Biomed Phys Eng Express. 2024 May 23. doi: 10.1088/2057-1976/ad4f8f. Online ahead of print.

ABSTRACT

Accurate lung segmentation in chest X-ray images plays a pivotal role in early disease detection and clinical decision-making. In this study, we introduce an innovative approach to enhance the precision of lung segmentation using the Segment Anything Model (SAM). Despite its versatility, SAM faces the challenge of prompt decoupling, often resulting in misclassifications, especially with intricate structures like the clavicle. Our research focuses on the integration of spatial at- tention mechanisms within SAM. This approach enables the model to concentrate specifically on the lung region, fostering adaptability to image variations and reducing the likelihood of false positives. This work has the potential to significantly advance lung segmentation, improving the identification and quantification of lung anomalies across diverse clinical contexts.

PMID:38781939 | DOI:10.1088/2057-1976/ad4f8f

Biomed Phys Eng Express. 2024 May 23. doi: 10.1088/2057-1976/ad4f91. Online ahead of print.

ABSTRACT

Congenital heart defects (CHD) are one of the serious problems that arise during pregnancy. Early CHD detection reduces death rates and morbidity but is hampered by the relatively low detection rates (i.e., 60%) of current screening technology. The detection rate could be increased by supplementing ultrasound imaging with fetal ultrasound image evaluation (FUSI) using deep learning techniques. As a result, the non-invasive foetal ultrasound image has clear potential in the diagnosis of CHD and should be considered in addition to foetal echocardiography. This review paper highlights cutting-edge technologies for detecting CHD using ultrasound images, which involve pre-processing, localization, segmentation, and classification. Existing technique of preprocessing includes spatial domain filter, non-linear mean filter, transform domain filter, and denoising methods based on Convolutional Neural Network (CNN); segmentation includes thresholding-based techniques, region growing-based techniques, edge detection techniques, Artificial Neural Network (ANN) based segmentation methods, non-deep learning approaches and deep learning approaches. The paper also suggests future research directions for improving current methodologies.

PMID:38781934 | DOI:10.1088/2057-1976/ad4f91

Biomed Phys Eng Express. 2024 May 23. doi: 10.1088/2057-1976/ad4f8e. Online ahead of print.

ABSTRACT

Deep learning models have demonstrated remarkable performance in the classification of motor imagery BCI systems. However, these models exhibit sensitivity to challenging trials, often called hard trials, leading to performance degradation. In this paper, we address this issue by proposing two novel methods for identifying and mitigating the impact of hard trials on model performance. The first method leverages model prediction scores to discern hard trials. The second approach employs a quantitative explainable artificial intelligence (XAI) approach, enabling a more transparent and interpretable means of hard trials identification.
The identified hard trials are removed from the entire motor imagery training and validation dataset, and the deep learning model is further re-trained using the dataset without hard trials.
To evaluate the efficacy of these proposed methods, experiments were conducted on the Open
BMI dataset. The results for hold-out analysis show that, the proposed quantitative XAI based hard trial removal method has statistically improved the average classification accuracy of the baseline deep CNN model from 63.77 % to 68.70 %, with p-value = 7.66 -11 for the subject specific MI classification. Additionally, analyzing the scalp map depicting the average relevance scores of correctly classified trials compared to a misclassified trial provides a deeper insight into identifying hard trials. The results indicates that the proposed quantitative based XAI approach outperforms the prediction-score based approach in hard trial identification.&#xD.

PMID:38781932 | DOI:10.1088/2057-1976/ad4f8e

Eur J Radiol. 2024 May 18;176:111510. doi: 10.1016/j.ejrad.2024.111510. Online ahead of print.

ABSTRACT

PURPOSE: To evaluate the diagnostic accuracy of computed tomography (CT)-based radiomic algorithms and deep learning models to preoperatively identify lymph node metastasis (LNM) in patients with pancreatic ductal adenocarcinoma (PDAC).

