Zhonghua Yan Ke Za Zhi. 2024 Mar 11;60(3):257-264. doi: 10.3760/cma.j.cn112142-20231210-00281.

ABSTRACT

Objective: To achieve automatic segmentation, quantification, and grading of different regions of leopard spots fundus (FT) using deep learning technology. The analysis includes exploring the correlation between novel quantitative indicators, leopard spot fundus grades, and various systemic and ocular parameters. Methods: This was a cross-sectional study. The data were sourced from the Beijing Eye Study, a population-based longitudinal study. In 2001, a group of individuals aged 40 and above were surveyed in five urban communities in Haidian District and three rural communities in Daxing District of Beijing. A follow-up was conducted in 2011. This study included individuals aged 50 and above who participated in the second 5-year follow-up in 2011, considering only the data from the right eye. Color fundus images centered on the macula of the right eye were input into the leopard spot segmentation model and macular detection network. Using the macular center as the origin, with inner circle diameters of 1 mm, 3 mm, and outer circle diameter of 6 mm, fine segmentation of the fundus was achieved. This allowed the calculation of the leopard spot density (FTD) and leopard spot grade for each region. Further analyses of the differences in ocular and systemic parameters among different regions' FTD and leopard spot grades were conducted. The participants were categorized into three refractive types based on equivalent spherical power (SE): myopia (SE<-0.25 D), emmetropia (-0.25 D≤SE≤0.25 D), and hyperopia (SE>0.25 D). Based on axial length, the participants were divided into groups with axial length<24 mm, 24-26 mm, and>26 mm for the analysis of different types of FTD. Statistical analyses were performed using one-way analysis of variance, Kruskal-Wallis test, Bonferroni test, and Spearman correlation analysis. Results: The study included 3 369 participants (3 369 eyes) with an average age of (63.9±10.6) years; among them, 1 886 were female (56.0%) and 1, 483 were male (64.0%). The overall FTD for all eyes was 0.060 (0.016, 0.163); inner circle FTD was 0.000 (0.000, 0.025); middle circle FTD was 0.030 (0.000, 0.130); outer circle FTD was 0.055 (0.009, 0.171). The results of the univariate analysis indicated that FTD in various regions was correlated with axial length (overall: r=0.38, P<0.001; inner circle: r=0.31, P<0.001; middle circle: r=0.36, P<0.001; outer circle: r=0.39, P<0.001), subfoveal choroidal thickness (SFCT) (overall: r=-0.69, P<0.001; inner circle: r=-0.57, P<0.001; middle circle: r=-0.68, P<0.001; outer circle: r=-0.72, P<0.001), age (overall: r=0.34, P<0.001; inner circle: r=0.30, P<0.001; middle circle: r=0.31, P<0.001; outer circle: r=0.35, P<0.001), gender (overall: r=-0.11, P<0.001; inner circle: r=-0.04, P<0.001; middle circle: r=-0.07, P<0.001; outer circle: r=-0.11, P<0.001), SE (overall: r=-0.20; P<0.001; inner circle: r=-0.19, P<0.001; middle circle: r=-0.20, P<0.001; outer circle: r=-0.20, P<0.001), uncorrected visual acuity (overall: r=-0.18, P<0.001; inner circle: r=-0.26, P<0.001; middle circle: r=-0.24, P<0.001; outer circle: r=-0.22, P<0.001), and body mass index (BMI) (overall: r=-0.11, P<0.001; inner circle: r=-0.13, P<0.001; middle circle: r=-0.14, P<0.001; outer circle: r=-0.13, P<0.001). Further multivariate analysis results indicated that different region FTD was correlated with axial length (overall: β=0.020, P<0.001; inner circle: β=-0.022, P<0.001; middle circle: β=0.027, P<0.001; outer circle: β=0.022, P<0.001), SFCT (overall: β=-0.001, P<0.001; inner circle: β=-0.001, P<0.001; middle circle: β=-0.001, P<0.001; outer circle: β=-0.001, P<0.001), and age (overall: β=0.002, P<0.001; inner circle: β=0.001, P<0.001; middle circle: β=0.002, P<0.001; outer circle: β=0.002, P<0.001). The distribution of overall (H=56.76, P<0.001), inner circle (H=72.22, P<0.001), middle circle (H=75.83, P<0.001), and outer circle (H=70.34, P<0.001) FTD differed significantly among different refractive types. The distribution of overall (H=373.15, P<0.001), inner circle (H=367.67, P<0.001), middle circle (H=389.14, P<0.001), and outer circle (H=386.89, P<0.001) FTD differed significantly among different axial length groups. Furthermore, comparing various levels of FTD with systemic and ocular parameters, significant differences were found in axial length (F=142.85, P<0.001) and SFCT (F=530.46, P<0.001). Conclusions: The use of deep learning technology enables automatic segmentation and quantification of different regions of theFT, as well as preliminary grading. Different region FTD is significantly correlated with axial length, SFCT, and age. Individuals with older age, myopia, and longer axial length tend to have higher FTD and more advanced FT grades.

