J Electrocardiol. 2024 Dec 25;89:153862. doi: 10.1016/j.jelectrocard.2024.153862. Online ahead of print.

ABSTRACT

BACKGROUND AND PURPOSE: Atrial fibrillation (AF), a common arrhythmia, is linked with atrial electrical and structural changes, notably low voltage areas (LVAs) which are associated with poor ablation outcomes and increased thromboembolic risk. This study aims to evaluate the efficacy of a deep learning model applied to 12‑lead ECGs for non-invasively predicting the presence of LVAs, potentially guiding pre-ablation strategies and improving patient outcomes.

METHODS: A retrospective analysis was conducted on 204 AF patients, who underwent catheter ablation. Pre-procedural sinus rhythm ECGs and electroanatomical maps (EAM) were utilized alongside demographic data to train a deep learning model combining Long Short-Term Memory networks and Convolutional Neural Networks with a cross-attention layer. Model performance was evaluated using a 5-fold cross-validation strategy.

RESULTS: The model effectively identified the presence of LVA on the examined atrial walls, achieving accuracies of 78 % for both the anterior and posterior walls, and 82 % for the LA roof. Moreover, it accurately predicted the global left atrial (LA) average voltage <0.7 mV, with an accuracy of 88 %.

CONCLUSION: The study showcases the potential of deep learning applied to 12‑lead ECGs to effectively predict regional LVAs and global LA voltage in AF patients non-invasively. This model offers a promising tool for the pre-ablation assessment of atrial substrate, facilitating personalized therapeutic strategies and potentially enhancing ablation success rates.

PMID:39742814 | DOI:10.1016/j.jelectrocard.2024.153862

J Electrocardiol. 2024 Dec 27;89:153858. doi: 10.1016/j.jelectrocard.2024.153858. Online ahead of print.

ABSTRACT

Electrocardiography (ECG) is a widely used, non-invasive, and cost-effective diagnostic method that plays a crucial role in the early detection and management of cardiac conditions. However, the ECG signal is easily disrupted by various noise signals in the real world, leading to a decrease in signal quality and potentially compromising accurate clinical interpretation. With the goal of reducing noise in ECG signals, this research proposes an end-to-end multi-resolution deep learning network with attention mechanism, namely the MrSeNet to perform effective denoising of ECG data. Our MrSeNet fuses features at different scales for effective denoising with the squeeze-and-excitation module to enhance the features of the ECG signal channel. CPSC2018 database and the MIT-BIH database were used to verify the validity of the model by adding different intensity noises based on NSTDB. Using Pearson correlation coefficient, signal-to-noise ratio, and root mean square error performance evaluation model, the experimental results show that MrSeNet performs better than the traditional method, the model can achieve a good denoising effect to different degrees of noise signal data, and has a good future application prospect.

PMID:39742813 | DOI:10.1016/j.jelectrocard.2024.153858

Clin Imaging. 2024 Dec 24;119:110392. doi: 10.1016/j.clinimag.2024.110392. Online ahead of print.

ABSTRACT

BACKGROUND: Thyroid cancer, a common endocrine malignancy, has seen increasing incidence, making lymph node metastasis (LNM) a critical factor for recurrence and survival. Radiomics and deep learning (DL) advancements offer the potential for improved LNM prediction using CT and MRI, though challenges in diagnostic accuracy remain.

METHODS: A systematic review and meta-analysis were conducted per established guidelines, with searches across PubMed, Scopus, Web of Science, and Embase up to February 15, 2024. Studies developing CT/MRI-based radiomics and/or DL models for preoperative LNM assessment in thyroid cancer patients were included. Data were extracted and analyzed using R software.

