Int Forum Allergy Rhinol. 2024 Jul 16. doi: 10.1002/alr.23411. Online ahead of print.

ABSTRACT

This follow-up dual-institutional and longitudinal study further evaluated for underlying gender biases in LORs for rhinology fellowship. Explicit and implicit linguistic gender bias was found, heavily favoring male applicants.

PMID:39010845 | DOI:10.1002/alr.23411

J Cachexia Sarcopenia Muscle. 2024 Jul 15. doi: 10.1002/jcsm.13527. Online ahead of print.

ABSTRACT

BACKGROUND: There is increasing evidence that myosteatosis, which is currently not assessed in clinical routine, plays an important role in risk estimation in individuals with impaired glucose metabolism, as it is associated with the progression of insulin resistance. With advances in artificial intelligence, automated and accurate algorithms have become feasible to fill this gap.

METHODS: In this retrospective study, we developed and tested a fully automated deep learning model using data from two prospective cohort studies (German National Cohort [NAKO] and Cooperative Health Research in the Region of Augsburg [KORA]) to quantify myosteatosis on whole-body T1-weighted Dixon magnetic resonance imaging as (1) intramuscular adipose tissue (IMAT; the current standard) and (2) quantitative skeletal muscle (SM) fat fraction (SMFF). Subsequently, we investigated the two measures for their discrimination of and association with impaired glucose metabolism beyond baseline demographics (age, sex and body mass index [BMI]) and cardiometabolic risk factors (lipid panel, systolic blood pressure, smoking status and alcohol consumption) in asymptomatic individuals from the KORA study. Impaired glucose metabolism was defined as impaired fasting glucose or impaired glucose tolerance (140-200 mg/dL) or prevalent diabetes mellitus.

RESULTS: Model performance was high, with Dice coefficients of ≥0.81 for IMAT and ≥0.91 for SM in the internal (NAKO) and external (KORA) testing sets. In the target population (380 KORA participants: mean age of 53.6 ± 9.2 years, BMI of 28.2 ± 4.9 kg/m2, 57.4% male), individuals with impaired glucose metabolism (n = 146; 38.4%) were older and more likely men and showed a higher cardiometabolic risk profile, higher IMAT (4.5 ± 2.2% vs. 3.9 ± 1.7%) and higher SMFF (22.0 ± 4.7% vs. 18.9 ± 3.9%) compared to normoglycaemic controls (all P ≤ 0.005). SMFF showed better discrimination for impaired glucose metabolism than IMAT (area under the receiver operating characteristic curve [AUC] 0.693 vs. 0.582, 95% confidence interval [CI] [0.06-0.16]; P < 0.001) but was not significantly different from BMI (AUC 0.733 vs. 0.693, 95% CI [-0.09 to 0.01]; P = 0.15). In univariable logistic regression, IMAT (odds ratio [OR] = 1.18, 95% CI [1.06-1.32]; P = 0.004) and SMFF (OR = 1.19, 95% CI [1.13-1.26]; P < 0.001) were associated with a higher risk of impaired glucose metabolism. This signal remained robust after multivariable adjustment for baseline demographics and cardiometabolic risk factors for SMFF (OR = 1.10, 95% CI [1.01-1.19]; P = 0.028) but not for IMAT (OR = 1.14, 95% CI [0.97-1.33]; P = 0.11).

CONCLUSIONS: Quantitative SMFF, but not IMAT, is an independent predictor of impaired glucose metabolism, and discrimination is not significantly different from BMI, making it a promising alternative for the currently established approach. Automated methods such as the proposed model may provide a feasible option for opportunistic screening of myosteatosis and, thus, a low-cost personalized risk assessment solution.

PMID:39009381 | DOI:10.1002/jcsm.13527

Magn Reson Med Sci. 2024 Jul 12. doi: 10.2463/mrms.mp.2023-0125. Online ahead of print.

ABSTRACT

PURPOSE: To develop a new method to generate synthetic MR spectroscopic imaging (MRSI) data for training machine learning models.

