BMC Vet Res. 2024 Sep 9;20(1):401. doi: 10.1186/s12917-024-04242-1.

ABSTRACT

Successful identification of estrum or other stages in a cycling bitch often requires a combination of methods, including assessment of its behavior, exfoliative vaginal cytology, vaginoscopy, and hormonal assays. Vaginoscopy is a handy and inexpensive tool for the assessment of the breeding period. The present study introduces an innovative method for identifying the stages in the estrous cycle of female canines. With a dataset of 210 vaginoscopic images covering four reproductive stages, this approach extracts deep features using the inception v3 and Residual Networks (ResNet) 152 models. Binary gray wolf optimization (BGWO) is applied for feature optimization, and classification is performed with the extreme gradient boosting (XGBoost) algorithm. Both models are compared with the support vector machine (SVM) with the Gaussian and linear kernel, k-nearest neighbor (KNN), and convolutional neural network (CNN), based on performance metrics such as accuracy, specificity, F1 score, sensitivity, precision, matthew correlation coefficient (MCC), and runtime. The outcomes demonstrate the superiority of the deep model of ResNet 152 with XGBoost classifier, achieving an average model accuracy of 90.37%. The method gave a specific accuracy of 90.91%, 96.38%, 88.37%, and 88.24% in predicting the proestrus, estrus, diestrus, and anestrus stages, respectively. When performing deep feature analysis using inception v3 with the same classifiers, the model achieved an accuracy of 89.41%, which is comparable to the results obtained with the ResNet model. The proposed model offers a reliable system for identifying the optimal mating period, providing breeders and veterinarians with an efficient tool to enhance the success of their breeding programs.

PMID:39245728 | DOI:10.1186/s12917-024-04242-1

Sci Rep. 2024 Sep 8;14(1):20908. doi: 10.1038/s41598-024-71873-7.

ABSTRACT

The global interest in market prediction has driven the adoption of advanced technologies beyond traditional statistical models. This paper explores the use of machine learning and deep learning techniques for stock market forecasting. We propose a comprehensive approach that includes efficient feature selection, data preprocessing, and classification methodologies. The wavelet transform method is employed for data cleaning and noise reduction. Feature selection is optimized using the Dandelion Optimization Algorithm (DOA), identifying the most relevant input features. A novel hybrid model, 3D-CNN-GRU, integrating a 3D convolutional neural network with a gated recurrent unit, is developed for stock market data analysis. Hyperparameter tuning is facilitated by the Blood Coagulation Algorithm (BCA), enhancing model performance. Our methodology achieves a remarkable prediction accuracy of 99.14%, demonstrating robustness and efficacy in stock market forecasting applications. While our model shows significant promise, it is limited by the scope of the dataset, which includes only the Nifty 50 index. Broader implications of this work suggest that incorporating additional datasets and exploring different market scenarios could further validate and enhance the model's applicability. Future research could focus on implementing this approach in varied financial contexts to ensure robustness and generalizability.

PMID:39245700 | DOI:10.1038/s41598-024-71873-7

Sci Rep. 2024 Sep 8;14(1):20915. doi: 10.1038/s41598-024-71763-y.

ABSTRACT

This paper presents the design and development of a coastal fisheries monitoring system that harnesses artificial intelligence technologies. Application of the system across the Pacific region promises to revolutionize coastal fisheries management. The program is built on a centralized, cloud-based monitoring system to automate data extraction and analysis processes. The system leverages YoloV4, OpenCV, and ResNet101 to extract information from images of fish and invertebrates collected as part of in-country monitoring programs overseen by national fisheries authorities. As of December 2023, the system has facilitated automated identification of over six hundred nearshore finfish species, and automated length and weight measurements of more than 80,000 specimens across the Pacific. The system integrates other key fisheries monitoring data such as catch rates, fishing locations and habitats, volumes, pricing, and market characteristics. The collection of these metrics supports much needed rapid fishery assessments. The system's co-development with national fisheries authorities and the geographic extent of its application enables capacity development and broader local inclusion of fishing communities in fisheries management. In doing so, the system empowers fishers to work with fisheries authorities to enable data-informed decision-making for more effective adaptive fisheries management. The system overcomes historically entrenched technical and financial barriers in fisheries management in many Pacific island communities.