RESULTS: Sixteen studies were analyzed. In internal validation sets, sensitivity was 81.1 % (95 % CI: 75.6 %-85.6 %) and specificity 76.4 % (95 % CI: 68.4 %-82.9 %). Training sets showed a sensitivity of 84.4 % (95 % CI: 81.5 %-87 %) and a specificity of 84.7 % (95 % CI: 74.4 %-91.4 %). The pooled AUC was 86 % for internal validation and 87 % for training. Handcrafted radiomics had a sensitivity of 79.4 % and specificity of 69.2 %, while DL models showed 80.8 % sensitivity and 78.7 % specificity. Subgroup analysis revealed that models for papillary thyroid cancer (PTC) had a pooled specificity of 76.3 %, while those including other or unspecified cancers showed 68.3 % specificity. Despite heterogeneity, significant differences (p = 0.037) were noted between models with and without clinical data.

CONCLUSION: Radiomics and DL models show promising potential for detecting LNM in thyroid cancer, particularly in PTC. However, study heterogeneity underscores the need for further research to optimize these imaging tools.

PMID:39742800 | DOI:10.1016/j.clinimag.2024.110392

Clin Imaging. 2024 Dec 26;119:110386. doi: 10.1016/j.clinimag.2024.110386. Online ahead of print.

ABSTRACT

We aimed to systematically review and meta-analyze the predictive value of magnetic resonance imaging (MRI)-derived radiomics/end-to-end deep learning (DL) models in predicting glioma alpha thalassemia/mental retardation syndrome X-linked (ATRX) status. We conducted a comprehensive search across four major databases-Web of Science, PubMed, Scopus, and Embase. All the studies that assessed the performance of radiomics and/or end-to-end DL models for predicting glioma ATRX status were included. Quality assessment was performed using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) criteria and the METhodological RadiomICs Score (METRICS). Pooled estimates for performance metrics were calculated. I-squared was used to assess heterogeneity, while subgroup and sensitivity analyses were performed to find its potential sources. Publication bias was assessed using Deeks' funnel plots. Seventeen and eleven studies were included in the systematic review and meta-analysis, respectively. Most of the studies had a low risk of bias and low concern for applicability according to the QUADAS-2. Also, most of them had good quality according to the METRICS. Meta-analysis showed a pooled sensitivity of 0.80 (95%CI: 0.71-0.96), a specificity of 0.82 (95%CI: 0.67-0.93), a positive diagnostic likelihood ratio (DLR) of 6.77 (95%CI: 4.67-9.82), a negative DLR of 0.15 (95%CI: 0.06-0.38), a diagnostic odds ratio of 30.36 (95%CI: 15.87-58.05), and an area under the curve (AUC) of 0.92 (95%CI: 0.89-0.94). Subgroup analysis revealed significant intergroup differences based on several factors. Radiomics models can accurately predict ATRX status in gliomas, enhancing non-invasive tumor characterization and guiding treatment strategies.

PMID:39742798 | DOI:10.1016/j.clinimag.2024.110386

Biochem Biophys Res Commun. 2024 Dec 24;746:151240. doi: 10.1016/j.bbrc.2024.151240. Online ahead of print.

ABSTRACT

Utilizing Artificial Intelligence (AI) in computational biology techniques could offer significant advantages in alleviating the growing workloads faced by structural biologists, especially with the emergence of big data. In this study, we employed Delaunay tessellation as a promising method to obtain the overall structural topology of proteins. Subsequently, we developed multi-class deep neural network models to classify protein superfamilies based on their local topology. Our models achieved a test accuracy of approximately 0.92 in classifying proteins into 18 well-populated superfamilies. We believe that the results of this study hold substantial value since, to the best of our knowledge, no previous studies have reported the utilization of protein topological data for protein classification through deep learning and Delaunay tessellation.

PMID:39742787 | DOI:10.1016/j.bbrc.2024.151240

Comput Methods Programs Biomed. 2024 Dec 28;260:108571. doi: 10.1016/j.cmpb.2024.108571. Online ahead of print.

ABSTRACT

BACKGROUND: Data sharing in healthcare is vital for advancing research and personalized medicine. However, the process is hindered by privacy, ethical, and legal challenges associated with patient data. Synthetic data generation emerges as a promising solution, replicating statistical properties of real data while enhancing privacy protection.

