Epilepsy Behav. 2024 Mar 22;154:109735. doi: 10.1016/j.yebeh.2024.109735. Online ahead of print.

ABSTRACT

Seizure events can manifest as transient disruptions in the control of movements which may be organized in distinct behavioral sequences, accompanied or not by other observable features such as altered facial expressions. The analysis of these clinical signs, referred to as semiology, is subject to observer variations when specialists evaluate video-recorded events in the clinical setting. To enhance the accuracy and consistency of evaluations, computer-aided video analysis of seizures has emerged as a natural avenue. In the field of medical applications, deep learning and computer vision approaches have driven substantial advancements. Historically, these approaches have been used for disease detection, classification, and prediction using diagnostic data; however, there has been limited exploration of their application in evaluating video-based motion detection in the clinical epileptology setting. While vision-based technologies do not aim to replace clinical expertise, they can significantly contribute to medical decision-making and patient care by providing quantitative evidence and decision support. Behavior monitoring tools offer several advantages such as providing objective information, detecting challenging-to-observe events, reducing documentation efforts, and extending assessment capabilities to areas with limited expertise. The main applications of these could be (1) improved seizure detection methods; (2) refined semiology analysis for predicting seizure type and cerebral localization. In this paper, we detail the foundation technologies used in vision-based systems in the analysis of seizure videos, highlighting their success in semiology detection and analysis, focusing on work published in the last 7 years. We systematically present these methods and indicate how the adoption of deep learning for the analysis of video recordings of seizures could be approached. Additionally, we illustrate how existing technologies can be interconnected through an integrated system for video-based semiology analysis. Each module can be customized and improved by adapting more accurate and robust deep learning approaches as these evolve. Finally, we discuss challenges and research directions for future studies.

PMID:38522192 | DOI:10.1016/j.yebeh.2024.109735

Blood Adv. 2024 Mar 24:bloodadvances.2023012200. doi: 10.1182/bloodadvances.2023012200. Online ahead of print.

ABSTRACT

Venous thromboembolism (VTE) is a leading cause of preventable in-hospital mortality. Monitoring VTE cases is limited by the challenges of manual chart review and diagnosis code interpretation. Natural language processing (NLP) can automate the process. Rule-based NLP methods are effective but time consuming. Machine learning (ML)-NLP methods present a promising solution. We conducted a systematic review and meta-analysis of studies published before May 2023 that use ML-NLP to identify VTE diagnoses in the electronic health records. Four reviewers screened all manuscripts, excluding studies that only used a rule-based method. A meta-analysis evaluated the pooled performance of each study's best performing model that evaluated for pulmonary embolism (PE) and/or deep vein thrombosis (DVT). Pooled sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) with confidence interval (CI) were calculated by DerSimonian and Laird method using a random-effects model. Study quality was assessed using an adapted TRIPOD tool. Thirteen studies were included in the systematic review and 8 had data available for meta-analysis. Pooled sensitivity was 0.931 (95% CI 0.881-0.962), specificity 0.984 (95% CI 0.967-0.992), PPV 0.910 (95% CI 0.865-0.941) and NPV 0.985 (95% CI 0.977-0.990). All studies met at least 13 of the 21 NLP-modified TRIPOD items, demonstrating fair quality. The highest performing models used vectorization rather than bag-of-words, and deep learning techniques such as convolutional neural networks. There was significant heterogeneity in the studies and only four validated their model on an external dataset. Further standardization of ML studies can help progress this novel technology towards real-world implementation.

PMID:38522096 | DOI:10.1182/bloodadvances.2023012200

Acta Diabetol. 2024 Mar 23. doi: 10.1007/s00592-024-02257-6. Online ahead of print.

ABSTRACT

AIMS: This study aims to develop an advanced model for the classification of Diabetic Macular Edema (DME) using deep learning techniques. Specifically, the objective is to introduce a novel architecture, SSCSAC-Net, that leverages self-supervised learning and category-selective attention mechanisms to improve the precision of DME classification.

METHODS: The proposed SSCSAC-Net integrates self-supervised learning to effectively utilize unlabeled data for learning robust features related to DME. Additionally, it incorporates a category-specific attention mechanism and a domain-specific layer into the ResNet-152 base architecture. The model is trained using an ensemble of unsupervised and supervised learning techniques. Benchmark datasets are utilized for testing the model's performance, ensuring its robustness and generalizability across different data distributions.