PMID:39245678 | DOI:10.1038/s41598-024-71763-y

Sci Total Environ. 2024 Sep 6:176095. doi: 10.1016/j.scitotenv.2024.176095. Online ahead of print.

ABSTRACT

Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are widely used across a spectrum of industrial and consumer goods. Nonetheless, their persistent nature and tendency to accumulate in biological systems pose substantial environmental and health threats. Consequently, striking a balance between maximizing product efficiency and minimizing environmental and health risks by tailoring the molecular structure of PFAS has become a pivotal challenge in the fields of environmental chemistry and sustainable development. To address this issue, a computational workflow was proposed for designing an environmentally friendly PFAS by incorporating deep learning (DL) and molecular generative models. The hybrid DL architecture MolHGT+ based on heterogeneous graph neural network with transformer-like attention was applied to predict the surface tension, bioaccumulation, and hepatotoxicity of the molecules. Through virtual screening of the PFAS master database using MolHGT+, the findings indicate that incorporating the siloxane group and betaine fragment can effectively decrease both the bioaccumulation and hepatotoxicity of PFAS while preserving low surface tension. In addition, molecular generative models were employed to create a structurally diverse pool of novel PFASs with the aforementioned hit molecules serving as the initial template structures. Overall, our study presents a promising AI-driven method for advancing the development of environmentally friendly PFAS.

PMID:39245376 | DOI:10.1016/j.scitotenv.2024.176095

Heart Rhythm. 2024 Sep 6:S1547-5271(24)03293-4. doi: 10.1016/j.hrthm.2024.09.003. Online ahead of print.

ABSTRACT

Sudden cardiac death (SCD) remains a pressing health issue, affecting hundreds of thousands each year globally. The heterogeneity among SCD victims, ranging from individuals with severe heart failure to seemingly healthy individuals, poses a significant challenge for effective risk assessment. Conventional risk stratification, which primarily relies on left ventricular ejection fraction, has resulted in only modest efficacy of implantable cardioverter-defibrillators (ICD) for SCD prevention. In response, artificial intelligence (AI) holds promise for personalised SCD risk prediction and tailoring preventive strategies to the unique profiles of individual patients. Machine and deep learning algorithms have the capability to learn intricate non-linear patterns between complex data and defined endpoints, and leverage these to identify subtle indicators and predictors of SCD that may not be apparent through traditional statistical analysis. However, despite the potential of AI to improve SCD risk stratification, there are important limitations that need to be addressed. We aim to provide an overview of the current state-of-the-art of AI prediction models for SCD, highlight the opportunities for these models in clinical practice, and identify the key challenges hindering widespread adoption.

PMID:39245250 | DOI:10.1016/j.hrthm.2024.09.003

J Dairy Sci. 2024 Sep 6:S0022-0302(24)01108-1. doi: 10.3168/jds.2023-24513. Online ahead of print.