PMID:38462374 | DOI:10.3760/cma.j.cn112142-20231210-00281

Anal Chim Acta. 2024 Apr 15;1298:342419. doi: 10.1016/j.aca.2024.342419. Epub 2024 Feb 26.

ABSTRACT

BACKGROUND: As a potential natural active substance, natural biologically active peptides (NBAPs) are recently attracting increasing attention. The traditional proteolysis methods of obtaining effective NBAPs are considerably vexing, especially since multiple proteases can be used, which blocks the exploration of available NBAPs. Although the development of virtual digesting brings some degree of convenience, the activity of the obtained peptides remains unclear, which would still not allow efficient access to the NBAPs. It is necessary to develop an efficient and accurate strategy for acquiring NBAPs.

RESULTS: A new in silico scheme named SSA-LSTM-VD, which combines a sparrow search algorithm-long short-term memory (SSA-LSTM) deep learning and virtually digested, was presented to optimize the proteolysis acquisition of NBAPs. Therein, SSA-LSTM reached the highest Efficiency value reached 98.00 % compared to traditional machine learning algorithms, and basic LSTM algorithm. SSA-LSTM was trained to predict the activity of peptides in the proteins virtually digested results, obtain the percentage of target active peptide, and select the appropriate protease for the actual experiment. As an application, SSA-LSTM was employed to predict the percentage of neuroprotective peptides in the virtual digested result of walnut protein, and trypsin was ultimately found to possess the highest value (85.29 %). The walnut protein was digested by trypsin (WPTrH) and the peptide sequence obtained was analyzed closely matches the theoretical neuroprotective peptide. More importantly, the neuroprotective effects of WPTrH had been demonstrated in nerve damage mouse models.

SIGNIFICANCE: The proposed SSA-LSTM-VD in this paper makes the acquisition of NBAPs efficient and accurate. The approach combines deep learning and virtually digested skillfully. Utilizing the SSA-LSTM-VD based strategy holds promise for discovering and developing peptides with neuroprotective properties or other desired biological activities.

PMID:38462343 | DOI:10.1016/j.aca.2024.342419

Artif Intell Med. 2024 Mar;149:102809. doi: 10.1016/j.artmed.2024.102809. Epub 2024 Feb 9.

ABSTRACT

Cardiovascular diseases, particularly arrhythmias, remain a leading cause of mortality worldwide. Electrocardiogram (ECG) analysis plays a pivotal role in cardiovascular disease diagnosis. Although previous studies have focused on waveform analysis and model training, integrating additional clinical information, especially demographic data, remains challenging. In this study, we present an innovative approach to ECG classification by incorporating demographic information from patients' medical histories through a colorization technique. Our proposed method maps demographic features onto the (R, G, B) color space through normalized scaling. Each demographic feature corresponds to a distinct color, allowing for different ECG leads to be colored. This approach preserves the relationships between data by maintaining the color correlations in the statistical features, enhancing ECG analytics and supporting precision medicine. We conducted experiments with PTB-XL dataset and achieved 1%-6% improvements in the area under the receiving operator characteristic curve performance compared with other methods for various classification problems. Notably, our method excelled in multiclass and challenging classification tasks. The combined use of color features and the original waveform shape features enhanced prediction accuracy for various deep learning models. Our findings suggest that colorization is a promising avenue for advancing ECG classification and diagnosis, contributing to improved prediction and diagnosis of cardiovascular diseases and ultimately enhancing clinical outcomes.

PMID:38462295 | DOI:10.1016/j.artmed.2024.102809

Artif Intell Med. 2024 Mar;149:102803. doi: 10.1016/j.artmed.2024.102803. Epub 2024 Feb 6.