METHODS: This study targeted routine MRI examination protocols with single voxel spectroscopy (SVS). A novel model derived from pix2pix generative adversarial networks was proposed to generate synthetic MRSI data using MRI and SVS data as inputs. T1- and T2-weighted, SVS, and reference MRSI data were acquired from healthy brains with clinically available sequences. The proposed model was trained to generate synthetic MRSI data. Quantitative evaluation involved the calculation of the mean squared error (MSE) against the reference and metabolite ratio value. The effect of the location of and the number of the SVS data on the quality of the synthetic MRSI data was investigated using the MSE.

RESULTS: The synthetic MRSI data generated from the proposed model were visually closer to the reference. The 95% confidence interval (CI) of the metabolite ratio value of synthetic MRSI data overlapped with the reference for seven of eight metabolite ratios. The MSEs tended to be lower in the same location than in different locations. The MSEs among groups of numbers of SVS data were not significantly different.

CONCLUSION: A new method was developed to generate MRSI data by integrating MRI and SVS data. Our method can potentially increase the volume of MRSI data training for other machine learning models by adding SVS acquisition to routine MRI examinations.

PMID:39010240 | DOI:10.2463/mrms.mp.2023-0125

J Oral Sci. 2024;66(3):163-168. doi: 10.2334/josnusd.24-0031.

ABSTRACT

PURPOSE: Using X-ray micro-computed tomography (micro-CT), the aim of this study was to measure the porosity of two tricalcium silicate sealers (EndoSequence BC and NeoSealer Flo) applied using three obturation techniques (single-cone, warm-vertical, and cold-lateral) to six single-rooted human teeth.

METHODS: Six extracted, single-rooted human teeth were shaped with ProTaper Next rotary files and obturated with EndoSequence BC or NeoSealer Flo sealers and gutta-percha (GP) using one of the three techniques above. Micro-CT was used to map the full length of the canals. Deep learning cross-sectional segmentation was used to analyze image slices of the apical (0-2 mm) and coronal (14-16 mm from the apex) regions (n = 230-261 per tooth) for the areas of GP and sealer, as well as porosity. Median (%) with interquartile range of porosity were calculated , and the results were statistically analyzed with the Kruskal-Wallis test.

RESULTS: In the apical region, EndoSequence BC had significantly fewer pores than NeoSealer Flo with the single-cone obturation (% median-interquartile range, IQR: 0.00-1.62) and warm-vertical condensation (5.57-10.32) techniques, whereas in the coronal region, NeoSealer Flo had significantly fewer pores than EndoSequence BC with these two techniques (0.39-5.02) and (0.10-0.19), respectively. There was no significant difference in porosity between the two sealers for the cold-lateral condensation technique in both the apical and coronal regions.

CONCLUSION: For optimal obturation, the choice of technique and sealer is critical.

PMID:39010164 | DOI:10.2334/josnusd.24-0031

J Cheminform. 2024 Jul 15;16(1):80. doi: 10.1186/s13321-024-00877-2.

ABSTRACT

MOTIVATION: Retrosynthesis planning poses a formidable challenge in the organic chemical industry, particularly in pharmaceuticals. Single-step retrosynthesis prediction, a crucial step in the planning process, has witnessed a surge in interest in recent years due to advancements in AI for science. Various deep learning-based methods have been proposed for this task in recent years, incorporating diverse levels of additional chemical knowledge dependency.

RESULTS: This paper introduces UAlign, a template-free graph-to-sequence pipeline for retrosynthesis prediction. By combining graph neural networks and Transformers, our method can more effectively leverage the inherent graph structure of molecules. Based on the fact that the majority of molecule structures remain unchanged during a chemical reaction, we propose a simple yet effective SMILES alignment technique to facilitate the reuse of unchanged structures for reactant generation. Extensive experiments show that our method substantially outperforms state-of-the-art template-free and semi-template-based approaches. Importantly, our template-free method achieves effectiveness comparable to, or even surpasses, established powerful template-based methods.

