Quant Imaging Med Surg. 2024 May 1;14(5):3534-3543. doi: 10.21037/qims-23-1488. Epub 2024 Apr 11.

ABSTRACT

BACKGROUND: Deep-learning-based reconstruction (DLR) improves the quality of magnetic resonance (MR) images which allows faster acquisitions. The aim of this study was to compare the image quality of standard and accelerated T2 weighted turbo-spin-echo (TSE) images of the prostate reconstructed with and without DLR and to find associations between perceived image quality and calculated image characteristics.

METHODS: In a cohort of 47 prospectively enrolled consecutive patients referred for bi-parametric prostate magnetic resonance imaging (MRI), two T2-TSE acquisitions in the transverse plane were acquired on a 3T scanner-a standard T2-TSE sequence and a short sequence accelerated by a factor of two using compressed sensing (CS). The images were reconstructed with and without DLR in super-resolution mode. The image quality was rated in six domains. Signal-to-noise ratio (SNR), and image sharpness were measured.

RESULTS: The mean acquisition time was 281±23 s for the standard and 140±12 s for the short acquisition (P<0.0001). DLR images had higher sharpness compared to non-DLR (P<0.001). Short and short-DLR had lower SNR than the standard and standard-DLR (P<0.001). The perceived image quality of short-DLR was rated better in all categories compared to the standard sequence (P<0.001 to P=0.004). All domains of subjective evaluation were correlated with measured image sharpness (P<0.001).

CONCLUSIONS: T2-TSE acquisition of the prostate accelerated using CS combined with DLR reconstruction provides images with increased sharpness that have a superior quality as perceived by human readers compared to standard T2-TSE. The perceived image quality is correlated with measured image contrast.

PMID:38720867 | PMC:PMC11074762 | DOI:10.21037/qims-23-1488

Quant Imaging Med Surg. 2024 May 1;14(5):3707-3716. doi: 10.21037/qims-23-1384. Epub 2024 Apr 18.

ABSTRACT

BACKGROUND: Automated diagnosis of infant hip dysplasia is heavily affected by the individual differences among infants and ultrasound machines.

METHODS: Hip sonographic images of 493 infants from various ultrasound machines were collected in the Department of Orthopedics in Yangzhou Maternal and Child Health Care Service Centre. Herein, we propose a semi-supervised learning method based on a feature pyramid network (FPN) and a contrastive learning scheme based on a Siamese architecture. A large amount of unlabeled data of ultrasound images was used via the Siamese network in the pre-training step, and then a small amount of annotated data for anatomical structures was adopted to train the model for landmark identification and standard plane recognition. The method was evaluated on our collected dataset.

RESULTS: The method achieved a mean Dice similarity coefficient (DSC) of 0.7873 and a mean Hausdorff distance (HD) of 5.0102 in landmark identification, compared to the model without contrastive learning, which had a mean DSC of 0.7734 and a mean HD of 6.1586. The accuracy, precision, and recall of standard plane recognition were 95.4%, 91.64%, and 94.86%, respectively. The corresponding area under the curve (AUC) was 0.982.

CONCLUSIONS: This study proposes a semi-supervised deep learning method following Graf's principle, which can better utilize a large volume of ultrasound images from various devices and infants. This method can identify the landmarks of infant hips more accurately than manual operators, thereby improving the efficiency of diagnosis of infant hip dysplasia.

PMID:38720865 | PMC:PMC11074738 | DOI:10.21037/qims-23-1384

Quant Imaging Med Surg. 2024 May 1;14(5):3432-3446. doi: 10.21037/qims-23-1379. Epub 2024 Apr 10.

ABSTRACT

BACKGROUND: Image-based assessment of prostate cancer (PCa) is increasingly emphasized in the diagnostic workflow for selecting biopsy targets and possibly predicting clinically significant prostate cancer (csPCa). Assessment is based on Prostate Imaging-Reporting and Data System (PI-RADS) which is largely dependent on T2-weighted image (T2WI) and diffusion weighted image (DWI). This study aims to determine whether deep learning reconstruction (DLR) can improve the image quality of DWI and affect the assessment of PI-RADS ≥4 in patients with PCa.