ABSTRACT

Diabetic Retinopathy (DR), an ocular complication of diabetes, is a leading cause of blindness worldwide. Traditionally, DR is monitored using Color Fundus Photography (CFP), a widespread 2-D imaging modality. However, DR classifications based on CFP have poor predictive power, resulting in suboptimal DR management. Optical Coherence Tomography Angiography (OCTA) is a recent 3-D imaging modality offering enhanced structural and functional information (blood flow) with a wider field of view. This paper investigates automatic DR severity assessment using 3-D OCTA. A straightforward solution to this task is a 3-D neural network classifier. However, 3-D architectures have numerous parameters and typically require many training samples. A lighter solution consists in using 2-D neural network classifiers processing 2-D en-face (or frontal) projections and/or 2-D cross-sectional slices. Such an approach mimics the way ophthalmologists analyze OCTA acquisitions: (1) en-face flow maps are often used to detect avascular zones and neovascularization, and (2) cross-sectional slices are commonly analyzed to detect macular edemas, for instance. However, arbitrary data reduction or selection might result in information loss. Two complementary strategies are thus proposed to optimally summarize OCTA volumes with 2-D images: (1) a parametric en-face projection optimized through deep learning and (2) a cross-sectional slice selection process controlled through gradient-based attribution. The full summarization and DR classification pipeline is trained from end to end. The automatic 2-D summary can be displayed in a viewer or printed in a report to support the decision. We show that the proposed 2-D summarization and classification pipeline outperforms direct 3-D classification with the advantage of improved interpretability.

PMID:38462293 | DOI:10.1016/j.artmed.2024.102803

Artif Intell Med. 2024 Mar;149:102798. doi: 10.1016/j.artmed.2024.102798. Epub 2024 Feb 5.

ABSTRACT

The ability to reconstruct high-quality images from undersampled MRI data is vital in improving MRI temporal resolution and reducing acquisition times. Deep learning methods have been proposed for this task, but the lack of verified methods to quantify the uncertainty in the reconstructed images hampered clinical applicability. We introduce "NPB-REC", a non-parametric fully Bayesian framework, for MRI reconstruction from undersampled data with uncertainty estimation. We use Stochastic Gradient Langevin Dynamics during training to characterize the posterior distribution of the network parameters. This enables us to both improve the quality of the reconstructed images and quantify the uncertainty in the reconstructed images. We demonstrate the efficacy of our approach on a multi-coil MRI dataset from the fastMRI challenge and compare it to the baseline End-to-End Variational Network (E2E-VarNet). Our approach outperforms the baseline in terms of reconstruction accuracy by means of PSNR and SSIM (34.55, 0.908 vs. 33.08, 0.897, p<0.01, acceleration rate R=8) and provides uncertainty measures that correlate better with the reconstruction error (Pearson correlation, R=0.94 vs. R=0.91). Additionally, our approach exhibits better generalization capabilities against anatomical distribution shifts (PSNR and SSIM of 32.38, 0.849 vs. 31.63, 0.836, p<0.01, training on brain data, inference on knee data, acceleration rate R=8). NPB-REC has the potential to facilitate the safe utilization of deep learning-based methods for MRI reconstruction from undersampled data. Code and trained models are available at https://github.com/samahkh/NPB-REC.

PMID:38462289 | DOI:10.1016/j.artmed.2024.102798

Artif Intell Med. 2024 Mar;149:102788. doi: 10.1016/j.artmed.2024.102788. Epub 2024 Feb 2.

ABSTRACT

BACKGROUND: Deep learning methods have shown great potential in processing multi-modal Magnetic Resonance Imaging (MRI) data, enabling improved accuracy in brain tumor segmentation. However, the performance of these methods can suffer when dealing with incomplete modalities, which is a common issue in clinical practice. Existing solutions, such as missing modality synthesis, knowledge distillation, and architecture-based methods, suffer from drawbacks such as long training times, high model complexity, and poor scalability.

METHOD: This paper proposes IMS2Trans, a novel lightweight scalable Swin Transformer network by utilizing a single encoder to extract latent feature maps from all available modalities. This unified feature extraction process enables efficient information sharing and fusion among the modalities, resulting in efficiency without compromising segmentation performance even in the presence of missing modalities.

RESULTS: Two datasets, BraTS 2018 and BraTS 2020, containing incomplete modalities for brain tumor segmentation are evaluated against popular benchmarks. On the BraTS 2018 dataset, our model achieved higher average Dice similarity coefficient (DSC) scores for the whole tumor, tumor core, and enhancing tumor regions (86.57, 75.67, and 58.28, respectively), in comparison with a state-of-the-art model, i.e. mmFormer (86.45, 75.51, and 57.79, respectively). Similarly, on the BraTS 2020 dataset, our model scored higher DSC scores in these three brain tumor regions (87.33, 79.09, and 62.11, respectively) compared to mmFormer (86.17, 78.34, and 60.36, respectively). We also conducted a Wilcoxon test on the experimental results, and the generated p-value confirmed that our model's performance was statistically significant. Moreover, our model exhibits significantly reduced complexity with only 4.47 M parameters, 121.89G FLOPs, and a model size of 77.13 MB, whereas mmFormer comprises 34.96 M parameters, 265.79 G FLOPs, and a model size of 559.74 MB. These indicate our model, being light-weighted with significantly reduced parameters, is still able to achieve better performance than a state-of-the-art model.