METHODS: PubMed, CENTRAL, Scopus, Web of Science and IEEE databases were searched to identify relevant studies published up until February 11, 2024. Two reviewers screened all papers independently for eligibility. Studies reporting the accuracy of CT-based radiomics or deep learning models for detecting LNM in PDAC, using histopathology as the reference standard, were included. Quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies 2, the Radiomics Quality Score (RQS) and the the METhodological RadiomICs Score (METRICS). Overall sensitivity (SE), specificity (SP), diagnostic odds ratio (DOR), and the area under the curve (AUC) were calculated.

RESULTS: Four radiomics studies comprising 213 patients and four deep learning studies with 272 patients were included. The average RQS total score was 12.00 ± 3.89, corresponding to an RQS percentage of 33.33 ± 10.80, while the average METRICS score was 63.60 ± 10.88. A significant and strong positive correlation was found between RQS and METRICS (p = 0.016; r = 0.810). The pooled SE, SP, DOR, and AUC of all the studies were 0.83 (95 %CI = 0.77-0.88), 0.76 (95 %CI = 0.62-0.86), 15.70 (95 %CI = 8.12-27.50) and 0.85 (95 %CI = 0.77-0.88). Meta-regression analysis results indicated that neither the study type (radiomics vs deep learning) nor the dataset size of the studies had a significant effect on the DOR (p = 0.09 and p = 0.26, respectively).

CONCLUSION: Based on our meta-analysis findings, preoperative CT-based radiomics algorithms and deep learning models demonstrate favorable performance in predicting LNM in patients with PDAC, with a strong correlation between RQS and METRICS of the included studies.

PMID:38781919 | DOI:10.1016/j.ejrad.2024.111510

Mult Scler Relat Disord. 2024 May 11;87:105682. doi: 10.1016/j.msard.2024.105682. Online ahead of print.

ABSTRACT

BACKGROUND: Magnetic resonance imaging [MRI] findings in Neuromyelitis optica spectrum disorder [NMOSD] and Multiple Sclerosis [MS] patients could lead us to discriminate toward them. For instance, U-fiber and Dawson's finger-type lesions are suggestive of MS, however linear ependymal lesions raise the possibility of NMOSD. Recently, artificial intelligence [AI] models have been used to discriminate between NMOSD and MS based on MRI features. In this study, we aim to systematically review the capability of AI algorithms in NMOSD and MS discrimination based on MRI features.

METHOD: We searched PubMed, Scopus, Web of Sciences, Embase, and IEEE databases up to August 2023. All studies that used AI-based algorithms to discriminate between NMOSD and MS using MRI features were included, without any restriction in time, region, race, and age. Data on NMOSD and MS patients, Aquaporin-4 antibodies [AQP4-Ab] status, diagnosis criteria, performance metrics (accuracy, sensitivity, specificity, and AUC), artificial intelligence paradigm, MR imaging, and used features were extracted. This study is registered with PROSPERO, CRD42023465265.

RESULTS: Fifteen studies were included in this systematic review, with sample sizes ranging between 53 and 351. 1,362 MS patients and 1,118 NMOSD patients were included in our systematic review. AQP4-Ab was positive in 94.9% of NMOSD patients in 9 studies. Eight studies used machine learning [ML] as a classifier, while 7 used deep learning [DL]. AI models based on only MRI or MRI and clinical features yielded a pooled accuracy of 82% (95% CI: 78-86%), sensitivity of 83% (95% CI: 79-88%), and specificity of 80% (95% CI: 75-86%). In subgroup analysis, using only MRI features yielded an accuracy, sensitivity, and specificity of 83% (95% CI: 78-88%), 81% (95% CI: 76-87%), and 84% (95% CI: 79-89%), respectively.

CONCLUSION: AI models based on MRI features showed a high potential to discriminate between NMOSD and MS. However, heterogeneity in MR imaging, model evaluation, and reporting performance metrics, among other confounders, affected the reliability of our results. Well-designed studies on multicentric datasets, standardized imaging and evaluation protocols, and detailed transparent reporting of results are needed to reach optimal performance.