ABSTRACT

The growing use of automated systems in the dairy industry generates a vast amount of cow-level data daily, creating opportunities for using these data to support real-time decision-making. Currently, various commercial systems offer built-in alert algorithms to identify cows requiring attention. To our knowledge, no work has been done to compare the use of models accounting for herd-level variability on their predictive ability against automated systems. Long Short-Term Memory (LSTM) models are machine learning models capable of learning temporal patterns and making predictions based on time series data. The objective of our study was to evaluate the ability of LSTM models to identify a health alert associated with a ketosis diagnosis (HAK) using deviations of daily milk yield, milk FPR, number of successful milkings, rumination time, and activity index from the herd median by parity and DIM, considering various time series lengths and numbers of d before HAK. Additionally, we aimed to use Explainable Artificial Intelligence method to understand the relationships between input variables and model outputs. Data on daily milk yield, milk fat-to-protein ratio (FPR), number of successful milkings, rumination time, activity, and health events during 0 to 21 d in milk (DIM) were retrospectively obtained from a commercial Holstein dairy farm in northern Indiana from February 2020 to January 2023. A total of 1,743 cows were included in the analysis (non-HAK = 1,550; HAK = 193). Variables were transformed based on deviations from the herd median by parity and DIM. Six LSTM models were developed to identify HAK 1, 2, and 3 d before farm diagnosis using historic cow-level data with varying time series lengths. Model performance was assessed using repeated stratified 10-fold cross-validation for 20 repeats. The Shapley additive explanations framework (SHAP) was used for model explanation. Model accuracy was 83, 74, and 70%, balanced error rate was 17 to 18, 26 to 28, and 34%, sensitivity was 81 to 83, 71 to 74, and 62%, specificity was 83, 74, and 71%, positive predictive value was 38, 25 to 27, and 21%, negative predictive value was 97 to 98, 95 to 96, and 94%, and area under the curve was 0.89 to 0.90, 0.80 to 0.81, and 0.72 for models identifying HAK 1, 2, and 3 d before diagnosis, respectively. Performance declined as the time interval between identification and farm diagnosis increased, and extending the time series length did not improve model performance. Model explanation revealed that cows with lower milk yield, number of successful milkings, rumination time, and activity, and higher milk FPR compared with herdmates of the same parity and DIM were more likely to be classified as HAK. Our results demonstrate the potential of LSTM models in identifying HAK using deviations of daily milk production variables, rumination time, and activity index from the herd median by parity and DIM. Future studies are needed to evaluate the performance of health alerts using LSTM models controlling for herd-specific metrics against commercial built-in algorithms in multiple farms and for other disorders.

PMID:39245172 | DOI:10.3168/jds.2023-24513

Radiother Oncol. 2024 Sep 6:110525. doi: 10.1016/j.radonc.2024.110525. Online ahead of print.

ABSTRACT

BACKGROUND AND PURPOSE: Fast and automated generation of treatment plans is desirable for magnetic resonance imaging (MRI)-guided adaptive radiotherapy (MRIgART). This study proposed a novel patient-specific auto-planning method and validated its feasibility in improving the existing online planning workflow.

MATERIALS AND METHODS: Data from 40 patients with prostate cancer were collected retrospectively. A patient-specific auto-planning method was proposed to generate adaptive treatment plans. First, a population dose-prediction model (M0) was trained using data from previous patients. Second, a patient-specific model (Mps) was created for each new patient by fine-tuning M0 with the patient's data. Finally, an auto plan was optimized using the parameters derived from the predicted dose distribution by Mps. The auto plans were compared with manual plans in terms of plan quality, efficiency, dosimetric verification, and clinical evaluation.

RESULTS: The auto plans improved target coverage, reduced irradiation to the rectum, and provided comparable protection to other organs-at-risk. Target coverage for the planning target volume (+0.61 %, P = 0.023) and clinical target volume 4000 (+1.60 %, P < 0.001) increased. V2900cGy (-1.06 %, P = 0.004) and V1810cGy (-2.49 %, P < 0.001) to the rectal wall and V1810cGy (-2.82 %, P = 0.012) to the rectum were significantly reduced. The auto plans required less planning time (-3.92 min, P = 0.001), monitor units (-46.48, P = 0.003), and delivery time (-0.26 min, P = 0.004), and their gamma pass rates (3 %/2 mm) were higher (+0.47 %, P = 0.014).

CONCLUSION: The proposed patient-specific auto-planning method demonstrated a robust level of automation and was able to generate high-quality treatment plans in less time for MRIgART in prostate cancer.

PMID:39245067 | DOI:10.1016/j.radonc.2024.110525

J Breast Imaging. 2024 Sep 8:wbae056. doi: 10.1093/jbi/wbae056. Online ahead of print.

ABSTRACT

OBJECTIVE: To evaluate potential screening mammography performance and workload impact using a commercial artificial intelligence (AI)-based triage device in a population-based screening sample.