CONCLUSION: By leveraging a single encoder for processing the available modalities, IMS2Trans offers notable scalability advantages over methods that rely on multiple encoders. This streamlined approach eliminates the need for maintaining separate encoders for each modality, resulting in a lightweight and scalable network architecture. The source code of IMS2Trans and the associated weights are both publicly available at https://github.com/hudscomdz/IMS2Trans.

PMID:38462288 | DOI:10.1016/j.artmed.2024.102788

Artif Intell Med. 2024 Mar;149:102785. doi: 10.1016/j.artmed.2024.102785. Epub 2024 Jan 26.

ABSTRACT

Early detection of acute kidney injury (AKI) may provide a crucial window of opportunity to prevent further injury, which helps improve clinical outcomes. This study aimed to develop a deep interpretable network for continuously predicting the 24-hour AKI risk in real-time and evaluate its performance internally and externally in critically ill patients. A total of 21,163 patients' electronic health records sourced from Beth Israel Deaconess Medical Center (BIDMC) were first included in building the model. Two external validation populations included 3025 patients from the Philips eICU Research Institute and 2625 patients from Zhongda Hospital Southeast University. A total of 152 intelligently engineered predictors were extracted on an hourly basis. The prediction model referred to as DeepAKI was designed with the basic framework of squeeze-and-excitation networks with dilated causal convolution embedded. The integrated gradients method was utilized to explain the prediction model. When performed on the internal validation set (3175 [15 %] patients from BIDMC) and the two external validation sets, DeepAKI obtained the area under the curve of 0.799 (95 % CI 0.791-0.806), 0.763 (95 % CI 0.755-0.771) and 0.676 (95 % CI 0.668-0.684) for continuousAKI prediction, respectively. For model interpretability, clinically relevant important variables contributing to the model prediction were informed, and individual explanations along the timeline were explored to show how AKI risk arose. The potential threats to generalisability in deep learning-based models when deployed across health systems in real-world settings were analyzed.

PMID:38462285 | DOI:10.1016/j.artmed.2024.102785

Artif Intell Med. 2024 Mar;149:102782. doi: 10.1016/j.artmed.2024.102782. Epub 2024 Jan 17.

ABSTRACT

Diabetic retinopathy (DR) is the most prevalent cause of visual impairment in adults worldwide. Typically, patients with DR do not show symptoms until later stages, by which time it may be too late to receive effective treatment. DR Grading is challenging because of the small size and variation in lesion patterns. The key to fine-grained DR grading is to discover more discriminating elements such as cotton wool, hard exudates, hemorrhages, microaneurysms etc. Although deep learning models like convolutional neural networks (CNN) seem ideal for the automated detection of abnormalities in advanced clinical imaging, small-size lesions are very hard to distinguish by using traditional networks. This work proposes a bi-directional spatial and channel-wise parallel attention based network to learn discriminative features for diabetic retinopathy grading. The proposed attention block plugged with a backbone network helps to extract features specific to fine-grained DR-grading. This scheme boosts classification performance along with the detection of small-sized lesion parts. Extensive experiments are performed on four widely used benchmark datasets for DR grading, and performance is evaluated on different quality metrics. Also, for model interpretability, activation maps are generated using the LIME method to visualize the predicted lesion parts. In comparison with state-of-the-art methods, the proposed IDANet exhibits better performance for DR grading and lesion detection.

PMID:38462283 | DOI:10.1016/j.artmed.2024.102782

Artif Intell Med. 2024 Mar;149:102779. doi: 10.1016/j.artmed.2024.102779. Epub 2024 Jan 24.