METHODS: In this retrospective study, 3.0T post-biopsy prostate magnetic resonance imaging (MRI) of 70 patients with PCa in Korea University Ansan Hospital from November 2021 to July 2022 was reconstructed with and without using DLR. Four DWI image sets were made: (I) conventional DWI (CDWI): DWI with acceleration factor 2 and conventional parallel imaging reconstruction, (II) DL1: DWI with acceleration factor 2 using DLR, (III) DL2: DWI with acceleration factor 3 using DLR, and (IV) DL3: DWI with acceleration factor 3 and half average b-value using DLR. Apparent diffusion coefficient (ADC) value, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were measured by one reviewer, while two reviewers independently assessed overall image quality, noise, and lesion conspicuity using a four-point visual scoring system from each DWI image set. Two reviewers also performed PI-RADSv2.1 scoring on lesions suspected of malignancy.

RESULTS: A total of 70 patients (mean age, 70.8±9.7 years) were analyzed. The image acquisition time was 4:46 min for CDWI and DL1, 3:40 min for DL2, and 2:00 min for DL3. DL1 and DL2 images resulted in better lesion conspicuity compared to CDWI images assessed by both readers (P<0.05). DLR resulted in a significant increase in SNR, from 38.4±14.7 in CDWI to 56.9±21.0 in DL1. CNR increased from 25.1±11.5 in CDWI to 43.1±17.8 in DL1 (P<0.001). PI-RADS v2.1 scoring for PCa lesions was more agreeable with the DL1 reconstruction method than with CDWI (κ value CDWI, DL1; 0.40, 0.61, respectively). A statistically significant number of lesions were upgraded from PI-RADS <4 in CDWI image to PI-RADS ≥4 in DL1 images for both readers (P<0.05). Most of the PI-RADS upgraded lesions were from higher than unfavorable intermediate-risk groups according to the 2023 National Comprehensive Cancer Network guidelines with statistically significant difference of marginal probability in DL1 and DL2 for both readers (P<0.05).

CONCLUSIONS: DLR in DWI for PCa can provide options for improving image quality with a significant impact on PI-RADS evaluation or about a 23% reduction in acquisition time without compromising image quality.

PMID:38720859 | PMC:PMC11074768 | DOI:10.21037/qims-23-1379

Quant Imaging Med Surg. 2024 May 1;14(5):3619-3627. doi: 10.21037/qims-23-1864. Epub 2024 Apr 26.

ABSTRACT

BACKGROUND: Cardiac ultrasound is one of the most important examinations in cardiovascular medicine, but the technical requirements for the operator are relatively high, which to some extent affects the scope of its use. This study was dedicated to investigating the agreement of ejection fraction between coronary computed tomography (CT) and cardiac ultrasound and diagnostic performance in evaluating the clinical diagnosis of patients with chronic heart failure.

METHODS: We conducted a single-center-based retrospective study including 343 consecutive patients enrolled between January 2019 to April 2020, all of whom presented with suspected symptoms of heart failure within one month. All enrolled cases performed cardiac ultrasound and coronary CT scans. The CT images were analyzed using accurate left ventricle (AccuLV) artificial intelligence (AI) software to calculate the ejection fraction-computed tomography (EF-CT) and it was compared with the ejection fraction (EF) obtained based on ultrasound. Cardiac insufficiency was determined if the EF measured by ultrasound was below 50%. Diagnostic performance analysis, correlation analysis and Bland-Altman plot were used to compare agreement between EF-CT and CT.

RESULTS: Of the 319 successfully performed patients, 220 (69%) were identified as cardiac insufficiency. Quantitative consistency analysis showed a good correlation between EF-CT and EF values in all cases (R square =0.704, r=0.837). Bland-Altman analysis showed mean bias of 6.6%, mean percentage error of 27.5% and 95% limit of agreement of -17% to 30% between EF and EF-CT. The results of the qualitative diagnostic study showed that the sensitivity and specificity of EF measured by coronary CT reached a high level of 91% [95% confidence interval (CI): 86-94%], and the positive diagnostic value was up to 96% (95% CI: 92-98%).