METHODS: This systematic review examines deep learning techniques for synthetic data generation in healthcare, focusing on their ability to maintain data utility and enhance privacy. Studies from Scopus, Web of Science, PubMed, and IEEE databases published between 2019 and 2023 were analyzed. Key methods explored include Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and Diffusion Models. Evaluation metrics encompass data resemblance, utility, and privacy preservation, with special attention to privacy-enhancing methods like differential privacy and federated learning.

RESULTS: GANs and VAEs demonstrated robust capabilities in generating realistic synthetic data for tabular, signal, image, and multi-modal datasets. Privacy-preserving approaches such as differential privacy and adversarial training significantly reduced re-identification risks while maintaining data fidelity. However, challenges persist in preserving temporal correlations, reducing biases, and aligning with regulatory frameworks, particularly for longitudinal and high-dimensional data.

CONCLUSION: Synthetic data generation holds significant potential for privacy-preserving data sharing in healthcare. Ongoing research is required to develop advanced algorithms and evaluation frameworks, ensuring synthetic data's quality and privacy. Collaboration between technologists and policymakers is essential to create comprehensive guidelines, fostering secure and effective data sharing in healthcare.

PMID:39742693 | DOI:10.1016/j.cmpb.2024.108571

Front Radiol. 2024 Dec 17;4:1523389. doi: 10.3389/fradi.2024.1523389. eCollection 2024.

NO ABSTRACT

PMID:39742350 | PMC:PMC11685185 | DOI:10.3389/fradi.2024.1523389

Front Radiol. 2024 Dec 16;4:1498411. doi: 10.3389/fradi.2024.1498411. eCollection 2024.

ABSTRACT

BACKGROUND: MR fingerprinting (MRF) is a novel method for quantitative assessment of in vivo MR relaxometry that has shown high precision and accuracy. However, the method requires data acquisition using customized, complex acquisition strategies and dedicated post processing methods thereby limiting its widespread application.

OBJECTIVE: To develop a deep learning (DL) network for synthesizing MRF signals from conventional magnitude-only MR imaging data and to compare the results to the actual MRF signal acquired.

METHODS: A U-Net DL network was developed to synthesize MRF signals from magnitude-only 3D T 1-weighted brain MRI data acquired from 37 volunteers aged between 21 and 62 years of age. Network performance was evaluated by comparison of the relaxometry data (T 1, T 2) generated from dictionary matching of the deep learning synthesized and actual MRF data from 47 segmented anatomic regions. Clustered bootstrapping involving 10,000 bootstraps followed by calculation of the concordance correlation coefficient were performed for both T 1 and T 2 MRF data pairs. 95% confidence limits and the mean difference between true and DL relaxometry values were also calculated.

RESULTS: The concordance correlation coefficient (and 95% confidence limits) for T 1 and T 2 MRF data pairs over the 47 anatomic segments were 0.8793 (0.8136-0.9383) and 0.9078 (0.8981-0.9145) respectively. The mean difference (and 95% confidence limits) were 48.23 (23.0-77.3) s and 2.02 (-1.4 to 4.8) s.

CONCLUSION: It is possible to synthesize MRF signals from MRI data using a DL network, thereby creating the potential for performing quantitative relaxometry assessment without the need for a dedicated MRF pulse sequence.

PMID:39742349 | PMC:PMC11686891 | DOI:10.3389/fradi.2024.1498411

Imeta. 2024 Oct 17;3(6):e244. doi: 10.1002/imt2.244. eCollection 2024 Dec.

ABSTRACT

Antimicrobial peptides (AMPs) have become a viable source of novel antibiotics that are effective against human pathogenic bacteria. In this study, we construct a bank of culturable marine biofilm bacteria constituting 713 strains and their nearly complete genomes and predict AMPs using ribosome profiling and deep learning. Compared with previous approaches, ribosome profiling has improved the identification and validation of small open reading frames (sORFs) for AMP prediction. Among the 80,430 expressed sORFs, 341 are identified as candidate AMPs with high probability. Most potential AMPs have less than 40% similarity in their amino acid sequence compared to those listed in public databases. Furthermore, these AMPs are associated with bacterial groups that are not previously known to produce AMPs. Therefore, our deep learning model has acquired characteristics of unfamiliar AMPs. Chemical synthesis of 60 potential AMP sequences yields 54 compounds with antimicrobial activity, including potent inhibitory effects on various drug-resistant human pathogens. This study extends the range of AMP compounds by investigating marine biofilm microbiomes using a novel approach, accelerating AMP discovery.