ABSTRACT

The healthcare sector, characterized by vast datasets and many diseases, is pivotal in shaping community health and overall quality of life. Traditional healthcare methods, often characterized by limitations in disease prevention, predominantly react to illnesses after their onset rather than proactively averting them. The advent of Artificial Intelligence (AI) has ushered in a wave of transformative applications designed to enhance healthcare services, with Machine Learning (ML) as a noteworthy subset of AI. ML empowers computers to analyze extensive datasets, while Deep Learning (DL), a specific ML methodology, excels at extracting meaningful patterns from these data troves. Despite notable technological advancements in recent years, the full potential of these applications within medical contexts remains largely untapped, primarily due to the medical community's cautious stance toward novel technologies. The motivation of this paper lies in recognizing the pivotal role of the healthcare sector in community well-being and the necessity for a shift toward proactive healthcare approaches. To our knowledge, there is a notable absence of a comprehensive published review that delves into ML, DL and distributed systems, all aimed at elevating the Quality of Service (QoS) in healthcare. This study seeks to bridge this gap by presenting a systematic and organized review of prevailing ML, DL, and distributed system algorithms as applied in healthcare settings. Within our work, we outline key challenges that both current and future developers may encounter, with a particular focus on aspects such as approach, data utilization, strategy, and development processes. Our study findings reveal that the Internet of Things (IoT) stands out as the most frequently utilized platform (44.3 %), with disease diagnosis emerging as the predominant healthcare application (47.8 %). Notably, discussions center significantly on the prevention and identification of cardiovascular diseases (29.2 %). The studies under examination employ a diverse range of ML and DL methods, along with distributed systems, with Convolutional Neural Networks (CNNs) being the most commonly used (16.7 %), followed by Long Short-Term Memory (LSTM) networks (14.6 %) and shallow learning networks (12.5 %). In evaluating QoS, the predominant emphasis revolves around the accuracy parameter (80 %). This study highlights how ML, DL, and distributed systems reshape healthcare. It contributes to advancing healthcare quality, bridging the gap between technology and medical adoption, and benefiting practitioners and patients.

PMID:38462281 | DOI:10.1016/j.artmed.2024.102779

Int J Cardiovasc Imaging. 2024 Mar 10. doi: 10.1007/s10554-024-03063-5. Online ahead of print.

ABSTRACT

We evaluated the diagnostic performance of a deep-learning model (DLM) (CorEx®, Spimed-AI, Paris, France) designed to automatically detect > 50% coronary stenosis on coronary computed tomography angiography (CCTA) images. We studied inter-observer variability as an additional aim. CCTA images obtained before transcatheter aortic valve implantation (TAVI) were assessed by two radiologists and the DLM, and the results were compared to those of invasive coronary angiography (ICA) used as the reference standard. 165 consecutive patients underwent both CCTA and ICA as part of their TAVI work-up. We excluded the 42 (25.5%) patients with a history of stenting or bypass grafting and the 23 (13.9%) patients with low-quality images. We retrospectively subjected the CCTA images from the remaining 100 patients to evaluation by the DLM and compared the DLM and ICA results. All 25 patients with > 50% stenosis by ICA also had > 50% stenosis by DLM evaluation of CCTA: thus, the DLM had 100% sensitivity and 100% negative predictive value. False-positive DLM results were common, yielding a positive predictive value of only 39% (95% CI, 27-51%). Two radiologists with 3 and 25 years' experience, respectively, performed similarly to the DLM in evaluating the CCTA images; thus, accuracy did not differ significantly between each reader and the DLM (p = 0.625 and p = 0.375, respectively). The DLM had 100% negative predictive value for > 50% stenosis and performed similarly to experienced radiologists. This tool may hold promise for identifying the up to one-third of patients who do not require ICA before TAVI.

PMID:38461472 | DOI:10.1007/s10554-024-03063-5

Comput Methods Programs Biomed. 2024 Mar 8;247:108109. doi: 10.1016/j.cmpb.2024.108109. Online ahead of print.

ABSTRACT

BACKGROUND AND OBJECTIVE: Automatic needle tip detection is important in real-time ultrasound (US) images that are utilized to guide interventional needle puncture procedures in clinical settings. However, due to the spatial indiscernibility problem caused by the severe background interferences and the tip characteristics of small size, being grayscale and indistinctive appearance patterns, tip detection in US images is challenging.

METHODS: To achieve precise tip detection in US images against spatial indiscernibility, a novel multi-keyframe motion-aware framework called TipDet is proposed. It can identify tips based on their short-term spatial-temporal pattern and long-term motion pattern. In TipDet, first, an adaptive keyframe model (AKM) is proposed to decide whether a frame is informative to serve as a keyframe for long-term motion pattern learning. Second, candidate tip detection is conducted using a two-stream backbone (TSB) based on their short-term spatial-temporal pattern. Third, to further identify the true one in the candidate tips, a novel method for learning the long-term motion pattern of the tips is proposed based on the proposed optical-flow-aware multi-head cross-attention (OFA-MHCA).