SCIENTIFIC CONTRIBUTION: We present a novel graph-to-sequence template-free retrosynthesis prediction pipeline that overcomes the limitations of Transformer-based methods in molecular representation learning and insufficient utilization of chemical information. We propose an unsupervised learning mechanism for establishing product-atom correspondence with reactant SMILES tokens, achieving even better results than supervised SMILES alignment methods. Extensive experiments demonstrate that UAlign significantly outperforms state-of-the-art template-free methods and rivals or surpasses template-based approaches, with up to 5% (top-5) and 5.4% (top-10) increased accuracy over the strongest baseline.

PMID:39010144 | DOI:10.1186/s13321-024-00877-2

BMC Musculoskelet Disord. 2024 Jul 16;25(1):547. doi: 10.1186/s12891-024-07669-7.

ABSTRACT

OBJECTIVE: This study aimed to evaluate a new deep-learning model for diagnosing avascular necrosis of the femoral head (AVNFH) by analyzing pelvic anteroposterior digital radiography.

METHODS: The study sample included 1167 hips. The radiographs were independently classified into 6 stages by a radiologist using their simultaneous MRIs. After that, the radiographs were given to train and test the deep learning models of the project including SVM and ANFIS layer using the Python programming language and TensorFlow library. In the last step, the test set of hip radiographs was provided to two independent radiologists with different work experiences to compare their diagnosis performance to the deep learning models' performance using the F1 score and Mcnemar test analysis.

RESULTS: The performance of SVM for AVNFH detection (AUC = 82.88%) was slightly higher than less experienced radiologists (79.68%) and slightly lower than experienced radiologists (88.4%) without reaching significance (p-value > 0.05). Evaluation of the performance of SVM for pre-collapse AVNFH detection with an AUC of 73.58% showed significantly higher performance than less experienced radiologists (AUC = 60.70%, p-value < 0.001). On the other hand, no significant difference is noted between experienced radiologists and SVM for pre-collapse detection. ANFIS algorithm for AVNFH detection with an AUC of 86.60% showed significantly higher performance than less experienced radiologists (AUC = 79.68%, p-value = 0.04). Although reaching less performance compared to experienced radiologists statistically not significant (AUC = 88.40%, p-value = 0.20).

CONCLUSIONS: Our study has shed light on the remarkable capabilities of SVM and ANFIS as diagnostic tools for AVNFH detection in radiography. Their ability to achieve high accuracy with remarkable efficiency makes them promising candidates for early detection and intervention, ultimately contributing to improved patient outcomes.

PMID:39010001 | DOI:10.1186/s12891-024-07669-7

Curr Oncol Rep. 2024 Jul 16. doi: 10.1007/s11912-024-01580-z. Online ahead of print.

ABSTRACT

PURPOSE OF REVIEW: Isocitrate dehydrogenase wild-type glioblastoma is the most aggressive primary brain tumour in adults. Its infiltrative nature and heterogeneity confer a dismal prognosis, despite multimodal treatment. Precision medicine is increasingly advocated to improve survival rates in glioblastoma management; however, conventional neuroimaging techniques are insufficient in providing the detail required for accurate diagnosis of this complex condition.

RECENT FINDINGS: Advanced magnetic resonance imaging allows more comprehensive understanding of the tumour microenvironment. Combining diffusion and perfusion magnetic resonance imaging to create a multiparametric scan enhances diagnostic power and can overcome the unreliability of tumour characterisation by standard imaging. Recent progress in deep learning algorithms establishes their remarkable ability in image-recognition tasks. Integrating these with multiparametric scans could transform the diagnosis and monitoring of patients by ensuring that the entire tumour is captured. As a corollary, radiomics has emerged as a powerful approach to offer insights into diagnosis, prognosis, treatment, and tumour response through extraction of information from radiological scans, and transformation of these tumour characteristics into quantitative data. Radiogenomics, which links imaging features with genomic profiles, has exhibited its ability in characterising glioblastoma, and determining therapeutic response, with the potential to revolutionise management of glioblastoma. The integration of deep learning algorithms into radiogenomic models has established an automated, highly reproducible means to predict glioblastoma molecular signatures, further aiding prognosis and targeted therapy. However, challenges including lack of large cohorts, absence of standardised guidelines and the 'black-box' nature of deep learning algorithms, must first be overcome before this workflow can be applied in clinical practice.