RESULTS: Evaluation of the SSCSAC-Net on multiple datasets demonstrates its superior performance compared to existing techniques. The model achieves high accuracy, precision, and recall rates, with an accuracy of 98.7%, precision of 98.6%, and recall of 98.8%. Furthermore, the incorporation of self-supervised learning reduces the dependency on extensive labeled data, making the solution more scalable and cost-effective.

CONCLUSIONS: The proposed SSCSAC-Net represents a significant advancement in automated DME classification. By effectively using self-supervised learning and attention mechanisms, the model offers improved accuracy in identifying DME-related features within retinal images. Its robustness and generalizability across different datasets highlight its potential for clinical applications, providing a valuable tool for clinicians in diagnosing DME effectively.

PMID:38521818 | DOI:10.1007/s00592-024-02257-6

Acad Radiol. 2024 Mar 22:S1076-6332(24)00139-9. doi: 10.1016/j.acra.2024.02.049. Online ahead of print.

ABSTRACT

OBJECTIVES: To investigate the clinical feasibility and image quality of accelerated brain diffusion-weighted imaging (DWI) with deep learning image reconstruction and super resolution.

METHODS: 85 consecutive patients with clinically indicated MRI at a 3 T scanner were prospectively included. Conventional diffusion-weighted data (c-DWI) with four averages were obtained. Reconstructions of one and two averages, as well as deep learning diffusion-weighted imaging (DL-DWI), were accomplished. Three experienced readers evaluated the acquired data using a 5-point Likert scale regarding overall image quality, overall contrast, diagnostic confidence, occurrence of artefacts and evaluation of the central region, basal ganglia, brainstem, and cerebellum. To assess interrater agreement, Fleiss' kappa (ϰ) was determined. Signal intensity (SI) levels for basal ganglia and the central region were estimated via automated segmentation, and SI values of detected pathologies were measured.

RESULTS: Intracranial pathologies were identified in 35 patients. DL-DWI was significantly superior for all defined parameters, independently from applied averages (p-value <0.001). Optimum image quality was achieved with DL-DWI by utilizing a single average (p-value <0.001), demonstrating very good (80.9%) to excellent image quality (14.5%) in nearly all cases, compared to 12.5% with very good and 0% with excellent image quality for c-MRI (p-value <0.001). Comparable results could be shown for diagnostic confidence. Inter-rater Fleiss' Kappa demonstrated moderate to substantial agreement for virtually all defined parameters, with good accordance, particularly for the assessment of pathologies (p = 0.74). Regarding SI values, no significant difference was found.

CONCLUSION: Ultra-fast diffusion-weighted imaging with super resolution is feasible, resulting in highly accelerated brain imaging while increasing diagnostic image quality.

PMID:38521612 | DOI:10.1016/j.acra.2024.02.049

Environ Pollut. 2024 Mar 21:123838. doi: 10.1016/j.envpol.2024.123838. Online ahead of print.

ABSTRACT

Accurate fine-mode and coarse-mode aerosol knowledge is crucial for understanding their impacts on the climate and Earth's ecosystems. However, current satellite-based Fine-Mode Aerosol Optical Depth (FAOD) and Coarse-Mode Aerosol Optical Depth (CAOD) methods have drawbacks including inaccuracies, low spatial coverage, and limited temporal duration. To overcome these issues, we developed new global-scale FAOD and CAOD from 2005 to 2020 using a novel deep learning model capable of the synergistic retrieval of two aerosol sizes. After validation with the aerosol robotic network (AERONET) and sky radiometer network (SKYNET), the new FAOD and CAOD showed significant improvements in accuracy and spatial coverage. From 2005 to 2020, the new data showed that China had the greatest decrease in FAOD and CAOD. In contrast, India and South Latin America had a significant increase in FAOD versus North Africa in CAOD. FAOD in the regions of Argentina, Paraguay, and Uruguay in South America have shown an upward trend. The results reveal that FAOD and CAOD display distinct patterns of change in the same regions, particularly on the west coast of the United States where FAOD is increasing, while CAOD is decreasing. Aside from the year 2020 due to the global COVID-19 pandemic, the analysis showed that although China has seen at least an +85% increase in energy consumption and urban expansion in 2019 compared to 2005 due to the needs of development and construction, the implementation of China's air pollution control policies has led to a significant decrease in FAOD (-46%) and CAOD (-65%) after 2013. This research enriches our comprehension of global fine and coarse aerosol patterns, additional investigations are needed to determine the potential global implications of these changes.