RESULTS: On the clinical human puncture dataset, which includes 4195 B-mode images, the experimental results show that the proposed TipDet can achieve precise tip detection against the spatial indiscernibility problem, achieving 78.7 % AP0.1:0.5 and 8.9 % improvement over the base detector at approximately 20 FPS. Moreover, a tip localization error of 1.3±0.6 % is achieved, exceeding the existing method.

CONCLUSIONS: The proposed TipDet can facilitate a wider and easier application of US-guided interventional procedures by providing robust and precise needle tip localization. The codes and data are available at https://github.com/ResonWang/TipDet.

PMID:38460346 | DOI:10.1016/j.cmpb.2024.108109

Sci Rep. 2024 Mar 9;14(1):5791. doi: 10.1038/s41598-024-56389-4.

ABSTRACT

Diabetic retinopathy (DR) is a serious ocular complication that can pose a serious risk to a patient's vision and overall health. Currently, the automatic grading of DR is mainly using deep learning techniques. However, the lesion information in DR images is complex, variable in shape and size, and randomly distributed in the images, which leads to some shortcomings of the current research methods, i.e., it is difficult to effectively extract the information of these various features, and it is difficult to establish the connection between the lesion information in different regions. To address these shortcomings, we design a multi-scale dynamic fusion (MSDF) module and combine it with graph convolution operations to propose a multi-scale dynamic graph convolutional network (MDGNet) in this paper. MDGNet firstly uses convolution kernels with different sizes to extract features with different shapes and sizes in the lesion regions, and then automatically learns the corresponding weights for feature fusion according to the contribution of different features to model grading. Finally, the graph convolution operation is used to link the lesion features in different regions. As a result, our proposed method can effectively combine local and global features, which is beneficial for the correct DR grading. We evaluate the effectiveness of method on two publicly available datasets, namely APTOS and DDR. Extensive experiments demonstrate that our proposed MDGNet achieves the best grading results on APTOS and DDR, and is more accurate and diverse for the extraction of lesion information.

PMID:38461342 | DOI:10.1038/s41598-024-56389-4

Ultrasound Med Biol. 2024 Mar 8:S0301-5629(24)00030-9. doi: 10.1016/j.ultrasmedbio.2024.01.016. Online ahead of print.

ABSTRACT

OBJECTIVE: Spontaneous echo contrast (SEC) is a vascular ultrasound finding associated with increased thromboembolism risk. However, identification requires expert determination and clinician time to report. We developed a deep learning model that can automatically identify SEC. Our model can be applied retrospectively without deviating from routine clinical practice. The retrospective nature of our model means future works could scan archival data to opportunistically correlate SEC findings with documented clinical outcomes.

METHODS: We curated a data set of 801 archival acquisitions along the femoral vein from 201 patients. We used a multisequence convolutional neural network (CNN) with ResNetv2 backbone and visualized keyframe importance using soft attention. We evaluated SEC prediction performance using an 80/20 train/test split. We report receiver operating characteristic area under the curve (ROC-AUC), along with the Youden threshold-associated sensitivity, specificity, F1 score, true negative, false negative, false positive and true positive.

RESULTS: Using soft attention, we can identify SEC with an AUC of 0.74, sensitivity of 0.73 and specificity of 0.68. Without soft attention, our model achieves an AUC of 0.69, sensitivity of 0.71 and specificity of 0.60. Additionally, we provide attention visualizations and note that our model assigns higher attention score to ultrasound frames containing more vessel lumen.

CONCLUSION: Our multisequence CNN model can identify the presence of SEC from ultrasound keyframes with an AUC of 0.74, which could enable screening applications and enable more SEC data discovery. The model does not require the expert intervention or additional clinician reporting time that are currently significant barriers to SEC adoption. Model and processed data sets are publicly available at https://github.com/Ouwen/automatic-spontaneous-echo-contrast.

PMID:38461036 | DOI:10.1016/j.ultrasmedbio.2024.01.016

BMJ Open Respir Res. 2024 Mar 9;11(1):e002226. doi: 10.1136/bmjresp-2023-002226.

ABSTRACT

PURPOSE: Acute exacerbation of idiopathic pulmonary fibrosis (AE-IPF) is the primary cause of death in patients with IPF, characterised by diffuse, bilateral ground-glass opacification on high-resolution CT (HRCT). This study proposes a three-dimensional (3D)-based deep learning algorithm for classifying AE-IPF using HRCT images.