METHODS: In this retrospective study, a sample of 2129 women who underwent screening mammograms were evaluated. The performance of a commercial AI-based triage device was compared with radiologists' reports, actual outcomes, and national benchmarks using commonly used mammography metrics. Up to 5 years of follow-up examination results were evaluated in cases to establish benignity. The algorithm sorted cases into groups of "suspicious" and "low suspicion." A theoretical workload reduction was calculated by subtracting cases triaged as "low suspicion" from the sample.

RESULTS: At the default 93% sensitivity setting, there was significant improvement (P <.05) in the following triage simulation mean performance measures compared with actual outcome: 45.5% improvement in recall rate (13.4% to 7.3%; 95% CI, 6.2-8.3), 119% improvement in positive predictive value (PPV) 1 (5.3% to 11.6%; 95% CI, 9.96-13.4), 28.5% improvement in PPV2 (24.6% to 31.6%; 95% CI, 24.8-39.1), 20% improvement in sensitivity (83.3% to 100%; 95% CI, 100-100), and 7.2% improvement in specificity (87.2% to 93.5%; 95% CI, 92.4-94.5). A theoretical 62.5% workload reduction was possible. At the ultrahigh 99% sensitivity setting, a theoretical 27% workload reduction was possible. No cancers were missed by the algorithm at either sensitivity.

CONCLUSION: Artificial intelligence-based triage in this simulation demonstrated potential for significant improvement in mammography performance and predicted substantial theoretical workload reduction without any missed cancers.

PMID:39245042 | DOI:10.1093/jbi/wbae056

Comput Biol Med. 2024 Sep 7;182:109103. doi: 10.1016/j.compbiomed.2024.109103. Online ahead of print.

ABSTRACT

The lung is characterized by high elasticity and complex structure, which implies that the lung is capable of undergoing complex deformation and the shape variable is substantial. Large deformation estimation poses significant challenges to lung image registration. The traditional U-Net architecture is difficult to cover complex deformation due to its limited receptive field. Moreover, the relationship between voxels weakens as the number of downsampling times increases, that is, the long-range dependence issue. In this paper, we propose a novel multilevel registration framework which enhances the correspondence between voxels to improve the ability of estimating large deformations. Our approach consists of a convolutional neural network (CNN) with a two-stream registration structure and a cross-scale mapping attention (CSMA) mechanism. The former extracts the robust features of image pairs within layers, while the latter establishes frequent connections between layers to maintain the correlation of image pairs. This method fully utilizes the context information of different scales to establish the mapping relationship between low-resolution and high-resolution feature maps. We have achieved remarkable results on DIRLAB (TRE 1.56 ± 1.60) and POPI (NCC 99.72% SSIM 91.42%) dataset, demonstrating that this strategy can effectively address the large deformation issues, mitigate long-range dependence, and ultimately achieve more robust lung CT image registration.

PMID:39244962 | DOI:10.1016/j.compbiomed.2024.109103

Oral Oncol. 2024 Sep 7;159:107019. doi: 10.1016/j.oraloncology.2024.107019. Online ahead of print.

NO ABSTRACT

PMID:39244858 | DOI:10.1016/j.oraloncology.2024.107019

Ultrasound Med Biol. 2024 Sep 6:S0301-5629(24)00299-0. doi: 10.1016/j.ultrasmedbio.2024.08.002. Online ahead of print.

ABSTRACT

OBJECTIVE: As metabolic dysfunction-associated steatotic liver disease (MASLD) becomes more prevalent worldwide, it is imperative to create more accurate technologies that make it easy to assess the liver in a point-of-care setting. The aim of this study is to test the performance of a new software tool implemented in Velacur (Sonic Incytes), a liver stiffness and ultrasound attenuation measurement device, on patients with MASLD. This tool employs a deep learning-based method to detect and segment shear waves in the liver tissue for subsequent analysis to improve tissue characterization for patient diagnosis.

METHODS: This new tool consists of a deep learning based algorithm, which was trained on 15,045 expert-segmented images from 103 patients, using a U-Net architecture. The algorithm was then tested on 4429 images from 36 volunteers and patients with MASLD. Test subjects were scanned at different clinics with different Velacur operators. Evaluation was performed on both individual images (image based) and averaged across all images collected from a patient (patient based). Ground truth was defined by expert segmentation of the shear waves within each image. For evaluation, sensitivity and specificity for correct wave detection in the image were calculated. For those images containing waves, the Dice coefficient was calculated. A prototype of the software tool was also implemented on Velacur and assessed by operators in real world settings.