CONCLUSIONS: The EF-CT and EF have excellent agreement, and AccuLV-based AI left ventricular function analysis software perhaps can be used as a clinical diagnostic reference.

PMID:38720849 | PMC:PMC11074755 | DOI:10.21037/qims-23-1864

Pathobiology. 2024 May 8. doi: 10.1159/000539010. Online ahead of print.

ABSTRACT

Lymph node metastasis is one of the most common ways of tumour metastasis. The presence or absence of lymph node involvement influences the cancer's stage, therapy, and prognosis. The integration of artificial intelligence systems in the histopathological diagnosis of lymph nodes after surgery is urgent. Here, we propose a pan-origin lymph node cancer metastasis detection system. The system is trained by over 700 whole slide images and is composed of two deep learning models to locate the lymph nodes and detect cancers. It achieved a area under the receiver operating characteristic (ROC) curve (AUC) of 0.958, with a 95.2% sensitivity and 72.2% specificity, on 1,402 whole-slide images (WSIs) from 49 organs at the National Cancer Center, China. Moreover, we demonstrated that the system could perform robustly with 1,051 WSIs from 52 organs from another medical center, with a AUC of 0.925. Our research represents a step forward in a pan-origin lymph node metastasis detection system, providing accurate pathological guidance by reducing the probability of missed diagnosis in routine clinical practice.

PMID:38718783 | DOI:10.1159/000539010

Med Image Anal. 2024 May 3;95:103188. doi: 10.1016/j.media.2024.103188. Online ahead of print.

ABSTRACT

In medical image diagnosis, fairness has become increasingly crucial. Without bias mitigation, deploying unfair AI would harm the interests of the underprivileged population and potentially tear society apart. Recent research addresses prediction biases in deep learning models concerning demographic groups (e.g., gender, age, and race) by utilizing demographic (sensitive attribute) information during training. However, many sensitive attributes naturally exist in dermatological disease images. If the trained model only targets fairness for a specific attribute, it remains unfair for other attributes. Moreover, training a model that can accommodate multiple sensitive attributes is impractical due to privacy concerns. To overcome this, we propose a method enabling fair predictions for sensitive attributes during the testing phase without using such information during training. Inspired by prior work highlighting the impact of feature entanglement on fairness, we enhance the model features by capturing the features related to the sensitive and target attributes and regularizing the feature entanglement between corresponding classes. This ensures that the model can only classify based on the features related to the target attribute without relying on features associated with sensitive attributes, thereby improving fairness and accuracy. Additionally, we use disease masks from the Segment Anything Model (SAM) to enhance the quality of the learned feature. Experimental results demonstrate that the proposed method can improve fairness in classification compared to state-of-the-art methods in two dermatological disease datasets.

PMID:38718715 | DOI:10.1016/j.media.2024.103188

Sci Rep. 2024 May 8;14(1):10560. doi: 10.1038/s41598-024-61286-x.

ABSTRACT

The research on video analytics especially in the area of human behavior recognition has become increasingly popular recently. It is widely applied in virtual reality, video surveillance, and video retrieval. With the advancement of deep learning algorithms and computer hardware, the conventional two-dimensional convolution technique for training video models has been replaced by three-dimensional convolution, which enables the extraction of spatio-temporal features. Specifically, the use of 3D convolution in human behavior recognition has been the subject of growing interest. However, the increased dimensionality has led to challenges such as the dramatic increase in the number of parameters, increased time complexity, and a strong dependence on GPUs for effective spatio-temporal feature extraction. The training speed can be considerably slow without the support of powerful GPU hardware. To address these issues, this study proposes an Adaptive Time Compression (ATC) module. Functioning as an independent component, ATC can be seamlessly integrated into existing architectures and achieves data compression by eliminating redundant frames within video data. The ATC module effectively reduces GPU computing load and time complexity with negligible loss of accuracy, thereby facilitating real-time human behavior recognition.

PMID:38720020 | DOI:10.1038/s41598-024-61286-x

Sci Rep. 2024 May 8;14(1):10569. doi: 10.1038/s41598-024-60901-1.