MATERIALS AND METHODS: A novel 3D-based deep learning algorithm, SlowFast, was developed by applying a database of 306 HRCT scans obtained from two centres. The scans were divided into four separate subsets (training set, n=105; internal validation set, n=26; temporal test set 1, n=79; and geographical test set 2, n=96). The final training data set consisted of 1050 samples with 33 600 images for algorithm training. Algorithm performance was evaluated using accuracy, sensitivity, specificity, positive predictive value, negative predictive value, receiver operating characteristic (ROC) curve and weighted κ coefficient.

RESULTS: The accuracy of the algorithm in classifying AE-IPF on the test sets 1 and 2 was 93.9% and 86.5%, respectively. Interobserver agreements between the algorithm and the majority opinion of the radiologists were good (κw=0.90 for test set 1 and κw=0.73 for test set 2, respectively). The ROC accuracy of the algorithm for classifying AE-IPF on the test sets 1 and 2 was 0.96 and 0.92, respectively. The algorithm performance was superior to visual analysis in accurately diagnosing radiological findings. Furthermore, the algorithm's categorisation was a significant predictor of IPF progression.

CONCLUSIONS: The deep learning algorithm provides high auxiliary diagnostic efficiency in patients with AE-IPF and may serve as a useful clinical aid for diagnosis.

PMID:38460976 | DOI:10.1136/bmjresp-2023-002226

Comput Biol Med. 2024 Feb 28;172:108235. doi: 10.1016/j.compbiomed.2024.108235. Online ahead of print.

ABSTRACT

Cardiovascular diseases (CVD) are a leading cause of death globally, and result in significant morbidity and reduced quality of life. The electrocardiogram (ECG) plays a crucial role in CVD diagnosis, prognosis, and prevention; however, different challenges still remain, such as an increasing unmet demand for skilled cardiologists capable of accurately interpreting ECG. This leads to higher workload and potential diagnostic inaccuracies. Data-driven approaches, such as machine learning (ML) and deep learning (DL) have emerged to improve existing computer-assisted solutions and enhance physicians' ECG interpretation of the complex mechanisms underlying CVD. However, many ML and DL models used to detect ECG-based CVD suffer from a lack of explainability, bias, as well as ethical, legal, and societal implications (ELSI). Despite the critical importance of these Trustworthy Artificial Intelligence (AI) aspects, there is a lack of comprehensive literature reviews that examine the current trends in ECG-based solutions for CVD diagnosis or prognosis that use ML and DL models and address the Trustworthy AI requirements. This review aims to bridge this knowledge gap by providing a systematic review to undertake a holistic analysis across multiple dimensions of these data-driven models such as type of CVD addressed, dataset characteristics, data input modalities, ML and DL algorithms (with a focus on DL), and aspects of Trustworthy AI like explainability, bias and ethical considerations. Additionally, within the analyzed dimensions, various challenges are identified. To these, we provide concrete recommendations, equipping other researchers with valuable insights to understand the current state of the field comprehensively.

PMID:38460311 | DOI:10.1016/j.compbiomed.2024.108235

Comput Biol Med. 2024 Mar 2;172:108239. doi: 10.1016/j.compbiomed.2024.108239. Online ahead of print.

ABSTRACT

The identification of compound-protein interactions (CPIs) plays a vital role in drug discovery. However, the huge cost and labor-intensive nature in vitro and vivo experiments make it urgent for researchers to develop novel CPI prediction methods. Despite emerging deep learning methods have achieved promising performance in CPI prediction, they also face ongoing challenges: (i) providing bidirectional interpretability from both the chemical and biological perspective for the prediction results; (ii) comprehensively evaluating model generalization performance; (iii) demonstrating the practical applicability of these models. To overcome the challenges posed by current deep learning methods, we propose a cross multi-head attention oriented bidirectional interpretable CPI prediction model (CmhAttCPI). First, CmhAttCPI takes molecular graphs and protein sequences as inputs, utilizing the GCW module to learn atom features and the CNN module to learn residue features, respectively. Second, the model applies cross multi-head attention module to compute attention weights for atoms and residues. Finally, CmhAttCPI employs a fully connected neural network to predict scores for CPIs. We evaluated the performance of CmhAttCPI on balanced datasets and imbalanced datasets. The results consistently show that CmhAttCPI outperforms multiple state-of-the-art methods. We constructed three scenarios based on compound and protein clustering and comprehensively evaluated the model generalization ability within these scenarios. The results demonstrate that the generalization ability of CmhAttCPI surpasses that of other models. Besides, the visualizations of attention weights reveal that CmhAttCPI provides chemical and biological interpretation for CPI prediction. Moreover, case studies confirm the practical applicability of CmhAttCPI in discovering anticancer candidates.