PMID:38521397 | DOI:10.1016/j.envpol.2024.123838

J Cardiovasc Magn Reson. 2024 Mar 21:101039. doi: 10.1016/j.jocmr.2024.101039. Online ahead of print.

ABSTRACT

BACKGROUND: MRI is an important imaging modality for the assessment and management of adult patients with congenital heart disease (CHD). However, conventional techniques for 3D whole-heart acquisition involve long and unpredictable scan times and methods that accelerate scans via k-space undersampling often rely on long iterative reconstructions. Deep-learning-based reconstruction methods have recently attracted much interest due to their capacity to provide fast reconstructions whilst often outperforming existing state-of-the-art methods. In this study, we sought to adapt and validate a non-rigid motion-corrected model-based deep-learning (MoCo-MoDL) reconstruction framework for 3D whole-heart MRI in a CHD patient cohort.

METHODS: The previously proposed deep-learning reconstruction framework MoCo-MoDL, which incorporates a non-rigid motion-estimation network and a denoising regularisation network within an unrolled iterative reconstruction, was trained in an end-to-end manner using 39 CHD patient datasets. Once trained, the framework was evaluated in 8 CHD patient datasets acquired with seven-fold prospective undersampling. Reconstruction quality was compared with the state-of-the-art non-rigid motion-corrected patch-based low-rank reconstruction method (NR-PROST) and against reference images (acquired with three-or-four-fold undersampling and reconstructed with NR-PROST).

RESULTS: Seven-fold undersampled scan times were 2.1 ± 0.3 minutes and reconstruction times were ~ 30 seconds, approximately 240 times faster than an NR-PROST reconstruction. Image quality comparable to the reference images was achieved using the proposed MoCo-MoDL framework, with no statistically significant differences found in any of the assessed quantitative or qualitative image quality measures. Additionally, expert image quality scores indicated the MoCo-MoDL reconstructions were consistently of a higher quality than the NR-PROST reconstructions of the same data, with the differences in 12 of the 22 scores measured for individual vascular structures found to be statistically significant.

CONCLUSION: The MoCo-MoDL framework was applied to an adult CHD patient cohort, achieving good quality 3D whole-heart images from ~ 2-minute scans with reconstruction times of ~ 30 seconds.

PMID:38521391 | DOI:10.1016/j.jocmr.2024.101039

J Biomed Inform. 2024 Mar 21:104626. doi: 10.1016/j.jbi.2024.104626. Online ahead of print.

ABSTRACT

OBJECTIVE: The accuracy of deep learning models for many disease prediction problems is affected by time-varying covariates, rare incidence, covariate imbalance and delayed diagnosis when using structured electronic health records data. The situation is further exasperated when predicting the risk of one disease on condition of another disease, such as the hepatocellular carcinoma risk among patients with nonalcoholic fatty liver disease due to slow, chronic progression, the scarce of data with both disease conditions and the sex bias of the diseases. The goal of this study is to investigate the extent to which the aforementioned issues influence deep learning performance, and then devised strategies to tackle these challenges. These strategies were applied to improve hepatocellular carcinoma risk prediction among patients with nonalcoholic fatty liver disease.

METHODS: We evaluated two representative deep learning models in the task of predicting the occurrence of hepatocellular carcinoma in a cohort of patients with nonalcoholic fatty liver disease (n = 220,838) from a national EHR database. The disease prediction task was carefully formulated as a classification problem while taking censorship and the length of follow-up into consideration.

RESULTS: We developed a novel backward masking scheme to deal with the issue of delayed diagnosis which is very common in EHR data analysis and evaluate how the length of longitudinal information after the index date affects disease prediction. We observed that modeling time-varying covariates improved the performance of the algorithms and transfer learning mitigated reduced performance caused by the lack of data. In addition, covariate imbalance, such as sex bias in data impaired performance. Deep learning models trained on one sex and evaluated in the other sex showed reduced performance, indicating the importance of assessing covariate imbalance while preparing data for model training.