RESULTS: The wave detection algorithm had a sensitivity of 81% and a specificity of 84%, with a Dice coefficient of 0.74 and 0.75 for image based and patient-based averages respectively. The implementation of this software tool as an overlay on the B-Mode ultrasound resulted in improved exam quality collected by operators.

CONCLUSION: The shear wave algorithm performed well on a test set of volunteers and patients with metabolic dysfunction-associated steatotic liver disease. The addition of this software tool, implemented on the Velacur system, improved the quality of the liver assessments performed in a real world, point of care setting.

PMID:39244483 | DOI:10.1016/j.ultrasmedbio.2024.08.002

J Biomed Inform. 2024 Sep 5:104722. doi: 10.1016/j.jbi.2024.104722. Online ahead of print.

ABSTRACT

OBJECTIVE: Keratitis is the primary cause of corneal blindness worldwide. Prompt identification and referral of patients with keratitis are fundamental measures to improve patient prognosis. Although deep learning can assist ophthalmologists in automatically detecting keratitis through a slit lamp camera, remote and underserved areas often lack this professional equipment. Smartphones, a widely available device, have recently been found to have potential in keratitis screening. However, given the limited data available from smartphones, employing traditional deep learning algorithms to construct a robust intelligent system presents a significant challenge. This study aimed to propose a meta-learning framework, cosine nearest centroid-based metric learning (CNCML), for developing a smartphone-based keratitis screening model in the case of insufficient smartphone data by leveraging the prior knowledge acquired from slit-lamp photographs.

METHODS: We developed and assessed CNCML based on 13,009 slit-lamp photographs and 4,075 smartphone photographs that were obtained from 3 independent clinical centers. To mimic real-world scenarios with various degrees of sample scarcity, we used training sets of different sizes (0 to 20 photographs per class) from the HUAWEI smartphone to train CNCML. We evaluated the performance of CNCML not only on an internal test dataset but also on two external datasets that were collected by two different brands of smartphones (VIVO and XIAOMI) in another clinical center. Furthermore, we compared the performance of CNCML with that of traditional deep learning models on these smartphone datasets. The accuracy and macro-average area under the curve (macro-AUC) were utilized to evaluate the performance of models.

RESULTS: With merely 15 smartphone photographs per class used for training, CNCML reached accuracies of 84.59%, 83.15%, and 89.99% on three smartphone datasets, with corresponding macro-AUCs of 0.96, 0.95, and 0.98, respectively. The accuracies of CNCML on these datasets were 0.56% to 9.65% higher than those of the most competitive traditional deep learning models.

CONCLUSIONS: CNCML exhibited fast learning capabilities, attaining remarkable performance with a small number of training samples. This approach presents a potential solution for transitioning intelligent keratitis detection from professional devices (e.g., slit-lamp cameras) to more ubiquitous devices (e.g., smartphones), making keratitis screening more convenient and effective.

PMID:39244181 | DOI:10.1016/j.jbi.2024.104722

Sci Total Environ. 2024 Sep 5:176048. doi: 10.1016/j.scitotenv.2024.176048. Online ahead of print.

ABSTRACT

Accurately predicting tree mortality in mixed forests sets a challenge for conventional models because of large uncertainty, especially under changing climate. Machine learning algorithms had potential for predicting individual tree mortality with higher accuracy via filtering the relevant climatic and environmental factors. In this study, the sensitivity of individual tree mortality to regional climate was validated by modeling in seminatural mixed coniferous forests based on 25-yearobservations in northeast of China. Three advanced machine learning and deep learning algorithms were employed, including support vector machines, multi-layer perceptron, and random forests. Mortality was predicted by the effects of multiple inherent and environmental factors, including tree size and growth, topography, competition, stand structure and regional climate. All three types of models performed satisfactorily with their values of the areas under receiving operating characteristic curve (AUC) > 0.9. With tree growth, competition and regional climate as input variables, a model based on the random forests showed the highest values of the explained variance score (0.862) and AUC (0.914). Since the trees were vulnerable despite their species, mortality could occur after growth limit induced by insufficient or excessive sun radiation during growing seasons, cold threat caused thermal insufficiency in winters, and annual moisture constraints in these mixed coniferous forests. Our findings could enrich basic knowledge on individual tree mortality caused by water and heat inadequacy with the negative impacts of global warming global warming. Successful individual tree mortality modeling via advanced algorithms in mixed forests could assist in adaptive forest ecology modeling in large areas.

