADR-Net can be utilized as a baseline for the assessment of cross-dataset and segmentation tasks.

PMID:38830932 | DOI:10.1038/s41598-024-63538-2

Med Phys. 2024 Jun 3. doi: 10.1002/mp.17238. Online ahead of print.

ABSTRACT

BACKGROUND: Direction Modulated Brachytherapy (DMBT) enables conformal dose distributions. However, clinicians may face challenges in creating viable treatment plans within a fast-paced clinical setting, especially for a novel technology like DMBT, where cumulative clinical experience is limited. Deep learning-based dose prediction methods have emerged as effective tools for enhancing efficiency.

PURPOSE: To develop a voxel-wise dose prediction model using an attention-gating mechanism and a 3D UNET for cervical cancer high-dose-rate (HDR) brachytherapy treatment planning with DMBT six-groove tandems with ovoids or ring applicators.

METHODS: A multi-institutional cohort of 122 retrospective clinical HDR brachytherapy plans treated to a prescription dose in the range of 4.8-7.0 Gy/fraction was used. A DMBT tandem model was constructed and incorporated onto a research version of BrachyVision Treatment Planning System (BV-TPS) as a 3D solid model applicator and retrospectively re-planned all cases by seasoned experts. Those plans were randomly divided into 64:16:20 as training, validating, and testing cohorts, respectively. Data augmentation was applied to the training and validation sets to increase the size by a factor of 4. An attention-gated 3D UNET architecture model was developed to predict full 3D dose distributions based on high-risk clinical target volume (CTVHR) and organs at risk (OARs) contour information. The model was trained using the mean absolute error loss function, Adam optimization algorithm, a learning rate of 0.001, 250 epochs, and a batch size of eight. In addition, a baseline UNET model was trained similarly for comparison. The model performance was evaluated on the testing dataset by analyzing the outcomes in terms of mean dose values and derived dose-volume-histogram indices from 3D dose distributions and comparing the generated dose distributions against the ground-truth dose distributions using dose statistics and clinically meaningful dosimetric indices.

RESULTS: The proposed attention-gated 3D UNET model showed competitive accuracy in predicting 3D dose distributions that closely resemble the ground-truth dose distributions. The average values of the mean absolute errors were 1.82 ± 29.09 Gy (vs. 6.41 ± 20.16 Gy for a baseline UNET) in CTVHR, 0.89 ± 1.25 Gy (vs. 0.94 ± 3.96 Gy for a baseline UNET) in the bladder, 0.33 ± 0.67 Gy (vs. 0.53 ± 1.66 Gy for a baseline UNET) in the rectum, and 0.55 ± 1.57 Gy (vs. 0.76 ± 2.89 Gy for a baseline UNET) in the sigmoid. The results showed that the mean absolute error (MAE) for the bladder, rectum, and sigmoid were 0.22 ± 1.22 Gy (3.62%) (p = 0.015), 0.21 ± 1.06 Gy (2.20%) (p = 0.172), and -0.03 ± 0.54 Gy (1.13%) (p = 0.774), respectively. The MAE for D90, V100%, and V150% of the CTVHR were 0.46 ± 2.44 Gy (8.14%) (p = 0.018), 0.57 ± 11.25% (5.23%) (p = 0.283), and -0.43 ± 19.36% (4.62%) (p = 0.190), respectively. The proposed model needs less than 5 s to predict a full 3D dose distribution of 64 × 64 × 64 voxels for any new patient plan, thus making it sufficient for near real-time applications and aiding with decision-making in the clinic.

CONCLUSIONS: Attention gated 3D-UNET model demonstrated a capability in predicting voxel-wise dose prediction, in comparison to 3D UNET, for DMBT intracavitary brachytherapy planning. The proposed model could be used to obtain dose distributions for near real-time decision-making before DMBT planning and quality assurance. This will guide future automated planning, making the workflow more efficient and clinically viable.

PMID:38830129 | DOI:10.1002/mp.17238

Rheumatology (Oxford). 2024 Jun 3:keae312. doi: 10.1093/rheumatology/keae312. Online ahead of print.

ABSTRACT

OBJECTIVES: This study aimed to investigate the value of a deep learning (DL) model based on greyscale ultrasound (US) images for precise assessment and accurate diagnosis of primary Sjögren's syndrome (pSS).