CONCLUSIONS: The strategies developed in this work can significantly improve the performance of hepatocellular carcinoma risk prediction among patients with nonalcoholic fatty liver disease. Furthermore, our novel strategies can be generalized to apply to other disease risk predictions using structured electronic health records, especially for disease risks on condition of another disease.

PMID:38521180 | DOI:10.1016/j.jbi.2024.104626

AJNR Am J Neuroradiol. 2024 Mar 23:ajnr.A8252. doi: 10.3174/ajnr.A8252. Online ahead of print.

ABSTRACT

BACKGROUND AND PURPOSE: Interest in artificial intelligence (AI) and machine learning (ML) has been growing in neuroradiology, but there is limited knowledge on how this interest has manifested into research and specifically, its qualities and characteristics. This study aims to characterize the emergence and evolution of AI/ML articles within neuroradiology and provide a comprehensive overview of the trends, challenges, and future directions of the field.

MATERIALS AND METHODS: We performed a bibliometric analysis of the American Journal of Neuroradiology (AJNR): the journal was queried for original research articles published since inception (Jan. 1, 1980) to Dec. 3, 2022 that contained any of the following key terms: "machine learning", "artificial intelligence", "radiomics", "deep learning", "neural network", "generative adversarial network", "object detection", or "natural language processing". Articles were screened by two independent reviewers, and categorized into Statistical Modelling (Type 1), AI/ML Development (Type 2), both representing developmental research work but without a direct clinical integration, or End-user Application (Type 3) which is the closest surrogate of potential AI/ML integration into day-to-day practice. To better understand the limiting factors to Type 3 articles being published, we analyzed Type 2 articles as they should represent the precursor work leading to Type 3.

RESULTS: A total of 182 articles were identified with 79% being non-integration focused (Type 1 n = 53, Type 2 n = 90) and 21% (n = 39) being Type 3. The total number of articles published grew roughly five-fold in the last five years, with the non-integration focused articles mainly driving this growth. Additionally, a minority of Type 2 articles addressed bias (22%) and explainability (16%). These articles were primarily led by radiologists (63%), with most of them (60%) having additional postgraduate degrees.

CONCLUSIONS: AI/ML publications have been rapidly increasing in neuroradiology with only a minority of this growth being attributable to end-user application. Areas identified for improvement include enhancing the quality of Type 2 articles, namely external validation, and addressing both bias and explainability. These results ultimately provide authors, editors, clinicians, and policymakers important insights to promote a shift towards integrating practical AI/ML solutions in neuroradiology.

ABBREVIATIONS: AI = artificial intelligence; ML = machine learning.

PMID:38521092 | DOI:10.3174/ajnr.A8252

Neural Netw. 2024 Mar 16;174:106242. doi: 10.1016/j.neunet.2024.106242. Online ahead of print.

ABSTRACT

In this paper, we introduce PDE-LEARN, a novel deep learning algorithm that can identify governing partial differential equations (PDEs) directly from noisy, limited measurements of a physical system of interest. PDE-LEARN uses a Rational Neural Network, U, to approximate the system response function and a sparse, trainable vector, ξ, to characterize the hidden PDE that the system response function satisfies. Our approach couples the training of U and ξ using a loss function that (1) makes U approximate the system response function, (2) encapsulates the fact that U satisfies a hidden PDE that ξ characterizes, and (3) promotes sparsity in ξ using ideas from iteratively reweighted least-squares. Further, PDE-LEARN can simultaneously learn from several data sets, allowing it to incorporate results from multiple experiments. This approach yields a robust algorithm to discover PDEs directly from realistic scientific data. We demonstrate the efficacy of PDE-LEARN by identifying several PDEs from noisy and limited measurements.

PMID:38521016 | DOI:10.1016/j.neunet.2024.106242

Comput Biol Med. 2024 Mar 20;173:108254. doi: 10.1016/j.compbiomed.2024.108254. Online ahead of print.