PMID:39742298 | PMC:PMC11683478 | DOI:10.1002/imt2.244

Rev Cardiovasc Med. 2024 Dec 24;25(12):454. doi: 10.31083/j.rcm2512454. eCollection 2024 Dec.

ABSTRACT

BACKGROUND: This study aimed to develop and evaluate the detection and classification performance of different deep learning models on carotid plaque ultrasound images to achieve efficient and precise ultrasound screening for carotid atherosclerotic plaques.

METHODS: This study collected 5611 carotid ultrasound images from 3683 patients from four hospitals between September 17, 2020, and December 17, 2022. By cropping redundant information from the images and annotating them using professional physicians, the dataset was divided into a training set (3927 images) and a test set (1684 images). Four deep learning models, You Only Look Once Version 7 (YOLO V7) and Faster Region-Based Convolutional Neural Network (Faster RCNN) were employed for image detection and classification to distinguish between vulnerable and stable carotid plaques. Model performance was evaluated using accuracy, sensitivity, specificity, F1 score, and area under curve (AUC), with p < 0.05 indicating a statistically significant difference.

RESULTS: We constructed and compared deep learning models based on different network architectures. In the test set, the Faster RCNN (ResNet 50) model exhibited the best classification performance (accuracy (ACC) = 0.88, sensitivity (SEN) = 0.94, specificity (SPE) = 0.71, AUC = 0.91), significantly outperforming the other models. The results suggest that deep learning technology has significant potential for application in detecting and classifying carotid plaque ultrasound images.

CONCLUSIONS: The Faster RCNN (ResNet 50) model demonstrated high accuracy and reliability in classifying carotid atherosclerotic plaques, with diagnostic capabilities approaching that of intermediate-level physicians. It has the potential to enhance the diagnostic abilities of primary-level ultrasound physicians and assist in formulating more effective strategies for preventing ischemic stroke.

PMID:39742249 | PMC:PMC11683696 | DOI:10.31083/j.rcm2512454

Rev Cardiovasc Med. 2024 Dec 28;25(12):463. doi: 10.31083/j.rcm2512463. eCollection 2024 Dec.

ABSTRACT

BACKGROUND: Obstructive sleep apnea (OSA) is a severe condition associated with numerous cardiovascular complications, including heart failure. The complex biological and morphological relationship between OSA and atherosclerotic cardiovascular disease (ASCVD) poses challenges in predicting adverse cardiovascular outcomes. While artificial intelligence (AI) has shown potential for predicting cardiovascular disease (CVD) and stroke risks in other conditions, there is a lack of detailed, bias-free, and compressed AI models for ASCVD and stroke risk stratification in OSA patients. This study aimed to address this gap by proposing three hypotheses: (i) a strong relationship exists between OSA and ASCVD/stroke, (ii) deep learning (DL) can stratify ASCVD/stroke risk in OSA patients using surrogate carotid imaging, and (iii) including OSA risk as a covariate with cardiovascular risk factors can improve CVD risk stratification.

METHODS: The study employed the Preferred Reporting Items for Systematic reviews and Meta-analyses (PRISMA) search strategy, yielding 191 studies that link OSA with coronary, carotid, and aortic atherosclerotic vascular diseases. This research investigated the link between OSA and CVD, explored DL solutions for OSA detection, and examined the role of DL in utilizing carotid surrogate biomarkers by saving costs. Lastly, we benchmark our strategy against previous studies.

RESULTS: (i) This study found that CVD and OSA are indirectly or directly related. (ii) DL models demonstrated significant potential in improving OSA detection and proved effective in CVD risk stratification using carotid ultrasound as a biomarker. (iii) Additionally, DL was shown to be useful for CVD risk stratification in OSA patients; (iv) There are important AI attributes such as AI-bias, AI-explainability, AI-pruning, and AI-cloud, which play an important role in CVD risk for OSA patients.