METHODS: This was a multicentre prospective analysis. All pSS patients were diagnosed according to 2016 ACR/EULAR criteria. 72 pSS patients and 72 sex- and age-matched healthy controls recruited between January 2022 and April 2023, together with 41 patients and 41 healthy controls recruited from June 2023 to February 2024 were used for DL model development and validation, respectively. DL model was constructed based on the ResNet 50, input with preprocessed all participants' bilateral submandibular glands (SMGs), parotid glands (PGs), and lacrimal glands (LGs) greyscale US images. Diagnostic performance of the model was compared with two radiologists. The accuracy of prediction and identification performance of DL model were evaluated by calibration curve.

RESULTS: 864 and 164 greyscale US images of SMGs, PGs, and LGs were collected for development and validation of the model. The AUCs of DL model in the SMG, PG, and LG were 0.92, 0.93, 0.91 in the model cohort, and were 0.90, 0.88, 0.87 in the validation cohort respectively, outperforming both radiologists. Calibration curves showed the prediction probability of DL model were consistent with the actual probability in both model cohort and validation cohort.

CONCLUSION: DL model based on greyscale US images showed diagnostic potential in the precise assessment of pSS patients in the SMG, PG, and LG, outperforming conventional radiologist evaluation.

PMID:38830044 | DOI:10.1093/rheumatology/keae312

IEEE Trans Biomed Eng. 2024 Jun 3;PP. doi: 10.1109/TBME.2024.3408232. Online ahead of print.

ABSTRACT

OBJECTIVE: This paper aims at quantifying biomarkers from the segmentation of retinal arteries in adaptive optics ophthalmoscopy images (AOO).

METHODS: The segmentation is based on the combination of deep learning and knowledge-driven deformable models to achieve a precise segmentation of the vessel walls, with a specific attention to bifurcations. Biomarkers (junction coefficient, branching coefficient, wall to lumen ratio (wlr) are derived from the resulting segmentation.

RESULTS: reliable and accurate segmentations (mse = 1.75 ± 1.24 pixel) and measurements are obtained, with high reproducibility with respect to images acquisition and users, and without bias.

SIGNIFICANCE: In a preliminary clinical study of patients with a genetic small vessel disease, some of them with vascular risk factors, an increased wlr was found in comparison to a control population.

CONCLUSION: The wlr estimated in AOO images with our method (AOV, Adaptive Optics Vessel analysis) seems to be a very robust biomarker as long as the wall is well contrasted.

PMID:38829761 | DOI:10.1109/TBME.2024.3408232

IEEE Trans Biomed Eng. 2024 Jun 3;PP. doi: 10.1109/TBME.2024.3408331. Online ahead of print.

ABSTRACT

OBJECTIVE: To remove signal contamination in electroencephalogram (EEG) traces coming from ocular, motion, and muscular artifacts which degrade signal quality. To do this in real-time, with low computational overhead, on a mobile platform in a channel count independent manner to enable portable Brain-Computer Interface (BCI) applications.

METHODS: We propose a Deep AutoEncoder (DAE) neural network for single-channel EEG artifact removal, and implement it on a smartphone via TensorFlow Lite. Delegate based acceleration is employed to allow real-time, low computational resource operation. Artifact removal performance is quantified by comparing corrupted and ground-truth clean EEG data from public datasets for a range of artifact types. The on-phone computational resources required are also measured when processing pre-saved data.

RESULTS: DAE cleaned EEG shows high correlations with ground-truth clean EEG, with average Correlation Coefficients of 0.96, 0.85, 0.70 and 0.79 for clean EEG reconstruction, and EOG, motion, and EMG artifact removal respectively. On-smartphone tests show the model processes a 4 s EEG window within 5 ms, substantially outperforming a comparison FastICA artifact removal algorithm.

CONCLUSION: The proposed DAE model shows effectiveness in single-channel EEG artifact removal. This is the first demonstration of a low-computational resource deep learning model for mobile EEG in smartphones with hardware/software acceleration.

SIGNIFICANCE: This work enables portable BCIs which require low latency real-time artifact removal, and potentially operation with a small number of EEG channels for wearability. It makes use of the artificial intelligence accelerators found in modern smartphones to improve computational performance compared to previous artifact removal approaches.

PMID:38829759 | DOI:10.1109/TBME.2024.3408331

IEEE Trans Med Imaging. 2024 Jun 3;PP. doi: 10.1109/TMI.2024.3400899. Online ahead of print.