ABSTRACT

Chronic Lung Allograft Dysfunction (CLAD) is a critical post-transplant complication that predominantly determines the long-term survival rates and quality of life of patients undergoing lung transplantation. The limited efficacy of current immunosuppressive strategies underscores our incomplete understanding of the immunological aspects of CLAD. Hence, there is an urgent need for more comprehensive and targeted research to unravel the complex interplay of immune cells in the development and progression of CLAD. This study conducts an in-depth analysis of the immune environment in CLAD. By examining the gene expression profiles of T cells, natural killer cells, B cells, macrophages, and monocytes, we have elucidated a unique immunological landscape in CLAD compared to healthy controls. We highlight the heterogeneity within the immune populations and provide a comprehensive understanding of the immune mechanisms driving CLAD. Enrichment analysis identified specific pathways that are either overactive or suppressed in CLAD, revealing potential molecular targets for therapeutic intervention. Our findings emphasize the crucial role of T cells in the pathophysiology of CLAD, coordinating the immune response and revealing an amplified immune cell network, potentially leading to maladaptive tissue responses. By integrating a comprehensive cellular and molecular portrait of the immune environment, our research not only deepens our understanding of the pathogenesis of CLAD but also lays a foundational approach for the development of targeted therapies.

PMID:38520924 | DOI:10.1016/j.compbiomed.2024.108254

Comput Biol Med. 2024 Mar 11;173:108257. doi: 10.1016/j.compbiomed.2024.108257. Online ahead of print.

ABSTRACT

We developed an attention model to predict future adverse glycemic events 30 min in advance based on the observation of past glycemic values over a 35 min period. The proposed model effectively encodes insulin administration and meal intake time using Time2Vec (T2V) for glucose prediction. The proposed impartial feature selection algorithm is designed to distribute rewards proportionally according to agent contributions. Agent contributions are calculated by a step-by-step negation of updated agents. Thus, the proposed feature selection algorithm optimizes features from electronic medical records to improve performance. For evaluation, we collected continuous glucose monitoring data from 102 patients with type 2 diabetes admitted to Cheonan Hospital, Soonchunhyang University. Using our proposed model, we achieved F1-scores of 89.0%, 60.6%, and 89.8% for normoglycemia, hypoglycemia, and hyperglycemia, respectively.

PMID:38520922 | DOI:10.1016/j.compbiomed.2024.108257

Ultrasonics. 2024 Jan 29;140:107251. doi: 10.1016/j.ultras.2024.107251. Online ahead of print.

ABSTRACT

Lung ultrasound (LUS) has emerged as a safe and cost-effective modality for assessing lung health, particularly during the COVID-19 pandemic. However, interpreting LUS images remains challenging due to its reliance on artefacts, leading to operator variability and limiting its practical uptake. To address this, we propose a deep learning pipeline for multi-class segmentation of objects (ribs, pleural line) and artefacts (A-lines, B-lines, B-line confluence) in ultrasound images of a lung training phantom. Lightweight models achieved a mean Dice Similarity Coefficient (DSC) of 0.74, requiring fewer than 500 training images. Applying this method in real-time, at up to 33.4 frames per second in inference, allows enhanced visualisation of these features in LUS images. This could be useful in providing LUS training and helping to address the skill gap. Moreover, the segmentation masks obtained from this model enable the development of explainable measures of disease severity, which have the potential to assist in the triage and management of patients. We suggest one such semi-quantitative measure called the B-line Artefact Score, which is related to the percentage of an intercostal space occupied by B-lines and in turn may be associated with the severity of a number of lung conditions. Moreover, we show how transfer learning could be used to train models for small datasets of clinical LUS images, identifying pathologies such as simple pleural effusions and lung consolidation with DSC values of 0.48 and 0.32 respectively. Finally, we demonstrate how such DL models could be translated into clinical practice, implementing the phantom model alongside a portable point-of-care ultrasound system, facilitating bedside assessment and improving the accessibility of LUS.

PMID:38520819 | DOI:10.1016/j.ultras.2024.107251

Eur J Radiol. 2024 Mar 20;175:111434. doi: 10.1016/j.ejrad.2024.111434. Online ahead of print.

ABSTRACT

PURPOSE: Artificial intelligence and deep learning solutions are increasingly utilized in healthcare and radiology. The number of studies addressing their enhancement of productivity and monetary impact is, however, still limited. Our hospital has faced a need to enhance MRI scanner throughput, and we investigate the utility of new commercial deep learning reconstruction (DLR) algorithm for this purpose. In this work, a multidisciplinary team evaluated the impact of the widespread deployment of a new commercial deep learning reconstruction (DLR) algorithm for our magnetic resonance imaging scanner fleet.