ABSTRACT

Within the medical field of human assisted reproductive technology, a method for interpretable, non-invasive, and objective oocyte evaluation is lacking. To address this clinical gap, a workflow utilizing machine learning techniques has been developed involving automatic multi-class segmentation of two-dimensional images, morphometric analysis, and prediction of developmental outcomes of mature denuded oocytes based on feature extraction and clinical variables. Two separate models have been developed for this purpose-a model to perform multiclass segmentation, and a classifier model to classify oocytes as likely or unlikely to develop into a blastocyst (Day 5-7 embryo). The segmentation model is highly accurate at segmenting the oocyte, ensuring high-quality segmented images (masks) are utilized as inputs for the classifier model (mask model). The mask model displayed an area under the curve (AUC) of 0.63, a sensitivity of 0.51, and a specificity of 0.66 on the test set. The AUC underwent a reduction to 0.57 when features extracted from the ooplasm were removed, suggesting the ooplasm holds the information most pertinent to oocyte developmental competence. The mask model was further compared to a deep learning model, which also utilized the segmented images as inputs. The performance of both models combined in an ensemble model was evaluated, showing an improvement (AUC 0.67) compared to either model alone. The results of this study indicate that direct assessments of the oocyte are warranted, providing the first objective insights into key features for developmental competence, a step above the current standard of care-solely utilizing oocyte age as a proxy for quality.

PMID:38719918 | DOI:10.1038/s41598-024-60901-1

Sci Rep. 2024 May 8;14(1):10609. doi: 10.1038/s41598-024-59558-7.

ABSTRACT

We propose a novel framework that combines state-of-the-art deep learning approaches with pre- and post-processing algorithms for particle detection in complex/heterogeneous backgrounds common in the manufacturing domain. Traditional methods, like size analyzers and those based on dilution, image processing, or deep learning, typically excel with homogeneous backgrounds. Yet, they often fall short in accurately detecting particles against the intricate and varied backgrounds characteristic of heterogeneous particle-substrate (HPS) interfaces in manufacturing. To address this, we've developed a flexible framework designed to detect particles in diverse environments and input types. Our modular framework hinges on model selection and AI-guided particle detection as its core, with preprocessing and postprocessing as integral components, creating a four-step process. This system is versatile, allowing for various preprocessing, AI model selections, and post-processing strategies. We demonstrate this with an entrainment-based particle delivery method, transferring various particles onto substrates that mimic the HPS interface. By altering particle and substrate properties (e.g., material type, size, roughness, shape) and process parameters (e.g., capillary number) during particle entrainment, we capture images under different ambient lighting conditions, introducing a range of HPS background complexities. In the preprocessing phase, we apply image enhancement and sharpening techniques to improve detection accuracy. Specifically, image enhancement adjusts the dynamic range and histogram, while sharpening increases contrast by combining the high pass filter output with the base image. We introduce an image classifier model (based on the type of heterogeneity), employing Transfer Learning with MobileNet as a Model Selector, to identify the most appropriate AI model (i.e., YOLO model) for analyzing each specific image, thereby enhancing detection accuracy across particle-substrate variations. Following image classification based on heterogeneity, the relevant YOLO model is employed for particle identification, with a distinct YOLO model generated for each heterogeneity type, improving overall classification performance. In the post-processing phase, domain knowledge is used to minimize false positives. Our analysis indicates that the AI-guided framework maintains consistent precision and recall across various HPS conditions, with the harmonic mean of these metrics comparable to those of individual AI model outcomes. This tool shows potential for advancing in-situ process monitoring across multiple manufacturing operations, including high-density powder-based 3D printing, powder metallurgy, extreme environment coatings, particle categorization, and semiconductor manufacturing.

PMID:38719876 | DOI:10.1038/s41598-024-59558-7

Sci Rep. 2024 May 8;14(1):9591. doi: 10.1038/s41598-024-59619-x.