ABSTRACT

Registering pre-operative modalities, such as magnetic resonance imaging or computed tomography, to ultrasound images is crucial for guiding clinicians during surgeries and biopsies. Recently, deep-learning approaches have been proposed to increase the speed and accuracy of this registration problem. However, all of these approaches need expensive supervision from the ultrasound domain. In this work, we propose a multitask generative framework that needs weak supervision only from the pre-operative imaging domain during training. To perform a deformable registration, the proposed framework translates a magnetic resonance image to the ultrasound domain while preserving the structural content. To demonstrate the efficacy of the proposed method, we tackle the registration problem of pre-operative 3D MR to transrectal ultrasonography images as necessary for targeted prostate biopsies. We use an in-house dataset of 600 patients, divided into 540 for training, 30 for validation, and the remaining for testing. An expert manually segmented the prostate in both modalities for validation and test sets to assess the performance of our framework. The proposed framework achieves a 3.58 mm target registration error on the expert-selected landmarks, 89.2% in the Dice score, and 1.81 mm 95th percentile Hausdorff distance on the prostate masks in the test set. Our experiments demonstrate that the proposed generative model successfully translates magnetic resonance images into the ultrasound domain. The translated image contains the structural content and fine details due to an ultrasound-specific two-path design of the generative model. The proposed framework enables training learning-based registration methods while only weak supervision from the pre-operative domain is available.

PMID:38829753 | DOI:10.1109/TMI.2024.3400899

IEEE Rev Biomed Eng. 2024 Jun 3;PP. doi: 10.1109/RBME.2024.3408456. Online ahead of print.

ABSTRACT

Increasing demands on medical imaging departments are taking a toll on the radiologist's ability to deliver timely and accurate reports. Recent technological advances in artificial intelligence have demonstrated great potential for automatic radiology report generation (ARRG), sparking an explosion of research. This survey paper conducts a methodological review of contemporary ARRG approaches by way of (i) assessing datasets based on characteristics, such as availability, size, and adoption rate, (ii) examining deep learning training methods, such as contrastive learning and reinforcement learning, (iii) exploring state-of-the-art model architectures, including variations of CNN and transformer models, (iv) outlining techniques integrating clinical knowledge through multimodal inputs and knowledge graphs, and (v) scrutinising current model evaluation techniques, including commonly applied NLP metrics and qualitative clinical reviews. Furthermore, the quantitative results of the reviewed models are analysed, where the top performing models are examined to seek further insights. Finally, potential new directions are highlighted, with the adoption of additional datasets from other radiological modalities and improved evaluation methods predicted as important areas of future development.

PMID:38829752 | DOI:10.1109/RBME.2024.3408456

IEEE Trans Ultrason Ferroelectr Freq Control. 2024 Jun 3;PP. doi: 10.1109/TUFFC.2024.3408314. Online ahead of print.

ABSTRACT

In Non-Destructive Evaluation (NDE), accurately characterizing defects within components relies on accurate sizing and localization to evaluate the severity or criticality of defects. This study presents for the first time a deep learning methodology using 3-Dimensional (3D) U-Net to localize and size defects in Carbon Fibre Reinforced Polymer (CFRP) composites through volumetric segmentation of ultrasonic testing data. Using a previously developed approach, synthetic training data closely representative of experimental data was used for the automatic generation of ground truth segmentation masks. The model's performance was compared to the conventional amplitude 6 dB drop analysis method used in industry against ultrasonic defect responses from 40 defects fabricated in CFRP components. The results showed good agreement with the 6 dB drop method for in-plane localization and excellent through-thickness localization, with Mean Absolute Errors (MAE) of 0.57 mm and 0.08 mm, respectively. Initial sizing results consistently oversized defects with a 55% higher mean average error than the 6 dB drop method. However, when a correction factor was applied to account for variation between the experimental and synthetic domains the final sizing accuracy resulted in a 35% reduction in MAE compared to the 6 dB drop technique. By working with volumetric ultrasonic data (as opposed to 2D images) this approach reduces pre-processing (such as signal gating) and allows for the generation of 3D defect masks which can be used for the generation of computer aided design files; greatly reducing the qualification reporting burden of NDE operators.

PMID:38829751 | DOI:10.1109/TUFFC.2024.3408314

Transl Vis Sci Technol. 2024 Jun 3;13(6):1. doi: 10.1167/tvst.13.6.1.

ABSTRACT

PURPOSE: Deep learning architectures can automatically learn complex features and patterns associated with glaucomatous optic neuropathy (GON). However, developing robust algorithms requires a large number of data sets. We sought to train an adversarial model for generating high-quality optic disc images from a large, diverse data set and then assessed the performance of models on generated synthetic images for detecting GON.