CONCLUSIONS: DL provides a powerful tool for CVD risk stratification in OSA patients. These results can promote several recommendations for developing unique, bias-free, and explainable AI algorithms for predicting ASCVD and stroke risks in patients with OSA.

PMID:39742217 | PMC:PMC11683711 | DOI:10.31083/j.rcm2512463

J Vitreoretin Dis. 2024 Dec 30:24741264241308495. doi: 10.1177/24741264241308495. Online ahead of print.

ABSTRACT

Purpose: To evaluate the combined relationship between ischemia, retinal fluid, and layer thickness measurements with visual acuity (VA) outcomes in patients with retinal vein occlusion (RVO). Methods: Swept-source optical coherence tomography (OCT) data were used to assess retinal layer thickness and quantify intraretinal fluid (IRF) and subretinal fluid (SRF) using a deep learning-based, macular fluid segmentation algorithm for treatment-naïve eyes diagnosed with visual impairment resulting from central RVO (CRVO) or branch RVO (BRVO). Patients received 3 loading doses of 2 mg intravitreal aflibercept injections and were then put on a treat-and-extend regimen. Image analysis was performed at baseline and postoperatively at 3 months and 6 months. The baseline OCT morphologic features and fluid measurements were correlated with the changes in best-corrected VA (BCVA) using the Pearson correlation coefficient (r). Results: The study comprised 49 eyes. A combined model incorporating thickness in the outer plexiform layer (OPL), retinal nerve fiber layer (RNFL), and presence of IRF had the strongest overall correlation for CRVO (r = 0.865; P < .05). For BRVO, the addition of IRF to the OPL-inner nasal model had a strong correlation (r = 0.803; P < .05). The baseline ischemic index in the deep capillary complex showed a notable correlation with the 6-month change in BCVA for CRVO (r = 0.9101; P < .001) and BRVO (r = 0.9200; P < .001). Conclusions: A combined model of IRF volume, OPL, and RNFL layer thicknesses, along with ischemic indices, provides the best correlation to BCVA changes. Combined fluid and layer segmentation of OCT images provides clinically useful biomarkers for RVO patients. These results give insight into the pathology of RVOs and describe the relationship between deep capillary complex ischemia and OPL/RNFL thickness in BCVA outcomes.

PMID:39742143 | PMC:PMC11683825 | DOI:10.1177/24741264241308495

Front Nutr. 2024 Dec 17;11:1469878. doi: 10.3389/fnut.2024.1469878. eCollection 2024.

ABSTRACT

INTRODUCTION: Nutrition is closely related to body health. A reasonable diet structure not only meets the body's needs for various nutrients but also effectively prevents many chronic diseases. However, due to the general lack of systematic nutritional knowledge, people often find it difficult to accurately assess the nutritional content of food. In this context, image-based nutritional evaluation technology can provide significant assistance. Therefore, we are dedicated to directly predicting the nutritional content of dishes through images. Currently, most related research focuses on estimating the volume or area of food through image segmentation tasks and then calculating its nutritional content based on the food category. However, this method often lacks real nutritional content labels as a reference, making it difficult to ensure the accuracy of the predictions.

METHODS: To address this issue, we combined segmentation and regression tasks and used the Nutrition5k dataset, which contains detailed nutritional content labels but no segmentation labels, for manual segmentation annotation. Based on these annotated data, we developed a nutritional content prediction model that performs segmentation first and regression afterward. Specifically, we first applied the UNet model to segment the food, then used a backbone network to extract features, and enhanced the feature expression capability through the Squeeze-and-Excitation structure. Finally, the extracted features were processed through several fully connected layers to obtain predictions for the weight, calories, fat, carbohydrates, and protein content.

RESULTS AND DISCUSSION: Our model achieved an outstanding average percentage mean absolute error (PMAE) of 17.06% for these components. All manually annotated segmentation labels can be found at https://doi.org/10.6084/m9.figshare.26252048.v1.