PMID:39009914 | DOI:10.1007/s11912-024-01580-z

Sci Rep. 2024 Jul 15;14(1):16254. doi: 10.1038/s41598-024-63242-1.

ABSTRACT

With technological innovations, enterprises in the real world are managing every iota of data as it can be mined to derive business intelligence (BI). However, when data comes from multiple sources, it may result in duplicate records. As data is given paramount importance, it is also significant to eliminate duplicate entities towards data integration, performance and resource optimization. To realize reliable systems for record deduplication, late, deep learning could offer exciting provisions with a learning-based approach. Deep ER is one of the deep learning-based methods used recently for dealing with the elimination of duplicates in structured data. Using it as a reference model, in this paper, we propose a framework known as Enhanced Deep Learning-based Record Deduplication (EDL-RD) for improving performance further. Towards this end, we exploited a variant of Long Short Term Memory (LSTM) along with various attribute compositions, similarity metrics, and numerical and null value resolution. We proposed an algorithm known as Efficient Learning based Record Deduplication (ELbRD). The algorithm extends the reference model with the aforementioned enhancements. An empirical study has revealed that the proposed framework with extensions outperforms existing methods.

PMID:39009682 | DOI:10.1038/s41598-024-63242-1

Sci Rep. 2024 Jul 15;14(1):16308. doi: 10.1038/s41598-024-67017-6.

ABSTRACT

Vertebral compression fractures (VCFs) of the thoracolumbar spine are commonly caused by osteoporosis or result from traumatic events. Early diagnosis of vertebral compression fractures can prevent further damage to patients. When assessing these fractures, plain radiographs are used as the primary diagnostic modality. In this study, we developed a deep learning based fracture detection model that could be used as a tool for primary care in the orthopedic department. We constructed a VCF dataset using 487 lateral radiographs, which included 598 fractures in the L1-T11 vertebra. For detecting VCFs, Mask R-CNN model was trained and optimized, and was compared to three other popular models on instance segmentation, Cascade Mask R-CNN, YOLOACT, and YOLOv5. With Mask R-CNN we achieved highest mean average precision score of 0.58, and were able to locate each fracture pixel-wise. In addition, the model showed high overall sensitivity, specificity, and accuracy, indicating that it detected fractures accurately and without misdiagnosis. Our model can be a potential tool for detecting VCFs from a simple radiograph and assisting doctors in making appropriate decisions in initial diagnosis.

PMID:39009647 | DOI:10.1038/s41598-024-67017-6

Sci Rep. 2024 Jul 15;14(1):16304. doi: 10.1038/s41598-024-67130-6.

ABSTRACT

This research paper introduces an efficient approach for the segmentation of active and inactive plaques within Fluid-attenuated inversion recovery (FLAIR) images, employing a convolutional neural network (CNN) model known as DeepLabV3Plus SE with the EfficientNetB0 backbone in Multiple sclerosis (MS), and demonstrates its superior performance compared to other CNN architectures. The study encompasses various critical components, including dataset pre-processing techniques, the utilization of the Squeeze and Excitation Network (SE-Block), and the atrous spatial separable pyramid Block to enhance segmentation capabilities. Detailed descriptions of pre-processing procedures, such as removing the cranial bone segment, image resizing, and normalization, are provided. This study analyzed a cross-sectional cohort of 100 MS patients with active brain plaques, examining 5000 MRI slices. After filtering, 1500 slices were utilized for labeling and deep learning. The training process adopts the dice coefficient as the loss function and utilizes Adam optimization. The study evaluated the model's performance using multiple metrics, including intersection over union (IOU), Dice Score, Precision, Recall, and F1-Score, and offers a comparative analysis with other CNN architectures. Results demonstrate the superior segmentation ability of the proposed model, as evidenced by an IOU of 69.87, Dice Score of 76.24, Precision of 88.89, Recall of 73.52, and F1-Score of 80.47 for the DeepLabV3+SE_EfficientNetB0 model. This research contributes to the advancement of plaque segmentation in FLAIR images and offers a compelling approach with substantial potential for medical image analysis and diagnosis.