ABSTRACT

Vaping involves the heating of chemical solutions (e-liquids) to high temperatures prior to lung inhalation. A risk exists that these chemicals undergo thermal decomposition to new chemical entities, the composition and health implications of which are largely unknown. To address this concern, a graph-convolutional neural network (NN) model was used to predict pyrolysis reactivity of 180 e-liquid chemical flavours. The output of this supervised machine learning approach was a dataset of probability ranked pyrolysis transformations and their associated 7307 products. To refine this dataset, the molecular weight of each NN predicted product was automatically correlated with experimental mass spectrometry (MS) fragmentation data for each flavour chemical. This blending of deep learning methods with experimental MS data identified 1169 molecular weight matches that prioritized these compounds for further analysis. The average number of discrete matches per flavour between NN predictions and MS fragmentation was 6.4 with 92.8% of flavours having at least one match. Globally harmonized system classifications for NN/MS matches were extracted from PubChem, revealing that 127 acute toxic, 153 health hazard and 225 irritant classifications were predicted. This approach may reveal the longer-term health risks of vaping in advance of clinical diseases emerging in the general population.

PMID:38719814 | DOI:10.1038/s41598-024-59619-x

Otolaryngol Clin North Am. 2024 May 7:S0030-6665(24)00059-8. doi: 10.1016/j.otc.2024.04.002. Online ahead of print.

ABSTRACT

Technological advancements in laryngology, broncho-esophagology, and sleep surgery have enabled the collection of increasing amounts of complex data for diagnosis and treatment of voice, swallowing, and sleep disorders. Clinicians face challenges in efficiently synthesizing these data for personalized patient care. Artificial intelligence (AI), specifically machine learning and deep learning, offers innovative solutions for processing and interpreting these data, revolutionizing diagnosis and management in these fields, and making care more efficient and effective. In this study, we review recent AI-based innovations in the fields of laryngology, broncho-esophagology, and sleep surgery.

PMID:38719714 | DOI:10.1016/j.otc.2024.04.002

AJNR Am J Neuroradiol. 2024 May 7:ajnr.A8334. doi: 10.3174/ajnr.A8334. Online ahead of print.

ABSTRACT

BACKGROUND AND PURPOSE: Intracranial steno-occlusive lesions are responsible for acute ischemic stroke. However, the clinical benefits of artificial intelligence-based methods for detecting pathologic lesions in intracranial arteries have not been evaluated. We aimed to validate the clinical utility of an artificial intelligence model for detecting steno-occlusive lesions in the intracranial arteries.

MATERIALS AND METHODS: Overall, 138 TOF-MRA images were collected from two institutions, which served as internal (n = 62) and external (n = 76) test sets, respectively. Each study was reviewed by five radiologists (two neuroradiologists and three radiology residents) to compare the usage and non-usage of our proposed artificial intelligence model for TOF-MRA interpretation. They identified the steno-occlusive lesions and recorded their reading time. Observer performance was assessed using the area under the Jackknife free-response receiver operating characteristic curve and reading time for comparison.

RESULTS: The average area under the Jackknife free-response receiver operating characteristic curve for the five radiologists demonstrated an improvement from 0.70 without artificial intelligence to 0.76 with artificial intelligence (P = .027). Notably, this improvement was most pronounced among the three radiology residents, whose performance metrics increased from 0.68 to 0.76 (P = .002). Despite an increased reading time upon using artificial intelligence, there was no significant change among the readings by radiology residents. Moreover, the use of artificial intelligence resulted in improved inter-observer agreement among the reviewers (the intraclass correlation coefficient increased from 0.734 to 0.752).

CONCLUSIONS: Our proposed artificial intelligence model offers a supportive tool for radiologists, potentially enhancing the accuracy of detecting intracranial steno-occlusion lesions on TOF-MRA. Less-experienced readers may benefit the most from this model.ABBREVIATIONS: AI = Artificial intelligence; AUC = Area under the receiver operating characteristic curve; AUFROC = Area under the Jackknife free-response receiver operating characteristic curve; DL = Deep learning; ICC = Intraclass correlation coefficient; IRB = Institutional Review Boards; JAFROC = Jackknife free-response receiver operating characteristic.

PMID:38719612 | DOI:10.3174/ajnr.A8334

J Stomatol Oral Maxillofac Surg. 2024 May 6:101912. doi: 10.1016/j.jormas.2024.101912. Online ahead of print.