PMID:39244065 | DOI:10.1016/j.scitotenv.2024.176048

J Stomatol Oral Maxillofac Surg. 2024 Sep 5:102048. doi: 10.1016/j.jormas.2024.102048. Online ahead of print.

ABSTRACT

INTRODUCTION: In orthodontic treatments, accurately assessing the upper airway volume and morphology is essential for proper diagnosis and planning. Cone beam computed tomography (CBCT) is used for assessing upper airway volume through manual, semi-automatic, and automatic airway segmentation methods. This study evaluates upper airway segmentation accuracy by comparing the results of an automatic model and a semi-automatic method against the gold standard manual method.

MATERIALS AND METHODS: An automatic segmentation model was trained using the MONAI Label framework to segment the upper airway from CBCT images. An open-source program, ITK-SNAP, was used for semi-automatic segmentation. The accuracy of both methods was evaluated against manual segmentations. Evaluation metrics included Dice Similarity Coefficient (DSC), Precision, Recall, 95% Hausdorff Distance (HD), and volumetric differences.

RESULTS: The automatic segmentation group averaged a DSC score of 0.915±0.041, while the semi-automatic group scored 0.940±0.021, indicating clinically acceptable accuracy for both methods. Analysis of the 95% HD revealed that semi-automatic segmentation (0.997±0.585) was more accurate and closer to manual segmentation than automatic segmentation (1.447±0.674). Volumetric comparisons revealed no statistically significant differences between automatic and manual segmentation for total, oropharyngeal, and velopharyngeal airway volumes. Similarly, no significant differences were noted between the semi-automatic and manual methods across these regions.

CONCLUSION: It has been observed that both automatic and semi-automatic methods, which utilise open-source software, align effectively with manual segmentation. Implementing these methods can aid in decision-making by allowing faster and easier upper airway segmentation with comparable accuracy in orthodontic practice.

PMID:39244033 | DOI:10.1016/j.jormas.2024.102048

J Cardiovasc Magn Reson. 2024 Sep 5:101090. doi: 10.1016/j.jocmr.2024.101090. Online ahead of print.

ABSTRACT

BACKGROUND: Cardiovascular magnetic resonance (CMR) chemical shift encoding (CSE) enables myocardial fat imaging. We sought to develop a deep learning network (FastCSE) to accelerate CSE.

METHODS: FastCSE was built on a super-resolution generative adversarial network extended to enhance complex-valued image sharpness. FastCSE enhances each echo image independently before water-fat separation. FastCSE was trained with retrospectively identified cines from 1519 patients (56 ± 16 years; 866 men) referred for clinical 3T CMR. In a prospective study of 16 participants (58 ± 19 years; 7 females) and 5 healthy individuals (32 ± 17 years; 5 females), dual-echo CSE images were collected with 1.5 × 1.5mm2, 2.5 × 1.5 mm2, and 3.8 × 1.9mm2 resolution using generalized autocalibrating partially parallel acquisition (GRAPPA). FastCSE was applied to images collected with resolution of 2.5 × 1.5mm2 and 3.8 × 1.9 mm2 to restore sharpness. Fat images obtained from two-point Dixon reconstruction were evaluated using a quantitative blur metric and analyzed with 5-way analysis of variance.