PMID:38781885 | DOI:10.1016/j.msard.2024.105682

Artif Intell Med. 2024 May 9;153:102888. doi: 10.1016/j.artmed.2024.102888. Online ahead of print.

ABSTRACT

BACKGROUND: When treating patients with coronary artery disease and concurrent renal concerns, we often encounter a conundrum: how to achieve a clearer view of vascular details while minimizing the contrast and radiation doses during percutaneous coronary intervention (PCI). Our goal is to use deep learning (DL) to create a real-time roadmap for guiding PCI. To this end, segmentation, a critical first step, paves the way for detailed vascular analysis. Unlike traditional supervised learning, which demands extensive labeling time and manpower, our strategy leans toward semi-supervised learning. This method not only economizes on labeling efforts but also aims at reducing contrast and radiation exposure.

METHODS AND RESULTS: CAG data sourced from eight tertiary centers in Taiwan, comprising 500 labeled and 8952 unlabeled images. Employing 400 labels for training and reserving 100 for validation, we built a U-Net based network within a teacher-student architecture. The initial teacher model was updated with 8952 unlabeled images inputted, employing a quality control strategy involving consistency regularization and RandAugment. The optimized teacher model produced pseudo-labels for label expansion, which were then utilized to train the final student model. We attained an average dice similarity coefficient of 0.9003 for segmentation, outperforming supervised learning methods with the same label count. Even with only 5 % labels for semi-supervised training, the results surpassed a supervised method with 100 % labels inputted. This semi-supervised approach's advantage extends beyond single-frame prediction, yielding consistently superior results in continuous angiography films.

CONCLUSIONS: High labeling cost hinders DL training. Semi-supervised learning, quality control, and pseudo-label expansion can overcome this. DL-assisted segmentation potentially provides a real-time PCI roadmap and further diminishes radiation and contrast doses.

PMID:38781870 | DOI:10.1016/j.artmed.2024.102888

Comput Methods Programs Biomed. 2024 May 11;252:108215. doi: 10.1016/j.cmpb.2024.108215. Online ahead of print.

ABSTRACT

BACKGROUND AND OBJECTIVE: Cell segmentation in bright-field histological slides is a crucial topic in medical image analysis. Having access to accurate segmentation allows researchers to examine the relationship between cellular morphology and clinical observations. Unfortunately, most segmentation methods known today are limited to nuclei and cannot segment the cytoplasm.

METHODS: We present a new network architecture Cyto R-CNN that is able to accurately segment whole cells (with both the nucleus and the cytoplasm) in bright-field images. We also present a new dataset CytoNuke, consisting of multiple thousand manual annotations of head and neck squamous cell carcinoma cells. Utilizing this dataset, we compared the performance of Cyto R-CNN to other popular cell segmentation algorithms, including QuPath's built-in algorithm, StarDist, Cellpose and a multi-scale Attention Deeplabv3+. To evaluate segmentation performance, we calculated AP50, AP75 and measured 17 morphological and staining-related features for all detected cells. We compared these measurements to the gold standard of manual segmentation using the Kolmogorov-Smirnov test.

RESULTS: Cyto R-CNN achieved an AP50 of 58.65% and an AP75 of 11.56% in whole-cell segmentation, outperforming all other methods (QuPath 19.46/0.91%; StarDist 45.33/2.32%; Cellpose 31.85/5.61%, Deeplabv3+ 3.97/1.01%). Cell features derived from Cyto R-CNN showed the best agreement to the gold standard (D¯=0.15) outperforming QuPath (D¯=0.22), StarDist (D¯=0.25), Cellpose (D¯=0.23) and Deeplabv3+ (D¯=0.33).