ABSTRACT

This study aimed to assess the diagnostic performance of a machine learning approach that utilized radiomic features extracted from Cone Beam Computer Tomography (CBCT) images and inflammatory biomarkers for distinguishing between Dentigerous Cysts (DCs), Odontogenic Keratocysts (OKCs), and Unicystic Ameloblastomas (UAs). This retrospective study involves 103 patients who underwent jaw lesion surgery in the Maxillofacial Surgery Unit of Federico II University Of Naples between January 2018 and January 2023. Nonparametric Wilcoxon-Mann-Whitney and Kruskal Wallis tests were used for continuous variables. Linear and non-logistic regression models (LRM and NLRM) were employed, along with machine learning techniques such as decision tree (DT), k-nearest neighbor (KNN), and support vector machine (SVM), to predict the outcomes. When individual inflammatory biomarkers were considered alone, their ability to differentiate between OKCs, UAs, and DCs was below 50% accuracy. However, a linear regression model combining four inflammatory biomarkers achieved an accuracy of 95% and an AUC of 0.96. The accuracy of single radiomics predictors was lower than that of inflammatory biomarkers, with an AUC of 0.83. The Fine Tree model, utilizing NLR, SII, and one radiomic feature, achieved an accuracy of 94.3% (AUC = 0.95) on the training and testing sets, and a validation set accuracy of 100%. The Fine Tree model demonstrated the capability to discriminate between OKCs, UAs, and DCs. However, the LRM utilizing four inflammatory biomarkers proved to be the most effective algorithm for distinguishing between OKCs, UAs, and DCs.

PMID:38719192 | DOI:10.1016/j.jormas.2024.101912

Nature. 2024 May 8. doi: 10.1038/s41586-024-07487-w. Online ahead of print.

ABSTRACT

The introduction of AlphaFold 21 has spurred a revolution in modelling the structure of proteins and their interactions, enabling a huge range of applications in protein modelling and design2-6. In this paper, we describe our AlphaFold 3 model with a substantially updated diffusion-based architecture, which is capable of joint structure prediction of complexes including proteins, nucleic acids, small molecules, ions, and modified residues. The new AlphaFold model demonstrates significantly improved accuracy over many previous specialised tools: far greater accuracy on protein-ligand interactions than state of the art docking tools, much higher accuracy on protein-nucleic acid interactions than nucleic-acid-specific predictors, and significantly higher antibody-antigen prediction accuracy than AlphaFold-Multimer v2.37,8. Together these results show that high accuracy modelling across biomolecular space is possible within a single unified deep learning framework.

PMID:38718835 | DOI:10.1038/s41586-024-07487-w

Phys Med Biol. 2024 May 8. doi: 10.1088/1361-6560/ad48f6. Online ahead of print.

ABSTRACT

To evaluate the feasibility of using a deep learning dose prediction approach to identify patients who could benefit most from proton therapy based on the normal tissue complication probability (NTCP) model.&#xD;Approach: Two 3D UNets were established to predict photon and proton doses. A dataset of 95 patients with localized prostate cancer was randomly partitioned into 55, 10, and 30 for training, validation, and testing, respectively. We selected NTCP models for late rectum bleeding and acute urinary urgency of grade 2 or higher to quantify the benefit of proton therapy. Propagated uncertainties of predicted ΔNTCPs resulting from the dose prediction errors were calculated. Patient selection accuracies for a single endpoint and a composite evaluation were assessed under different ΔNTCP thresholds.&#xD;Main results: Our deep learning-based dose prediction technique can reduce the time spent on plan comparison from approximately 2 days to as little as 5 seconds. The expanded uncertainty of predicted ΔNTCPs for rectum and bladder endpoints propagated from the dose prediction error were 0.0042 and 0.0016, respectively, which is less than one-third of the acceptable tolerance. The averaged selection accuracies for rectum bleeding, urinary urgency, and composite evaluation were 90%, 93.5%, and 93.5%, respectively.&#xD;Significance: Our study demonstrates that deep learning dose prediction and NTCP evaluation scheme could distinguish the NTCP differences between photon and proton treatment modalities. In addition, the dose prediction uncertainty does not significantly influence the decision accuracy of NTCP-based patient selection for proton therapy. Therefore, automated deep learning dose prediction and NTCP evaluation schemes can potentially be used to screen large patient populations and to avoid unnecessary delays in the start of prostate cancer radiotherapy in the future.