METHODS: A total of 17,060 (6874 glaucomatous and 10,186 healthy) fundus images were used to train deep convolutional generative adversarial networks (DCGANs) for synthesizing disc images for both classes. We then trained two models to detect GON, one solely on these synthetic images and another on a mixed data set (synthetic and real clinical images). Both the models were externally validated on a data set not used for training. The multiple classification metrics were evaluated with 95% confidence intervals. Models' decision-making processes were assessed using gradient-weighted class activation mapping (Grad-CAM) techniques.

RESULTS: Following receiver operating characteristic curve analysis, an optimal cup-to-disc ratio threshold for detecting GON from the training data was found to be 0.619. DCGANs generated high-quality synthetic disc images for healthy and glaucomatous eyes. When trained on a mixed data set, the model's area under the receiver operating characteristic curve attained 99.85% on internal validation and 86.45% on external validation. Grad-CAM saliency maps were primarily centered on the optic nerve head, indicating a more precise and clinically relevant attention area of the fundus image.

CONCLUSIONS: Although our model performed well on synthetic data, training on a mixed data set demonstrated better performance and generalization. Integrating synthetic and real clinical images can optimize the performance of a deep learning model in glaucoma detection.

TRANSLATIONAL RELEVANCE: Optimizing deep learning models for glaucoma detection through integrating DCGAN-generated synthetic and real-world clinical data can be improved and generalized in clinical practice.

PMID:38829624 | DOI:10.1167/tvst.13.6.1

J Imaging Inform Med. 2024 Jun 3. doi: 10.1007/s10278-024-01139-1. Online ahead of print.

ABSTRACT

High-resolution (HR) magnetic resonance imaging (MRI) can reveal rich anatomical structures for clinical diagnoses. However, due to hardware and signal-to-noise ratio limitations, MRI images are often collected with low resolution (LR) which is not conducive to diagnosing and analyzing clinical diseases. Recently, deep learning super-resolution (SR) methods have demonstrated great potential in enhancing the resolution of MRI images; however, most of them did not take the cross-modality and internal priors of MR seriously, which hinders the SR performance. In this paper, we propose a cross-modality reference and feature mutual-projection (CRFM) method to enhance the spatial resolution of brain MRI images. Specifically, we feed the gradients of HR MRI images from referenced imaging modality into the SR network to transform true clear textures to LR feature maps. Meanwhile, we design a plug-in feature mutual-projection (FMP) method to capture the cross-scale dependency and cross-modality similarity details of MRI images. Finally, we fuse all feature maps with parallel attentions to produce and refine the HR features adaptively. Extensive experiments on MRI images in the image domain and k-space show that our CRFM method outperforms existing state-of-the-art MRI SR methods.

PMID:38829472 | DOI:10.1007/s10278-024-01139-1

Comput Methods Biomech Biomed Engin. 2024 Jun 3:1-13. doi: 10.1080/10255842.2024.2359635. Online ahead of print.

ABSTRACT

Obstructive sleep apnea (OSA) is a non-communicable sleep-related medical condition marked by repeated disruptions in breathing during sleep. It may induce various cardiovascular and neurocognitive complications. Electrocardiography (ECG) is a useful method for detecting numerous health-related disorders. ECG signals provide a less complex and non-invasive solution for the screening of OSA. Automated and accurate detection of OSA may enhance diagnostic performance and reduce the clinician's workload. Traditional machine learning methods typically involve several labor-intensive manual procedures, including signal decomposition, feature evaluation, selection, and categorization. This article presents the time-frequency (T-F) spectrum classification of de-noised ECG data for the automatic screening of OSA patients using deep convolutional neural networks (DCNNs). At first, a filter-fusion algorithm is used to eliminate the artifacts from the raw ECG data. Stock-well transform (S-T) is employed to change filtered time-domain ECG into T-F spectrums. To discriminate between apnea and normal ECG signals, the obtained T-F spectrums are categorized using benchmark Alex-Net and Squeeze-Net, along with a less complex DCNN. The superiority of the presented system is measured by computing the sensitivity, specificity, accuracy, negative predicted value, precision, F1-score, and Fowlkes-Mallows index. The results of comparing all three utilized DCNNs reveal that the proposed DCNN requires fewer learning parameters and provides higher accuracy. An average accuracy of 95.31% is yielded using the proposed system. The presented deep learning system is lightweight and faster than Alex-Net and Squeeze-Net as it utilizes fewer learnable parameters, making it simple and reliable.

PMID:38829354 | DOI:10.1080/10255842.2024.2359635