METHODS: Our analysis centers on the DLR algorithm's effects on patient throughput and investment costs, contrasting these with alternative strategies for capacity expansion-namely, acquiring additional MRI scanners and increasing device utilization on weekends. We provide a framework for assessing the financial implications of new technologies in a trial phase, aiding in informed decision-making for healthcare investments.

RESULTS: We demonstrate substantial reductions in total operating costs compared to other capacity-enhancing methods. Specifically, the cost of adopting the deep learning technology for our entire scanner fleet is only 11 % compared to procuring an additional scanner and 20 % compared to the weekend utilization costs of existing devices.

CONCLUSIONS: Procuring DLR for our existing five-scanner fleet allows us to sustain our current MRI service levels without the need for an additional scanner, thereby achieving considerable cost savings. These reductions highlight the efficiency and economic viability of DLR in optimizing MRI service delivery.

PMID:38520806 | DOI:10.1016/j.ejrad.2024.111434

J Am Med Inform Assoc. 2024 Mar 23:ocae060. doi: 10.1093/jamia/ocae060. Online ahead of print.

ABSTRACT

OBJECTIVES: Leveraging artificial intelligence (AI) in conjunction with electronic health records (EHRs) holds transformative potential to improve healthcare. However, addressing bias in AI, which risks worsening healthcare disparities, cannot be overlooked. This study reviews methods to handle various biases in AI models developed using EHR data.

MATERIALS AND METHODS: We conducted a systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines, analyzing articles from PubMed, Web of Science, and IEEE published between January 01, 2010 and December 17, 2023. The review identified key biases, outlined strategies for detecting and mitigating bias throughout the AI model development, and analyzed metrics for bias assessment.

RESULTS: Of the 450 articles retrieved, 20 met our criteria, revealing 6 major bias types: algorithmic, confounding, implicit, measurement, selection, and temporal. The AI models were primarily developed for predictive tasks, yet none have been deployed in real-world healthcare settings. Five studies concentrated on the detection of implicit and algorithmic biases employing fairness metrics like statistical parity, equal opportunity, and predictive equity. Fifteen studies proposed strategies for mitigating biases, especially targeting implicit and selection biases. These strategies, evaluated through both performance and fairness metrics, predominantly involved data collection and preprocessing techniques like resampling and reweighting.

DISCUSSION: This review highlights evolving strategies to mitigate bias in EHR-based AI models, emphasizing the urgent need for both standardized and detailed reporting of the methodologies and systematic real-world testing and evaluation. Such measures are essential for gauging models' practical impact and fostering ethical AI that ensures fairness and equity in healthcare.

PMID:38520723 | DOI:10.1093/jamia/ocae060

Adv Sci (Weinh). 2024 Mar 23:e2310141. doi: 10.1002/advs.202310141. Online ahead of print.

ABSTRACT

Fiber-reinforced hydrogel composites are widely employed in many engineering applications, such as drug release, and flexible electronics, with more flexible mechanical properties than pure hydrogel materials. Comparing to the hydrogel strengthened by continuous fiber, the meta-fiber reinforced hydrogel provides stronger individualized design ability of deformation patterns and tunable stiffness, especially for the elaborate applications in joint, cartilage, and organ. In this paper, a novel structure design strategy based on deep learning algorithm is proposed for hydrogel reinforced by meta-fiber to achieve targeted mechanical properties, such as stress and displacement fields. A solid mechanic model for meta-fiber reinforced hydrogel is first developed to construct the dataset of fiber distribution and the corresponding mechanical properties of the composite. Generative adversarial network (GAN) is then trained to characterize the relationship between stress or displacement field, and meta-fiber distribution. The well-trained GAN is implemented to design meta-fiber reinforced hydrogel composite structure under specific operation conditions. The results show that the deep learning method may efficiently predict the structure of the hydrogel composite with satisfied confidence, and has great potential for applications in drug delivery and flexible electronics.

PMID:38520708 | DOI:10.1002/advs.202310141

Laryngoscope. 2024 Mar 23. doi: 10.1002/lary.31413. Online ahead of print.

ABSTRACT

OBJECTIVE: Computer aided diagnostics (CAD) systems can automate the differentiation of maxillary sinus (MS) with and without opacification, simplifying the typically laborious process and aiding in clinical insight discovery within large cohorts.