CONCLUSION: Our newly proposed Cyto R-CNN architecture outperforms current algorithms in whole-cell segmentation while providing more reliable cell measurements than any other model. This could improve digital pathology workflows, potentially leading to improved diagnosis. Moreover, our published dataset can be used to develop further models in the future.

PMID:38781811 | DOI:10.1016/j.cmpb.2024.108215

BMC Oral Health. 2024 May 23;24(1):601. doi: 10.1186/s12903-024-04307-5.

ABSTRACT

PROBLEM: Oral squamous cell carcinoma (OSCC) is the eighth most prevalent cancer globally, leading to the loss of structural integrity within the oral cavity layers and membranes. Despite its high prevalence, early diagnosis is crucial for effective treatment.

AIM: This study aimed to utilize recent advancements in deep learning for medical image classification to automate the early diagnosis of oral histopathology images, thereby facilitating prompt and accurate detection of oral cancer.

METHODS: A deep learning convolutional neural network (CNN) model categorizes benign and malignant oral biopsy histopathological images. By leveraging 17 pretrained DL-CNN models, a two-step statistical analysis identified the pretrained EfficientNetB0 model as the most superior. Further enhancement of EfficientNetB0 was achieved by incorporating a dual attention network (DAN) into the model architecture.

RESULTS: The improved EfficientNetB0 model demonstrated impressive performance metrics, including an accuracy of 91.1%, sensitivity of 92.2%, specificity of 91.0%, precision of 91.3%, false-positive rate (FPR) of 1.12%, F1 score of 92.3%, Matthews correlation coefficient (MCC) of 90.1%, kappa of 88.8%, and computational time of 66.41%. Notably, this model surpasses the performance of state-of-the-art approaches in the field.

CONCLUSION: Integrating deep learning techniques, specifically the enhanced EfficientNetB0 model with DAN, shows promising results for the automated early diagnosis of oral cancer through oral histopathology image analysis. This advancement has significant potential for improving the efficacy of oral cancer treatment strategies.

PMID:38783295 | DOI:10.1186/s12903-024-04307-5

Nucleic Acids Res. 2024 May 23:gkae412. doi: 10.1093/nar/gkae412. Online ahead of print.

ABSTRACT

Protein-protein interactions (PPIs) play a vital role in cellular functions and are essential for therapeutic development and understanding diseases. However, current predictive tools often struggle to balance efficiency and precision in predicting the effects of mutations on these complex interactions. To address this, we present DDMut-PPI, a deep learning model that efficiently and accurately predicts changes in PPI binding free energy upon single and multiple point mutations. Building on the robust Siamese network architecture with graph-based signatures from our prior work, DDMut, the DDMut-PPI model was enhanced with a graph convolutional network operated on the protein interaction interface. We used residue-specific embeddings from ProtT5 protein language model as node features, and a variety of molecular interactions as edge features. By integrating evolutionary context with spatial information, this framework enables DDMut-PPI to achieve a robust Pearson correlation of up to 0.75 (root mean squared error: 1.33 kcal/mol) in our evaluations, outperforming most existing methods. Importantly, the model demonstrated consistent performance across mutations that increase or decrease binding affinity. DDMut-PPI offers a significant advancement in the field and will serve as a valuable tool for researchers probing the complexities of protein interactions. DDMut-PPI is freely available as a web server and an application programming interface at https://biosig.lab.uq.edu.au/ddmut_ppi.

PMID:38783112 | DOI:10.1093/nar/gkae412

Nucleic Acids Res. 2024 May 23:gkae429. doi: 10.1093/nar/gkae429. Online ahead of print.