PMID:38718814 | DOI:10.1088/1361-6560/ad48f6

J Neural Eng. 2024 May 8. doi: 10.1088/1741-2552/ad48b9. Online ahead of print.

ABSTRACT

Objective&#xD;The objective of this study is to investigate the application of various channel attention mechanisms within the domain of brain-computer interface (BCI) for motor imagery decoding. Channel attention mechanisms can be seen as a powerful evolution of spatial filters traditionally used for motor imagery decoding. This study systematically compares such mechanisms by integrating them into a lightweight architecture framework to evaluate their impact. &#xD;Approach&#xD;We carefully construct a straightforward and lightweight baseline architecture designed to seamlessly integrate different channel attention mechanisms. This approach is contrary to previous works which only investigate one attention mechanism and usually build a very complex, sometimes nested architecture. Our framework allows us to evaluate and compare the impact of different attention mechanisms under the same circumstances. The easy integration of different channel attention mechanisms as well as the low computational complexity enables us to conduct a wide range of experiments on four datasets to thoroughly assess the effectiveness of the baseline model and the attention mechanisms. &#xD;Results&#xD;Our experiments demonstrate the strength and generalizability of our architecture framework as well as how channel attention mechanisms can improve the performance while maintaining the small memory footprint and low computational complexity of our baseline architecture.&#xD;Significance&#xD;Our architecture emphasizes simplicity, offering easy integration of channel attention mechanisms, while maintaining a high degree of generalizability across datasets, making it a versatile and efficient solution for EEG motor imagery decoding within brain-computer interfaces.

PMID:38718788 | DOI:10.1088/1741-2552/ad48b9

J Am Med Inform Assoc. 2024 May 8:ocae118. doi: 10.1093/jamia/ocae118. Online ahead of print.

ABSTRACT

OBJECTIVE: Social media-based public health research is crucial for epidemic surveillance, but most studies identify relevant corpora with keyword-matching. This study develops a system to streamline the process of curating colloquial medical dictionaries. We demonstrate the pipeline by curating a UMLS-colloquial symptom dictionary from COVID-19-related tweets as proof of concept.

METHODS: COVID-19-related tweets from February 1, 2020, to April 30, 2022 were used. The pipeline includes three modules: a named entity recognition module to detect symptoms in tweets; an entity normalization module to aggregate detected entities; and a mapping module that iteratively maps entities to Unified Medical Language System concepts. A random 500 entity sample were drawn from the final dictionary for accuracy validation. Additionally, we conducted a symptom frequency distribution analysis to compare our dictionary to a pre-defined lexicon from previous research.

RESULTS: We identified 498,480 unique symptom entity expressions from the tweets. Pre-processing reduces the number to 18,226. The final dictionary contains 38,175 unique expressions of symptoms that can be mapped to 966 UMLS concepts (accuracy = 95%). Symptom distribution analysis found that our dictionary detects more symptoms and is effective at identifying psychiatric disorders like anxiety and depression, often missed by pre-defined lexicons.

CONCLUSIONS: This study advances public health research by implementing a novel, systematic pipeline for curating symptom lexicons from social media data. The final lexicon's high accuracy, validated by medical professionals, underscores the potential of this methodology to reliably interpret and categorize vast amounts of unstructured social media data into actionable medical insights across diverse linguistic and regional landscapes.

PMID:38718216 | DOI:10.1093/jamia/ocae118

Bioinformatics. 2024 May 8:btae134. doi: 10.1093/bioinformatics/btae134. Online ahead of print.

ABSTRACT

MOTIVATION: Combination drug therapies are effective treatments for cancer. However, the genetic heterogeneity of the patients and exponentially large space of drug pairings pose significant challenges for finding the right combination for a specific patient. Current in silico prediction methods can be instrumental in reducing the vast number of candidate drug combinations. However, existing powerful methods are trained with cancer cell line gene expression data, which limits their applicability in clinical settings. While synergy measurements on cell line models are available at large scale, patient-derived samples are too few to train a complex model. On the other hand, patient-specific single-drug response data are relatively more available.