PMID:39742105 | PMC:PMC11685081 | DOI:10.3389/fnut.2024.1469878

Digit Health. 2024 Dec 25;10:20552076241307573. doi: 10.1177/20552076241307573. eCollection 2024 Jan-Dec.

ABSTRACT

BACKGROUND: Diabetic peripheral neuropathy (DPN) is a common complication of diabetes, and its early identification is crucial for improving patient outcomes. Corneal confocal microscopy (CCM) can non-invasively detect changes in corneal nerve fibers (CNFs), making it a potential tool for the early diagnosis of DPN. However, the existing CNF analysis methods have certain limitations, highlighting the need to develop a reliable automated analysis tool.

METHODS: This study is based on data from two independent clinical centers. Various popular deep learning (DL) models have been trained and evaluated for their performance in CCM image segmentation using DL-based image segmentation techniques. Subsequently, an image processing algorithm was designed to automatically extract and quantify various morphological parameters of CNFs. To validate the effectiveness of this tool, it was compared with manually annotated datasets and ACCMetrics, and the consistency of the results was assessed using Bland--Altman analysis and intraclass correlation coefficient (ICC).

RESULTS: The U2Net model performed the best in the CCM image segmentation task, achieving a mean Intersection over Union (mIoU) of 0.8115. The automated analysis tool based on U2Net demonstrated a significantly higher consistency with the manually annotated results in the quantitative analysis of various CNF morphological parameters than the previously popular automated tool ACCMetrics. The area under the curve for classifying DPN using the CNF morphology parameters calculated by this tool reached 0.75.

CONCLUSIONS: The DL-based automated tool developed in this study can effectively segment and quantify the CNF parameters in CCM images. This tool has the potential to be used for the early diagnosis of DPN, and further research will help validate its practical application value in clinical settings.

PMID:39741986 | PMC:PMC11686633 | DOI:10.1177/20552076241307573

Nat Sci Sleep. 2024 Dec 27;16:2243-2256. doi: 10.2147/NSS.S492146. eCollection 2024.

ABSTRACT

PURPOSE: This study aims to develop a deep learning methodology for quantitative assessing adenoid hypertrophy in nasopharyngoscopy images and to investigate its correlation with the apnea-hypopnea index (AHI) in pediatric patients with obstructive sleep apnea (OSA).

PATIENTS AND METHODS: A total of 1642 nasopharyngoscopy images were collected from pediatric patients aged 3 to 12 years. After excluding images with obscured secretions, incomplete adenoid exposure, 1500 images were retained for analysis. The adenoid-to-nasopharyngeal (A/N) ratio was manually annotated by two experienced otolaryngologists using MATLAB's imfreehand tool. Inter-annotator agreement was assessed using the Mann-Whitney U-test. Deep learning segmentation models were developed with the MMSegmentation framework, incorporating transfer learning and ensemble learning techniques. Model performance was evaluated using precision, recall, mean intersection over union (MIoU), overall accuracy, Cohen's Kappa, confusion matrices, and receiver operating characteristic (ROC) curves. The correlation between the A/N ratio and AHI, derived from polysomnography, was analyzed to evaluate clinical relevance.

RESULTS: Manual evaluation of adenoid hypertrophy by otolaryngologists (p=0.8507) and MATLAB calibration (p=0.679) demonstrated high consistency, with no significant differences. Among the deep learning models, the ensemble learning-based SUMNet outperformed others, achieving the highest precision (0.9616), MIoU (0.8046), overall accuracy (0.9182), and Kappa (0.87). SUMNet also exhibited superior consistency in classifying adenoid sizes. ROC analysis revealed that SUMNet (AUC=0.85) outperformed expert evaluations (AUC=0.74). A strong positive correlation was observed between the A/N ratio and AHI, with the correlation coefficients for SUMNet-derived ratios ranging from r=0.9052 (tonsils size+1) to r=0.4452 (tonsils size+3) and for expert-derived ratios ranging from r=0.4590 (tonsils size+1) to r=0.2681 (tonsils size+3).

CONCLUSION: This study introduces a precise and reliable deep learning-based method for quantifying adenoid hypertrophy and addresses the challenge posed limited sample sizes in deep learning applications. The significant correlation between adenoid hypertrophy and AHI underscores the clinical utility of this method in pediatric OSA diagnosis.