PMID:39009636 | DOI:10.1038/s41598-024-67130-6

Transl Psychiatry. 2024 Jul 15;14(1):287. doi: 10.1038/s41398-024-02992-y.

ABSTRACT

The causes of depression are complex, and the current diagnosis methods rely solely on psychiatric evaluations with no incorporation of laboratory biomarkers in clinical practices. We investigated the stability of blood DNA methylation depression signatures in six different populations using six public and two domestic cohorts (n = 1942) conducting mega-analysis and meta-analysis of the individual studies. We evaluated 12 machine learning and deep learning strategies for depression classification both in cross-validation (CV) and in hold-out tests using merged data from 8 separate batches, constructing models with both biased and unbiased feature selection. We found 1987 CpG sites related to depression in both mega- and meta-analysis at the nominal level, and the associated genes were nominally related to axon guidance and immune pathways based on enrichment analysis and eQTM data. Random forest classifiers achieved the highest performance (AUC 0.73 and 0.76) in CV and hold-out tests respectively on the batch-level processed data. In contrast, the methylation showed low predictive power (all AUCs < 0.57) for all classifiers in CV and no predictive power in hold-out tests when used with harmonized data. All models achieved significantly better performance (>14% gain in AUCs) with pre-selected features (selection bias), with some of the models (joint autoencoder-classifier) reaching AUCs of up to 0.91 in the final testing regardless of data preparation. Different algorithmic feature selection approaches may outperform limma, however, random forest models perform well regardless of the strategy. The results provide an overview over potential future biomarkers for depression and highlight many important methodological aspects for DNA methylation-based depression profiling including the use of machine learning strategies.

PMID:39009577 | DOI:10.1038/s41398-024-02992-y

Nat Commun. 2024 Jul 15;15(1):5945. doi: 10.1038/s41467-024-49381-z.

ABSTRACT

Understanding and interpreting dynamics of functional materials in situ is a grand challenge in physics and materials science due to the difficulty of experimentally probing materials at varied length and time scales. X-ray photon correlation spectroscopy (XPCS) is uniquely well-suited for characterizing materials dynamics over wide-ranging time scales. However, spatial and temporal heterogeneity in material behavior can make interpretation of experimental XPCS data difficult. In this work, we have developed an unsupervised deep learning (DL) framework for automated classification of relaxation dynamics from experimental data without requiring any prior physical knowledge of the system. We demonstrate how this method can be used to accelerate exploration of large datasets to identify samples of interest, and we apply this approach to directly correlate microscopic dynamics with macroscopic properties of a model system. Importantly, this DL framework is material and process agnostic, marking a concrete step towards autonomous materials discovery.

PMID:39009571 | DOI:10.1038/s41467-024-49381-z

Invest Radiol. 2024 Jul 16. doi: 10.1097/RLI.0000000000001103. Online ahead of print.

ABSTRACT

Interstitial lung disease (ILD) encompasses a variety of lung disorders with varying degrees of inflammation or fibrosis, requiring a combination of clinical, imaging, and pathologic data for evaluation. Imaging is essential for the noninvasive diagnosis of the disease, as well as for assessing disease severity, monitoring its progression, and evaluating treatment response. However, traditional visual assessments of ILD with computed tomography (CT) suffer from reader variability. Automated quantitative CT offers a more objective approach by using computer-based analysis to consistently evaluate and measure ILD. Advancements in technology have significantly improved the accuracy and reliability of these measurements. Recently, interstitial lung abnormalities (ILAs), which represent potential preclinical ILD incidentally found on CT scans and are characterized by abnormalities in over 5% of any lung zone, have gained attention and clinical importance. The challenge lies in the accurate and consistent identification of ILA, given that its definition relies on a subjective threshold, making quantitative tools crucial for precise ILA evaluation. This review highlights the state of CT quantification of ILD and ILA, addressing clinical and research disparities while emphasizing how machine learning or deep learning in quantitative imaging can improve diagnosis and management by providing more accurate assessments, and finally, suggests the future directions of quantitative CT in this area.