RESULTS: FastCSE successfully reconstructed CSE images inline. FastCSE acquisition, with a resolution of 2.5 × 1.5mm² and 3.8 × 1.9 mm², reduced the number of breath-holds without impacting visualization of fat by approximately 1.5-fold and 3-fold compared to GRAPPA acquisition with a resolution of 1.5 × 1.5 mm², from 3.0 ± 0.8 breath-holds to 2.0 ± 0.2 and 1.1 ± 0.4 breath-holds, respectively. FastCSE improved image sharpness and removed ringing artifacts in GRAPPA fat images acquired with a resolution of 2.5 × 1.5 mm2 (0.31 ± 0.03 vs. 0.35 ± 0.04, P < 0.001) and 3.8 × 1.9 mm2 (0.31 ± 0.03 vs. 0.42 ± 0.06, P < 0.001). Blurring in FastCSE images was similar to blurring in images with 1.5 × 1.5 mm² resolution (0.32 ±0.03 vs. 0.31 ± 0.03, P = 0.78; 0.32 ± 0.03 vs. 0.31 ± 0.03, P = 0.90).

CONCLUSION: We showed that a deep learning-accelerated CSE technique based on complex-valued resolution enhancement can reduce the number of breath-holds in CSE imaging without impacting the visualization of fat. FastCSE showed similar image sharpness compared to a standardized parallel imaging method.

PMID:39243889 | DOI:10.1016/j.jocmr.2024.101090

Radiother Oncol. 2024 Sep 5:110522. doi: 10.1016/j.radonc.2024.110522. Online ahead of print.

ABSTRACT

BACKGROUND AND PURPOSE: This study aimed to evaluate the plan quality of our deep learning-based automated treatment planning method for robustly optimized intensity-modulated proton therapy (IMPT) plans in patients with oropharyngeal carcinoma (OPC). The assessment was conducted through a retrospective and prospective study, blindly comparing manual plans with deep learning plans.

MATERIALS AND METHODS: A set of 95 OPC patients were split into training (n = 60), configuration (n = 10), test retrospective study (n = 10) and test prospective study (n = 15). Our deep learning optimization (DLO) method combines IMPT dose prediction using a deep learning model with a robust mimicking optimization algorithm. Dosimetrists manually adjusted the DLO plan for individual patients. In both studies, manual plans and manually adjusted deep learning (mDLO) plans were blindly assessed by a radiation oncologist, a dosimetrist and a physicist, through visual inspection, clinical goal evaluation, and comparison of normal tissue complication probability values. mDLO plans were completed within an average time of 2.5 h. In comparison, the manual planning process typically took around 2 days.

RESULTS: In the retrospective study, in 10/10 (100%) patients, the mDLO plans were preferred, while in the prospective study, 9 out of 15 (60%) mDLO plans were preferred. In 4 out of the remaining 6 cases, the manual and mDLO plans were considered comparable in quality. Differences between manual and mDLO plans were limited.

CONCLUSION: This study showed a high preference for mDLO plans over manual IMPT plans, with 92% of cases considering mDLO plans comparable or superior in quality for OPC patients.

PMID:39243863 | DOI:10.1016/j.radonc.2024.110522

Asian J Psychiatr. 2024 Aug 30;101:104205. doi: 10.1016/j.ajp.2024.104205. Online ahead of print.

ABSTRACT

Mental disorders are becoming increasingly prevalent worldwide, and accurate incidence forecasting is crucial for effective mental health strategies. This study developed a long short-term memory (LSTM)-based recurrent neural network model to predict schizophrenia in inpatients in Taiwan. Data was collected on individuals aged over 20 years and diagnosed with schizophrenia between 1998 and 2015 from the National Health Insurance Research Database (NHIRD). The study compared six models, including LSTM, exponential smoothing, autoregressive integrated moving average, particle swarm optimization (PSO), PSO-based support vector regression, and deep neural network models, in terms of their predictive performance. The results showed that the LSTM model had the best accuracy, with the lowest mean absolute percentage error (2.34), root mean square error (157.42), and mean average error (154,831.70). This finding highlights the reliability of the LSTM model for forecasting mental disorder incidence. The study's findings provide valuable insights that can help government administrators devise clinical strategies for schizophrenia, and policymakers can use these predictions to formulate healthcare education and financial planning initiatives, fostering support networks for patients, caregivers, and the public.