PMID:39741799 | PMC:PMC11687100 | DOI:10.2147/NSS.S492146

Front Hum Neurosci. 2024 Dec 17;18:1471634. doi: 10.3389/fnhum.2024.1471634. eCollection 2024.

ABSTRACT

Emotion recognition is a critical research topic within affective computing, with potential applications across various domains. Currently, EEG-based emotion recognition, utilizing deep learning frameworks, has been effectively applied and achieved commendable performance. However, existing deep learning-based models face challenges in capturing both the spatial activity features and spatial topology features of EEG signals simultaneously. To address this challenge, a domain-adaptation spatial-feature perception-network has been proposed for cross-subject EEG emotion recognition tasks, named DSP-EmotionNet. Firstly, a spatial activity topological feature extractor module has been designed to capture spatial activity features and spatial topology features of EEG signals, named SATFEM. Then, using SATFEM as the feature extractor, DSP-EmotionNet has been designed, significantly improving the accuracy of the model in cross-subject EEG emotion recognition tasks. The proposed model surpasses state-of-the-art methods in cross-subject EEG emotion recognition tasks, achieving an average recognition accuracy of 82.5% on the SEED dataset and 65.9% on the SEED-IV dataset.

PMID:39741785 | PMC:PMC11685119 | DOI:10.3389/fnhum.2024.1471634

Maturitas. 2024 Dec 23;193:108185. doi: 10.1016/j.maturitas.2024.108185. Online ahead of print.

ABSTRACT

OBJECTIVE: Given that Parkinson's disease is a progressive disorder, with symptoms that worsen over time, our goal is to enhance the diagnosis of Parkinson's disease by utilizing machine learning techniques and microbiome analysis. The primary objective is to identify specific microbiome signatures that can reproducibly differentiate patients with Parkinson's disease from healthy controls.

METHODS: We used four Parkinson-related datasets from the NCBI repository, focusing on stool samples. Then, we applied a DADA2-based script for amplicon sequence processing and the Recursive Ensemble Feature Selection (REF) algorithm for biomarker discovery. The discovery dataset was PRJEB14674, while PRJNA742875, PRJEB27564, and PRJNA594156 served as testing datasets. The Extra Trees classifier was used to validate the selected features.

RESULTS: The Recursive Ensemble Feature Selection algorithm identified 84 features (Amplicon Sequence Variants) from the discovery dataset, achieving an accuracy of over 80%. The Extra Trees classifier demonstrated good diagnostic accuracy with an area under the receiver operating characteristic curve of 0.74. In the testing phase, the classifier achieved areas under the receiver operating characteristic curves of 0.64, 0.71, and 0.62 for the respective datasets, indicating sufficient to good diagnostic accuracy. The study identified several bacterial taxa associated with Parkinson's disease, such as Lactobacillus, Bifidobacterium, and Roseburia, which were increased in patients with the disease.

CONCLUSION: This study successfully identified microbiome signatures that can differentiate patients with Parkinson's disease from healthy controls across different datasets. These findings highlight the potential of integrating machine learning and microbiome analysis for the diagnosis of Parkinson's disease. However, further research is needed to validate these microbiome signatures and to explore their therapeutic implications in developing targeted treatments and diagnostics for Parkinson's disease.

PMID:39740526 | DOI:10.1016/j.maturitas.2024.108185

Comput Biol Med. 2024 Dec 30;186:109602. doi: 10.1016/j.compbiomed.2024.109602. Online ahead of print.