ABSTRACT

Native prokaryotic promoters share common sequence patterns, but are species dependent. For understudied species with limited data, it is challenging to predict the strength of existing promoters and generate novel promoters. Here, we developed PromoGen, a collection of nucleotide language models to generate species-specific functional promoters, across dozens of species in a data and parameter efficient way. Twenty-seven species-specific models in this collection were finetuned from the pretrained model which was trained on multi-species promoters. When systematically compared with native promoters, the Escherichia coli- and Bacillus subtilis-specific artificial PromoGen-generated promoters (PGPs) were demonstrated to hold all distribution patterns of native promoters. A regression model was developed to score generated either by PromoGen or by another competitive neural network, and the overall score of PGPs is higher. Encouraged by in silico analysis, we further experimentally characterized twenty-two B. subtilis PGPs, results showed that four of tested PGPs reached the strong promoter level while all were active. Furthermore, we developed a user-friendly website to generate species-specific promoters for 27 different species by PromoGen. This work presented an efficient deep-learning strategy for de novo species-specific promoter generation even with limited datasets, providing valuable promoter toolboxes especially for the metabolic engineering of understudied microorganisms.

PMID:38783063 | DOI:10.1093/nar/gkae429

NPJ Precis Oncol. 2024 May 23;8(1):115. doi: 10.1038/s41698-024-00592-z.

ABSTRACT

In the spectrum of colorectal tumors, microsatellite-stable (MSS) tumors with DNA polymerase ε (POLE) mutations exhibit a hypermutated profile, holding the potential to respond to immunotherapy similarly to their microsatellite-instable (MSI) counterparts. Yet, due to their rarity and the associated testing costs, systematic screening for these mutations is not commonly pursued. Notably, the histopathological phenotype resulting from POLE mutations is theorized to resemble that of MSI. This resemblance not only could facilitate their detection by a transformer-based Deep Learning (DL) system trained on MSI pathology slides, but also indicates the possibility for MSS patients with POLE mutations to access enhanced treatment options, which might otherwise be overlooked. To harness this potential, we trained a Deep Learning classifier on a large dataset with the ground truth for microsatellite status and subsequently validated its capabilities for MSI and POLE detection across three external cohorts. Our model accurately identified MSI status in both the internal and external resection cohorts using pathology images alone. Notably, with a classification threshold of 0.5, over 75% of POLE driver mutant patients in the external resection cohorts were flagged as "positive" by a DL system trained on MSI status. In a clinical setting, deploying this DL model as a preliminary screening tool could facilitate the efficient identification of clinically relevant MSI and POLE mutations in colorectal tumors, in one go.

PMID:38783059 | DOI:10.1038/s41698-024-00592-z

Sci Rep. 2024 May 23;14(1):11816. doi: 10.1038/s41598-024-62003-4.

ABSTRACT

Efficient Waste management plays a crucial role to ensure clean and green environment in the smart cities. This study investigates the critical role of efficient trash classification in achieving sustainable solid waste management within smart city environments. We conduct a comparative analysis of various trash classification methods utilizing deep learning models built on convolutional neural networks (CNNs). Leveraging the PyTorch open-source framework and the TrashBox dataset, we perform experiments involving ten unique deep neural network models. Our approach aims to maximize training accuracy. Through extensive experimentation, we observe the consistent superiority of the ResNext-101 model compared to others, achieving exceptional training, validation, and test accuracies. These findings illuminate the potential of CNN-based techniques in significantly advancing trash classification for optimized solid waste management within smart city initiatives. Lastly, this study presents a distributed framework based on federated learning that can be used to optimize the performance of a combination of CNN models for trash detection.

PMID:38783026 | DOI:10.1038/s41598-024-62003-4

Sci Rep. 2024 May 23;14(1):11774. doi: 10.1038/s41598-024-62294-7.