PMID:39008898 | DOI:10.1097/RLI.0000000000001103

J Phys Chem A. 2024 Jul 15. doi: 10.1021/acs.jpca.4c00077. Online ahead of print.

ABSTRACT

We demonstrate the use of gradient-boosted ensemble models that accurately predict emission wavelengths in benzobis[1,2-d:4,5-d']oxazole (BBO) based fluorescent emitters. We have curated a database of 50 molecules from previously published data by the Jeffries-EL group using density functional theory (DFT) computed ground and excited state features. We consider two machine learning (ML) models based on (i) whole cruciform molecules and (ii) their constituent fragment molecules. Both ML models provide accurate predictions with root-mean-square errors between 30 and 36 nm, competitive with state-of-the-art deep learning models trained on orders of magnitude more molecules, and this accuracy holds even when tested on four new BBO emitters unseen by the models. We also provide an interpretable feature importance analysis and discuss the relevant relationships between DFT and changes in predicted emission wavelength.

PMID:39008894 | DOI:10.1021/acs.jpca.4c00077

Med Phys. 2024 Jul 15. doi: 10.1002/mp.17303. Online ahead of print.

ABSTRACT

BACKGROUND: Lesion detection is one of the most important clinical tasks in positron emission tomography (PET) for oncology. An anthropomorphic model observer (MO) designed to replicate human observers (HOs) in a detection task is an important tool for assessing task-based image quality. The channelized Hotelling observer (CHO) has been the most popular anthropomorphic MO. Recently, deep learning MOs (DLMOs), mostly based on convolutional neural networks (CNNs), have been investigated for various imaging modalities. However, there have been few studies on DLMOs for PET.

PURPOSE: The goal of the study is to investigate whether DLMOs can predict HOs better than conventional MOs such as CHO in a two-alternative forced-choice (2AFC) detection task using PET images with real anatomical variability.

METHODS: Two types of DLMOs were implemented: (1) CNN DLMO, and (2) CNN-SwinT DLMO that combines CNN and Swin Transformer (SwinT) encoders. Lesion-absent PET images were reconstructed from clinical data, and lesion-present images were reconstructed with adding simulated lesion sinogram data. Lesion-present and lesion-absent PET image pairs were labeled by eight HOs consisting of four radiologists and four image scientists in a 2AFC detection task. In total, 2268 pairs of lesion-present and lesion-absent images were used for training, 324 pairs for validation, and 324 pairs for test. CNN DLMO, CNN-SwinT DLMO, CHO with internal noise, and non-prewhitening matched filter (NPWMF) were compared in the same train-test paradigm. For comparison, six quantitative metrics including prediction accuracy, mean squared errors (MSEs) and correlation coefficients, which measure how well a MO predicts HOs, were calculated in a 9-fold cross-validation experiment.

RESULTS: In terms of the accuracy and MSE metrics, CNN DLMO and CNN-SwinT DLMO showed better performance than CHO and NPWMF, and CNN-SwinT DLMO showed the best performance among the MOs evaluated.

CONCLUSIONS: DLMO can predict HOs more accurately than conventional MOs such as CHO in PET lesion detection. Combining SwinT and CNN encoders can improve the DLMO prediction performance compared to using CNN only.

PMID:39008812 | DOI:10.1002/mp.17303

Med Phys. 2024 Jul 15. doi: 10.1002/mp.17291. Online ahead of print.