METHODS: This study uses Hamburg City Health Study (HCHS) a large, prospective, long-term, population-based cohort study of participants between 45 and 74 years of age. We develop a CAD system using an ensemble of 3D Convolutional Neural Network (CNN) to analyze cranial MRIs, distinguishing MS with opacifications (polyps, cysts, mucosal thickening) from MS without opacifications. The system is used to find correlations of participants with and without MS opacifications with clinical data (smoking, alcohol, BMI, asthma, bronchitis, sex, age, leukocyte count, C-reactive protein, allergies).

RESULTS: The evaluation metrics of CAD system (Area Under Receiver Operator Characteristic: 0.95, sensitivity: 0.85, specificity: 0.90) demonstrated the effectiveness of our approach. MS with opacification group exhibited higher alcohol consumption, higher BMI, higher incidence of intrinsic asthma and extrinsic asthma. Male sex had higher prevalence of MS opacifications. Participants with MS opacifications had higher incidence of hay fever and house dust allergy but lower incidence of bee/wasp venom allergy.

CONCLUSION: The study demonstrates a 3D CNN's ability to distinguish MS with and without opacifications, improving automated diagnosis and aiding in correlating clinical data in population studies.

LEVEL OF EVIDENCE: 3 Laryngoscope, 2024.

PMID:38520698 | DOI:10.1002/lary.31413

Int J Comput Assist Radiol Surg. 2024 Mar 23. doi: 10.1007/s11548-024-03077-3. Online ahead of print.

ABSTRACT

PURPOSE: Accurate and rapid needle localization on 3D magnetic resonance imaging (MRI) is critical for MRI-guided percutaneous interventions. The current workflow requires manual needle localization on 3D MRI, which is time-consuming and cumbersome. Automatic methods using 2D deep learning networks for needle segmentation require manual image plane localization, while 3D networks are challenged by the need for sufficient training datasets. This work aimed to develop an automatic deep learning-based pipeline for accurate and rapid 3D needle localization on in vivo intra-procedural 3D MRI using a limited training dataset.

METHODS: The proposed automatic pipeline adopted Shifted Window (Swin) Transformers and employed a coarse-to-fine segmentation strategy: (1) initial 3D needle feature segmentation with 3D Swin UNEt TRansfomer (UNETR); (2) generation of a 2D reformatted image containing the needle feature; (3) fine 2D needle feature segmentation with 2D Swin Transformer and calculation of 3D needle tip position and axis orientation. Pre-training and data augmentation were performed to improve network training. The pipeline was evaluated via cross-validation with 49 in vivo intra-procedural 3D MR images from preclinical pig experiments. The needle tip and axis localization errors were compared with human intra-reader variation using the Wilcoxon signed rank test, with p < 0.05 considered significant.

RESULTS: The average end-to-end computational time for the pipeline was 6 s per 3D volume. The median Dice scores of the 3D Swin UNETR and 2D Swin Transformer in the pipeline were 0.80 and 0.93, respectively. The median 3D needle tip and axis localization errors were 1.48 mm (1.09 pixels) and 0.98°, respectively. Needle tip localization errors were significantly smaller than human intra-reader variation (median 1.70 mm; p < 0.01).

CONCLUSION: The proposed automatic pipeline achieved rapid pixel-level 3D needle localization on intra-procedural 3D MRI without requiring a large 3D training dataset and has the potential to assist MRI-guided percutaneous interventions.

PMID:38520646 | DOI:10.1007/s11548-024-03077-3

Environ Monit Assess. 2024 Mar 23;196(4):393. doi: 10.1007/s10661-024-12549-7.