PMID:38460309 | DOI:10.1016/j.compbiomed.2024.108239

Sci Rep. 2024 Mar 9;14(1):5809. doi: 10.1038/s41598-024-56476-6.

ABSTRACT

This study aimed to develop a deep learning model to assess the quality of fetal echocardiography and to perform prospective clinical validation. The model was trained on data from the 18-22-week anomaly scan conducted in seven hospitals from 2008 to 2018. Prospective validation involved 100 patients from two hospitals. A total of 5363 images from 2551 pregnancies were used for training and validation. The model's segmentation accuracy depended on image quality measured by a quality score (QS). It achieved an overall average accuracy of 0.91 (SD 0.09) across the test set, with images having above-average QS scoring 0.97 (SD 0.03). During prospective validation of 192 images, clinicians rated 44.8% (SD 9.8) of images as equal in quality, 18.69% (SD 5.7) favoring auto-captured images and 36.51% (SD 9.0) preferring manually captured ones. Images with above average QS showed better agreement on segmentations (p < 0.001) and QS (p < 0.001) with fetal medicine experts. Auto-capture saved additional planes beyond protocol requirements, resulting in more comprehensive echocardiographies. Low QS had adverse effect on both model performance and clinician's agreement with model feedback. The findings highlight the importance of developing and evaluating AI models based on 'noisy' real-life data rather than pursuing the highest accuracy possible with retrospective academic-grade data.

PMID:38461322 | DOI:10.1038/s41598-024-56476-6

BMC Bioinformatics. 2024 Mar 9;25(1):105. doi: 10.1186/s12859-024-05723-8.

ABSTRACT

MOTIVATION: The prediction of cancer drug response is a challenging subject in modern personalized cancer therapy due to the uncertainty of drug efficacy and the heterogeneity of patients. It has been shown that the characteristics of the drug itself and the genomic characteristics of the patient can greatly influence the results of cancer drug response. Therefore, accurate, efficient, and comprehensive methods for drug feature extraction and genomics integration are crucial to improve the prediction accuracy.

RESULTS: Accurate prediction of cancer drug response is vital for guiding the design of anticancer drugs. In this study, we propose an end-to-end deep learning model named DeepAEG which is based on a complete-graph update mode to predict IC50. Specifically, we integrate an edge update mechanism on the basis of a hybrid graph convolutional network to comprehensively learn the potential high-dimensional representation of topological structures in drugs, including atomic characteristics and chemical bond information. Additionally, we present a novel approach for enhancing simplified molecular input line entry specification data by employing sequence recombination to eliminate the defect of single sequence representation of drug molecules. Our extensive experiments show that DeepAEG outperforms other existing methods across multiple evaluation parameters in multiple test sets. Furthermore, we identify several potential anticancer agents, including bortezomib, which has proven to be an effective clinical treatment option. Our results highlight the potential value of DeepAEG in guiding the design of specific cancer treatment regimens.

PMID:38461284 | DOI:10.1186/s12859-024-05723-8

Sci Rep. 2024 Mar 9;14(1):5787. doi: 10.1038/s41598-024-56522-3.

ABSTRACT

All-optical plasmonic switches (AOPSs) utilizing surface plasmon polaritons are well-suited for integration into photonic integrated circuits (PICs) and play a crucial role in advancing all-optical signal processing. The current AOPS design methods still rely on trial-and-error or empirical approaches. In contrast, recent deep learning (DL) advances have proven highly effective as computational tools, offering an alternative means to accelerate nanophotonics simulations. This paper proposes an innovative approach utilizing DL for spectrum prediction and inverse design of AOPS. The switches employ circular nonlinear plasmonic ring resonators (NPRRs) composed of interconnected metal-insulator-metal waveguides with a ring resonator. The NPRR switching performance is shown using the nonlinear Kerr effect. The forward model presented in this study demonstrates superior computational efficiency when compared to the finite-difference time-domain method. The model analyzes various structural parameters to predict transmission spectra with a distinctive dip. Inverse modeling enables the prediction of design parameters for desired transmission spectra. This model provides a rapid estimation of design parameters, offering a clear advantage over time-intensive conventional optimization approaches. The loss of prediction for both the forward and inverse models, when compared to simulations, is exceedingly low and on the order of 10-4. The results confirm the suitability of employing DL for forward and inverse design of AOPSs in PICs.

PMID:38461205 | DOI:10.1038/s41598-024-56522-3

Chest. 2024 Mar;165(3):e87-e89. doi: 10.1016/j.chest.2023.10.039.

NO ABSTRACT

PMID:38461027 | DOI:10.1016/j.chest.2023.10.039