ABSTRACT

BACKGROUND: Vessel-wall volume and localized three-dimensional ultrasound (3DUS) metrics are sensitive to the change of carotid atherosclerosis in response to medical/dietary interventions. Manual segmentation of the media-adventitia boundary (MAB) and lumen-intima boundary (LIB) required to obtain these metrics is time-consuming and prone to observer variability. Although supervised deep-learning segmentation models have been proposed, training of these models requires a sizeable manually segmented training set, making larger clinical studies prohibitive.

PURPOSE: We aim to develop a method to optimize pre-trained segmentation models without requiring manual segmentation to supervise the fine-tuning process.

METHODS: We developed an adversarial framework called the unsupervised shape-and-texture generative adversarial network (USTGAN) to fine-tune a convolutional neural network (CNN) pre-trained on a source dataset for accurate segmentation of a target dataset. The network integrates a novel texture-based discriminator with a shape-based discriminator, which together provide feedback for the CNN to segment the target images in a similar way as the source images. The texture-based discriminator increases the accuracy of the CNN in locating the artery, thereby lowering the number of failed segmentations. Failed segmentation was further reduced by a self-checking mechanism to flag longitudinal discontinuity of the artery and by self-correction strategies involving surface interpolation followed by a case-specific tuning of the CNN. The U-Net was pre-trained by the source dataset involving 224 3DUS volumes with 136, 44, and 44 volumes in the training, validation and testing sets. The training of USTGAN involved the same training group of 136 volumes in the source dataset and 533 volumes in the target dataset. No segmented boundaries for the target cohort were available for training USTGAN. The validation and testing of USTGAN involved 118 and 104 volumes from the target cohort, respectively. The segmentation accuracy was quantified by Dice Similarity Coefficient (DSC), and incorrect localization rate (ILR). Tukey's Honestly Significant Difference multiple comparison test was employed to quantify the difference of DSCs between models and settings, where p ≤ 0.05 $p\,\le \,0.05$ was considered statistically significant.

RESULTS: USTGAN attained a DSC of 85.7 ± 13.0 $85.7\,\pm \,13.0$ % in LIB and 86.2 ± 10.6 ${86.2}\,\pm \,{10.6}$ % in MAB, improving from the baseline performance of 74.6 ± 30.7 ${74.6}\,\pm \,{30.7}$ % in LIB (p < 10 - 12 $&lt;10^{-12}$ ) and 75.7 ± 28.9 ${75.7}\,\pm \,{28.9}$ % in MAB (p < 10 - 12 $&lt;10^{-12}$ ). Our approach outperformed six state-of-the-art domain-adaptation models (MAB: p ≤ 3.63 × 10 - 7 $p \le 3.63\,\times \,10^{-7}$ , LIB: p ≤ 9.34 × 10 - 8 $p\,\le \,9.34\,\times \,10^{-8}$ ). The proposed USTGAN also had the lowest ILR among the methods compared (LIB: 2.5%, MAB: 1.7%).

CONCLUSION: Our framework improves segmentation generalizability, thereby facilitating efficient carotid disease monitoring in multicenter trials and in clinics with less expertise in 3DUS imaging.

PMID:39008794 | DOI:10.1002/mp.17291

Med Phys. 2024 Jul 15. doi: 10.1002/mp.17306. Online ahead of print.

ABSTRACT

BACKGROUND: The image resolution of fetal brain magnetic resonance imaging (MRI) is a critical factor in brain development measures, which is mainly determined by the physical resolution configured in the MRI sequence. However, fetal brain MRI are commonly reconstructed to 3D images with a higher apparent resolution, compared to the original physical resolution.

PURPOSE: This work is to demonstrate that accurate segmentation can be achieved based on the MRI physical resolution, and the high apparent resolution segmentation can be achieved by a simple deep learning module.

METHODS: This retrospective study included 150 adult and 80 fetal brain MRIs. The adult brain MRIs were acquired at a high physical resolution, which were downsampled to visualize and quantify its impacts on the segmentation accuracy. The physical resolution of fetal images was estimated based on MRI acquisition settings and the images were downsampled accordingly before segmentation and restored using multiple upsampling strategies. Segmentation accuracy of ConvNet models were evaluated on the original and downsampled images. Dice coefficients were calculated, and compared to the original data.