RESULTS: In this work, we propose a deep learning framework, Personalized Deep Synergy Predictor (PDSP), that enables us to use the patient-specific single drug response data for customizing patient drug synergy predictions. PDSP is first trained to learn synergy scores of drug pairs and their single drug responses for a given cell line using drug structures and large scale cell line gene expression data. Then, the model is fine-tuned for patients with their patient gene expression data and associated single drug response measured on the patient ex vivo samples. In this study, we evaluate PDSP on data from three leukemia patients and observe that it improves the prediction accuracy by 27% compared to models trained on cancer cell line data.

AVAILABILITY: PDSP is available at https://github.com/hikuru/PDSP.

SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

PMID:38718189 | DOI:10.1093/bioinformatics/btae134

PLoS One. 2024 May 8;19(5):e0293786. doi: 10.1371/journal.pone.0293786. eCollection 2024.

ABSTRACT

α-zeins are amphiphilic maize seed storage proteins with material properties suitable for a multitude of applications e.g., in renewable plastics, foods, therapeutics and additive manufacturing (3D-printing). To exploit their full potential, molecular-level insights are essential. The difficulties in experimental atomic-resolution characterization of α-zeins have resulted in a diversity of published molecular models. However, deep-learning α-zein models are largely unexplored. Therefore, this work studies an AlphaFold2 (AF2) model of a highly expressed α-zein using molecular dynamics (MD) simulations. The sequence of the α-zein cZ19C2 gave a loosely packed AF2 model with 7 α-helical segments connected by turns/loops. Compact tertiary structure was limited to a C-terminal bundle of three α-helices, each showing notable agreement with a published consensus sequence. Aiming to chart possible α-zein conformations in practically relevant solvents, rather than the native solid-state, the AF2 model was subjected to MD simulations in water/ethanol mixtures with varying ethanol concentrations. Despite giving structurally diverse endpoints, the simulations showed several patterns: In water and low ethanol concentrations, the model rapidly formed compact globular structures, largely preserving the C-terminal bundle. At ≥ 50 mol% ethanol, extended conformations prevailed, consistent with previous SAXS studies. Tertiary structure was partially stabilized in water and low ethanol concentrations, but was disrupted in ≥ 50 mol% ethanol. Aggregated results indicated minor increases in helicity with ethanol concentration. β-sheet content was consistently low (∼1%) across all conditions. Beyond structural dynamics, the rapid formation of branched α-zein aggregates in aqueous environments was highlighted. Furthermore, aqueous simulations revealed favorable interactions between the protein and the crosslinking agent glycidyl methacrylate (GMA). The proximity of GMA epoxide carbons and side chain hydroxyl oxygens simultaneously suggested accessible reactive sites in compact α-zein conformations and pre-reaction geometries for methacrylation. The findings may assist in expanding the applications of these technologically significant proteins, e.g., by guiding chemical modifications.

PMID:38718010 | DOI:10.1371/journal.pone.0293786

IEEE Trans Vis Comput Graph. 2024 May 8;PP. doi: 10.1109/TVCG.2024.3398289. Online ahead of print.

ABSTRACT

The digital construction of cultural heritage promotes communication and sharing of digital cultural resources across time and space. Digital storage serves as the foundation for the digital construction of cultural artifacts. In the digital storage of Dunhuang murals, image stitching plays a critical role in restoring the complete image of the cave murals. Traditional image stitching methods are constrained by the detection accuracy of feature points and are not fit for stitching low-texture murals. Despite deep learning-based image stitching methods, parallax misalignment and ghosting are still prevalent issues. For this reason, we perform the first Dunhuang mural stitching based on deep learning in this paper. This is in response to the need for digitizing and storing Dunhuang murals. Two mural stitching datasets are constructed, and we design a progressive regression image alignment network and a feature differential reconstruction soft-coded seam stitching network. We also introduce a soft-coded seam quality evaluation method. The algorithm presented in this paper achieves state-of-the-art alignment and stitching performance in the mural stitching task through unsupervised learning with a smaller number of model parameters, which provides technical support for the digitization and preservation of Dunhuang murals. The codes and models will be available at https://github.com/MmelodYy/DunHuangStitch.

PMID:38717890 | DOI:10.1109/TVCG.2024.3398289