ABSTRACT

To develop and assess a deep learning (DL) pipeline to learn dynamic MR image reconstruction from publicly available natural videos (Inter4K). Learning was performed for a range of DL architectures (VarNet, 3D UNet, FastDVDNet) and corresponding sampling patterns (Cartesian, radial, spiral) either from true multi-coil cardiac MR data (N = 692) or from synthetic MR data simulated from Inter4K natural videos (N = 588). Real-time undersampled dynamic MR images were reconstructed using DL networks trained with cardiac data and natural videos, and compressed sensing (CS). Differences were assessed in simulations (N = 104 datasets) in terms of MSE, PSNR, and SSIM and prospectively for cardiac cine (short axis, four chambers, N = 20) and speech cine (N = 10) data in terms of subjective image quality ranking, SNR and Edge sharpness. Friedman Chi Square tests with post-hoc Nemenyi analysis were performed to assess statistical significance. In simulated data, DL networks trained with cardiac data outperformed DL networks trained with natural videos, both of which outperformed CS (p < 0.05). However, in prospective experiments DL reconstructions using both training datasets were ranked similarly (and higher than CS) and presented no statistical differences in SNR and Edge Sharpness for most conditions.The developed pipeline enabled learning dynamic MR reconstruction from natural videos preserving DL reconstruction advantages such as high quality fast and ultra-fast reconstructions while overcoming some limitations (data scarcity or sharing). The natural video dataset, code and pre-trained networks are made readily available on github.

PMID:38783018 | DOI:10.1038/s41598-024-62294-7

Commun Biol. 2024 May 23;7(1):625. doi: 10.1038/s42003-024-06318-y.

NO ABSTRACT

PMID:38783006 | DOI:10.1038/s42003-024-06318-y

Sci Rep. 2024 May 23;14(1):11785. doi: 10.1038/s41598-024-62527-9.

ABSTRACT

This paper presents a comparison of machine learning (ML) methods used for three-dimensional localization of partial discharges (PD) in a power transformer tank. The study examines ML and deep learning (DL) methods, ranging from support vector machines (SVM) to more complex approaches like convolutional neural networks (CNN). Multiple case studies are considered, each with different attributes, including sensor position, frequency content of the PD signal, and size of the transformer tank. The paper focuses on predicting the PD location in three-dimensional space using single-sensor electric field measurements. Various aspects of each method are analyzed, such as the input signal, core methodology, correlation coefficient between the predicted location and the actual location, and root mean square error (RMSE). These features are discussed and compared across the different methods. The results indicate that the CNN model exhibits superior performance in terms of location accuracy among the methods considered.

PMID:38782982 | DOI:10.1038/s41598-024-62527-9

Sci Rep. 2024 May 23;14(1):11810. doi: 10.1038/s41598-024-62394-4.

ABSTRACT

In this retrospective study, we aimed to assess the objective and subjective image quality of different reconstruction techniques and a deep learning-based software on non-contrast head computed tomography (CT) images. In total, 152 adult head CT scans (77 female, 75 male; mean age 69.4 ± 18.3 years) obtained from three different CT scanners using different protocols between March and April 2021 were included. CT images were reconstructed using filtered-back projection (FBP), iterative reconstruction (IR), and post-processed using a deep learning-based algorithm (PS). Post-processing significantly reduced noise in FBP-reconstructed images (up to 15.4% reduction) depending on the protocol, leading to improvements in signal-to-noise ratio of up to 19.7%. However, when deep learning-based post-processing was applied to FBP images compared to IR alone, the differences were inconsistent and partly non-significant, which appeared to be protocol or site specific. Subjective assessments showed no significant overall improvement in image quality for all reconstructions and post-processing. Inter-rater reliability was low and preferences varied. Deep learning-based denoising software improved objective image quality compared to FBP in routine head CT. A significant difference compared to IR was observed for only one protocol. Subjective assessments did not indicate a significant clinical impact in terms of improved subjective image quality, likely due to the low noise levels in full-dose images.

PMID:38782976 | DOI:10.1038/s41598-024-62394-4

Sci Rep. 2024 May 23;14(1):11768. doi: 10.1038/s41598-024-60355-5.