RESULTS: When the apparent resolution was higher than the physical resolution, the accuracy of fetal brain segmentation had negligible degradation (accuracy reduced by 0.26%, 1.1%, and 1.8% with downsampling factors of 4/3, 2, and 4 in each dimension, without significant differences from the original data). Using a downsampling factor of 4 in each dimension, the proposed method provided 7× smaller and 10× faster models.

CONCLUSION: Efficient and accurate fetal brain segmentation models can be developed based on the physical resolution of MRI acquisitions.

PMID:39008780 | DOI:10.1002/mp.17306

Anal Chem. 2024 Jul 15. doi: 10.1021/acs.analchem.4c01622. Online ahead of print.

ABSTRACT

Deep learning (DL) is becoming more popular as a useful tool in various scientific domains, especially in chemistry applications. In the infrared spectroscopy field, where identifying functional groups in unknown compounds poses a significant challenge, there is a growing need for innovative approaches to streamline and enhance analysis processes. This study introduces a transformative approach leveraging a DL methodology based on transformer attention models. With a data set containing approximately 8677 spectra, our model utilizes self-attention mechanisms to capture complex spectral features and precisely predict 17 functional groups, outperforming conventional architectures in both functional group prediction accuracy and compound-level precision. The success of our approach underscores the potential of transformer-based methodologies in enhancing spectral analysis techniques.

PMID:39008658 | DOI:10.1021/acs.analchem.4c01622

J Phys Chem A. 2024 Jul 15. doi: 10.1021/acs.jpca.4c00083. Online ahead of print.

ABSTRACT

Due to the vast chemical space, discovering materials with a specific function is challenging. Chemical formulas are obligated to conform to a set of exacting criteria, such as charge neutrality, balanced electronegativity, synthesizability, and mechanical stability. In response to this formidable task, we introduce a deep-learning-based generative model for material composition and structure design by learning and exploiting explicit and implicit chemical knowledge. Our pipeline first uses deep diffusion language models as the generator of compositions and then applies a template-based crystal structure prediction algorithm to predict their corresponding structures, which is then followed by structure relaxation using a universal graph neural network-based potential. Density functional theory (DFT) calculations of the formation energies and energy-above-the-hull analysis are used to validate new structures generated through our pipeline. Based on the DFT calculation results, six new materials, including Ti2HfO5, TaNbP, YMoN2, TaReO4, HfTiO2, and HfMnO2, with formation energy less than zero have been found. Remarkably, among these, four materials, namely, Ti2HfO5, TaNbP, YMoN2, and TaReO4, exhibit an e-above-hull energy of less than 0.3 eV. These findings have proved the effectiveness of our approach.

PMID:39008628 | DOI:10.1021/acs.jpca.4c00083

PLoS One. 2024 Jul 15;19(7):e0305404. doi: 10.1371/journal.pone.0305404. eCollection 2024.

ABSTRACT

This work investigates whether inclusion of the low-frequency components of heart sounds can increase the accuracy, sensitivity and specificity of diagnosis of cardiovascular disorders. We standardized the measurement method to minimize changes in signal characteristics. We used the Continuous Wavelet Transform to analyze changing frequency characteristics over time and to allocate frequencies appropriately between the low-frequency and audible frequency bands. We used a Convolutional Neural Network (CNN) and deep-learning (DL) for image classification, and a CNN equipped with long short-term memory to enable sequential feature extraction. The accuracy of the learning model was validated using the PhysioNet 2016 CinC dataset, then we used our collected dataset to show that incorporating low-frequency components in the dataset increased the DL model's accuracy by 2% and sensitivity by 4%. Furthermore, the LSTM layer was 0.8% more accurate than the dense layer.

PMID:39008512 | DOI:10.1371/journal.pone.0305404