ABSTRACT

Tropospheric ozone is an air pollutant at the ground level and a greenhouse gas which significantly contributes to the global warming. Strong anthropogenic emissions in and around urban environments enhance surface ozone pollution impacting the human health and vegetation adversely. However, observations are often scarce and the factors driving ozone variability remain uncertain in the developing regions of the world. In this regard, here, we conducted machine learning (ML) simulations of ozone variability and comprehensively examined the governing factors over a major urban environment (Ahmedabad) in western India. Ozone precursors (NO2, NO, CO, C5H8 and CH2O) from the CAMS (Copernicus Atmosphere Monitoring Service) reanalysis and meteorological parameters from the ERA5 (European Centre for Medium-Range Weather Forecast's (ECMWF) fifth-generation reanalysis) were included as features in the ML models. Automated ML (AutoML) fitted the deep learning model optimally and simulated the daily ozone with root mean square error (RMSE) of ~2 ppbv reproducing 84-88% of variability. The model performance achieved here is comparable to widely used ML models (RF-Random Forest and XGBoost-eXtreme Gradient Boosting). Explainability of the models is discussed through different schemes of feature importance, including SAGE (Shapley Additive Global importancE) and permutation importance. The leading features are found to be different from different feature importance schemes. We show that urban ozone could be simulated well (RMSE = 2.5 ppbv and R2 = 0.78) by considering first four leading features, from different schemes, which are consistent with ozone photochemistry. Our study underscores the need to conduct science-informed analysis of feature importance from multiple schemes to infer the roles of input variables in ozone variability. AutoML-based studies, exploiting potentials of long-term observations, can strongly complement the conventional chemistry-transport modelling and can also help in accurate simulation and forecast of urban ozone.

PMID:38520559 | DOI:10.1007/s10661-024-12549-7

Abdom Radiol (NY). 2024 Mar 23. doi: 10.1007/s00261-024-04256-1. Online ahead of print.

ABSTRACT

PURPOSE: To compare a previous model-based image reconstruction (MBIR) with a newly developed deep learning (DL)-based image reconstruction for providing improved signal-to-noise ratio (SNR) in high through-plane resolution (1 mm) T2-weighted spin-echo (T2SE) prostate MRI.

METHODS: Large-area contrast and high-contrast spatial resolution of the reconstruction methods were assessed quantitatively in experimental phantom studies. The methods were next evaluated radiologically in 17 subjects at 3.0 Tesla for whom prostate MRI was clinically indicated. For each subject, the axial T2SE raw data were directed to MBIR and to the DL reconstruction at three vendor-provided levels: (L)ow, (M)edium, and (H)igh. Thin-slice images from the four reconstructions were compared using evaluation criteria related to SNR, sharpness, contrast fidelity, and reviewer preference. Results were compared using the Wilcoxon signed-rank test using Bonferroni correction, and inter-reader comparisons were done using the Cohen and Krippendorf tests.

RESULTS: Baseline contrast and resolution in phantom studies were equivalent for all four reconstruction pathways as desired. In vivo, all three DL levels (L, M, H) provided improved SNR versus MBIR. For virtually, all other evaluation criteria DL L and M were superior to MBIR. DL L and M were evaluated as superior to DL H in fidelity of contrast. For 44 of the 51 evaluations, the DL M reconstruction was preferred.

CONCLUSION: The deep learning reconstruction method provides significant SNR improvement in thin-slice (1 mm) T2SE images of the prostate while retaining image contrast. However, if taken to too high a level (DL High), both radiological sharpness and fidelity of contrast diminish.

PMID:38520510 | DOI:10.1007/s00261-024-04256-1

Primates. 2024 Mar 23. doi: 10.1007/s10329-024-01123-x. Online ahead of print.

ABSTRACT

The latest advances in artificial intelligence technology have opened doors to the video analysis of complex behaviours. In light of this, ethologists are actively exploring the potential of these innovations to streamline the time-intensive behavioural analysis process using video data. Several tools have been developed for this purpose in primatology in the past decade. Nonetheless, each tool grapples with technical constraints. To address these limitations, we have established a comprehensive protocol designed to harness the capabilities of a cutting-edge artificial intelligence-assisted software, LabGym. The primary objective of this study was to evaluate the suitability of LabGym for the analysis of primate behaviour, focusing on Japanese macaques as our model subjects. First, we developed a model that accurately detects Japanese macaques, allowing us to analyse their actions using LabGym. Our behavioural analysis model succeeded in recognising stone-handling-like behaviours on video. However, the absence of quantitative data within the specified time frame limits the ability of our study to draw definitive conclusions regarding the quality of the behavioural analysis. Nevertheless, to the best of our knowledge, this study represents the first instance of applying the LabGym tool specifically for the analysis of primate behaviours, with our model focusing on the automated recognition and categorisation of specific behaviours in Japanese macaques. It lays the groundwork for future research in this promising field to complexify our model using the latest version of LabGym and associated tools, such as multi-class detection and interactive behaviour analysis.

PMID:38520479 | DOI:10.1007/s10329-024-01123-x