ABSTRACT

Automated segmentation and detection of tumors in CT scans of the liver and kidney have a significant potential in assisting clinicians with cancer diagnosis and treatment planning. However, current approaches, including state-of-the-art deep learning ones, still face many challenges. Many tumors are not detected by these approaches when tested on public datasets for tumor detection and segmentation such as the Kidney Tumor Segmentation Challenge (KiTS) and the Liver tumor segmentation challenge (LiTS). False negative rates by lesion as high as 50% are commonly observed, and this rate is even higher for smaller lesions as they exhibit a high degree of variability (heterogeneity) among themselves. Additionally, in numerous instances, these lesions share similarities (homogeneity) in intensity, size, and shape with other anatomical structures as well as blurriness and blending with surrounding tissue. To improve the detection and segmentation accuracy of lesions in CT scans of the liver and kidney, we propose a selective ensemble approach that uses the predictions of two models to select the best possible mask for lesions. Both models are based on the UNet architecture and use the ConvNext convolutional block in both the encoder and decoder. The first model is trained on lesion segmentation regardless of size, while the second is designed and fine-tuned to segment and detect small lesions. Once the segmentation mask is predicted from both models we extract intensity-based features from within the lesion, contrast them with features from surrounding tissue, and select the mask that maximizes features' separation between the two. We test our approach on three different datasets for lesion segmentation in the kidney and liver. Our proposed approach achieves an improved detection and segmentation performance and is able to increase the number of lesions detected in all three datasets when compared to current state-of-the-art models.

PMID:39740509 | DOI:10.1016/j.compbiomed.2024.109602

J Med Imaging Radiat Sci. 2024 Dec 30;56(2):101844. doi: 10.1016/j.jmir.2024.101844. Online ahead of print.

ABSTRACT

PURPOSE: Radiotherapy is a crucial part of breast cancer treatment. Precision in dose assessment is essential to minimize side effects. Traditionally, anatomical structures are delineated manually, a time-consuming process subject to variability. automatic segmentation, including methods based on multiple atlases and deep learning, offers a promising alternative. For the radiotherapy treatment of the left breast, the RTOG 1005 protocol highlights the importance of cardiac delineation and the need to minimize cardiac exposure to radiation. Our study aims to evaluate dose distribution in auto-segmented substructures and establish models to correlate them with dose in the cardiac area.

METHODS AND MATERIALS: Anatomical structures were auto-segmented using TotalSegmentator and Limbus AI. The relationship between the volume of the cardiac area and of organs at risk was assessed using log-linear regressions.

RESULTS: The mean dose distribution was considerable for LAD (left anterior descending coronary artery), heart, and left ventricle. The volumetric distribution of organs at risk is evaluated for specific RTOG 1005 isodoses. We highlight the greater variability in the absolute volumetric evaluation. Log-linear regression models are presented to estimate dose constraint parameters. We highlight a greater number of highly correlated comparisons for absolute dose-volume assessment.

CONCLUSIONS: Dose-volume assessment protocols in patients with left breast cancer often neglect cardiac substructures. However, automatic tools can overcome these technical difficulties. In this study, we correlated the dose in the cardiac area with the doses in specific substructures and suggested limits for planning evaluation. Our data also indicates that statistical models could be applied in the assessment of those substructures where an automatic segmentation tool is not available. Our data also shows a benefit in reporting absolute dose-volume thresholds for future cause-effect assessments.

PMID:39740303 | DOI:10.1016/j.jmir.2024.101844

J Dig Dis. 2024 Dec 30. doi: 10.1111/1751-2980.13324. Online ahead of print.

ABSTRACT

Pancreaticobiliary endoscopy is an essential tool for diagnosing and treating pancreaticobiliary diseases. However, it does not fully meet clinical needs, which presents challenges such as significant difficulty in operation and risks of missed diagnosis or misdiagnosis. In recent years, artificial intelligence (AI) has enhanced the diagnostic and treatment efficiency and quality of pancreaticobiliary endoscopy. Diagnosis and differential diagnosis based on endoscopic ultrasound (EUS) images, pathology of EUS-guided fine-needle aspiration or biopsy, need for endoscopic retrograde cholangiopancreatography (ERCP) and assessment of operational difficulty, postoperative complications and prediction of patient prognosis, and real-time procedure guidance. This review provides an overview of AI applications in pancreaticobiliary endoscopy and proposes future development directions in aspects such as data quality and algorithmic interpretability, aiming to provide new insights for the integration of AI technology with pancreaticobiliary endoscopy.

PMID:39740251 | DOI:10.1111/1751-2980.13324