ABSTRACT

Accurate selection of sampling positions is critical in renal artery ultrasound examinations, and the potential of utilizing deep learning (DL) for assisting in this selection has not been previously evaluated. This study aimed to evaluate the effectiveness of DL object detection technology applied to color Doppler sonography (CDS) images in assisting sampling position selection. A total of 2004 patients who underwent renal artery ultrasound examinations were included in the study. CDS images from these patients were categorized into four groups based on the scanning position: abdominal aorta (AO), normal renal artery (NRA), renal artery stenosis (RAS), and intrarenal interlobular artery (IRA). Seven object detection models, including three two-stage models (Faster R-CNN, Cascade R-CNN, and Double Head R-CNN) and four one-stage models (RetinaNet, YOLOv3, FoveaBox, and Deformable DETR), were trained to predict the sampling position, and their predictive accuracies were compared. The Double Head R-CNN model exhibited significantly higher average accuracies on both parameter optimization and validation datasets (89.3 ± 0.6% and 88.5 ± 0.3%, respectively) compared to other methods. On clinical validation data, the predictive accuracies of the Double Head R-CNN model for all four types of images were significantly higher than those of the other methods. The DL object detection model shows promise in assisting inexperienced physicians in improving the accuracy of sampling position selection during renal artery ultrasound examinations.

PMID:38782971 | DOI:10.1038/s41598-024-60355-5

Br J Radiol. 2024 May 23:tqae108. doi: 10.1093/bjr/tqae108. Online ahead of print.

ABSTRACT

The licensing of antifibrotic therapy for fibrotic lung diseases, including idiopathic pulmonary fibrosis IPF has created an urgent need for reliable biomarkers to predict disease progression and treatment response. Some patients experience stable disease trajectories, while others deteriorate rapidly, making treatment decisions challenging. High-resolution chest CT has become crucial for diagnosis, but visual assessments by radiologists suffer from low reproducibility and high interobserver variability. To address these issues, computer-based image analysis, called quantitative CT, has emerged. However, many quantitative CT methods rely on human input for training, therefore potentially incorporating human error into computer training. Rapid advances in artificial intelligence, specifically deep learning, aim to overcome this limitation by enabling autonomous quantitative analysis. While promising, deep learning also presents challenges including the need to minimize algorithm biases, ensuring explainability, and addressing accessibility and ethical concerns. This review explores the development and application of deep learning in improving the imaging process for fibrotic lung disease.

PMID:38781513 | DOI:10.1093/bjr/tqae108

Science. 2024 May 24;384(6698):eadh0559. doi: 10.1126/science.adh0559. Epub 2024 May 24.

ABSTRACT

Nucleotide changes in gene regulatory elements are important determinants of neuronal development and diseases. Using massively parallel reporter assays in primary human cells from mid-gestation cortex and cerebral organoids, we interrogated the cis-regulatory activity of 102,767 open chromatin regions, including thousands of sequences with cell type-specific accessibility and variants associated with brain gene regulation. In primary cells, we identified 46,802 active enhancer sequences and 164 variants that alter enhancer activity. Activity was comparable in organoids and primary cells, suggesting that organoids provide an adequate model for the developing cortex. Using deep learning we decoded the sequence basis and upstream regulators of enhancer activity. This work establishes a comprehensive catalog of functional gene regulatory elements and variants in human neuronal development.

PMID:38781390 | DOI:10.1126/science.adh0559

Science. 2024 May 24;384(6698):eadp7977. doi: 10.1126/science.adp7977. Epub 2024 May 23.

ABSTRACT

"AI-Powered Forecasting" was recently on the cover of Science, highlighting a new deep learning model for much faster and more accurate weather forecasting. Known as GraphCast, it outperformed the gold-standard system and had an accuracy of 99.7% for tropospheric predictions, the most important forecasting region that is closest to Earth's surface. Better warnings for extreme weather events such as hurricanes and cyclones will help save lives. The parallel in medicine is forecasting specific, actionable, high risk for individuals to prevent diseases or severe acute events. But we don't have a gold standard for predicting health outcomes. That is hopefully about to change.

PMID:38781357 | DOI:10.1126/science.adp7977