PMID:39243662 | DOI:10.1016/j.ajp.2024.104205

Med Image Anal. 2024 Aug 31;99:103331. doi: 10.1016/j.media.2024.103331. Online ahead of print.

ABSTRACT

Multi-modal Magnetic Resonance Imaging (MRI) offers complementary diagnostic information, but some modalities are limited by the long scanning time. To accelerate the whole acquisition process, MRI reconstruction of one modality from highly under-sampled k-space data with another fully-sampled reference modality is an efficient solution. However, the misalignment between modalities, which is common in clinic practice, can negatively affect reconstruction quality. Existing deep learning-based methods that account for inter-modality misalignment perform better, but still share two main common limitations: (1) The spatial alignment task is not adaptively integrated with the reconstruction process, resulting in insufficient complementarity between the two tasks; (2) the entire framework has weak interpretability. In this paper, we construct a novel Deep Unfolding Network with Spatial Alignment, termed DUN-SA, to appropriately embed the spatial alignment task into the reconstruction process. Concretely, we derive a novel joint alignment-reconstruction model with a specially designed aligned cross-modal prior term. By relaxing the model into cross-modal spatial alignment and multi-modal reconstruction tasks, we propose an effective algorithm to solve this model alternatively. Then, we unfold the iterative stages of the proposed algorithm and design corresponding network modules to build DUN-SA with interpretability. Through end-to-end training, we effectively compensate for spatial misalignment using only reconstruction loss, and utilize the progressively aligned reference modality to provide inter-modality prior to improve the reconstruction of the target modality. Comprehensive experiments on four real datasets demonstrate that our method exhibits superior reconstruction performance compared to state-of-the-art methods.

PMID:39243598 | DOI:10.1016/j.media.2024.103331

Comput Biol Med. 2024 Sep 6;182:109101. doi: 10.1016/j.compbiomed.2024.109101. Online ahead of print.

ABSTRACT

The COVID-19 pandemic has driven substantial evolution of the SARS-CoV-2 virus, yielding subvariants that exhibit enhanced infectiousness in humans. However, this adaptive advantage may not universally extend to zoonotic transmission. In this work, we hypothesize that viral adaptations favoring animal hosts do not necessarily correlate with increased human infectivity. In addition, we consider the potential for gain-of-function mutations that could facilitate the virus's rapid evolution in humans following adaptation in animal hosts. Specifically, we identify the SARS-CoV-2 receptor-binding domain (RBD) mutations that enhance human-animal cross-transmission. To this end, we construct a multitask deep learning model, MT-TopLap trained on multiple deep mutational scanning datasets, to accurately predict the binding free energy changes upon mutation for the RBD to ACE2 of various species, including humans, cats, bats, deer, and hamsters. By analyzing these changes, we identified key RBD mutations such as Q498H in SARS-CoV-2 and R493K in the BA.2 variant that are likely to increase the potential for human-animal cross-transmission.

PMID:39243518 | DOI:10.1016/j.compbiomed.2024.109101

Neural Netw. 2024 Sep 3;180:106699. doi: 10.1016/j.neunet.2024.106699. Online ahead of print.

ABSTRACT

Despite significant success of deep learning in object detection tasks, the standard training of deep neural networks requires access to a substantial quantity of annotated images across all classes. Data annotation is an arduous and time-consuming endeavor, particularly when dealing with infrequent objects. Few-shot object detection (FSOD) methods have emerged as a solution to the limitations of classic object detection approaches based on deep learning. FSOD methods demonstrate remarkable performance by achieving robust object detection using a significantly smaller amount of training data. A challenge for FSOD is that instances from novel classes that do not belong to the fixed set of training classes appear in the background and the base model may pick them up as potential objects. These objects behave similarly to label noise because they are classified as one of the training dataset classes, leading to FSOD performance degradation. We develop a semi-supervised algorithm to detect and then utilize these unlabeled novel objects as positive samples during the FSOD training stage to improve FSOD performance. Specifically, we develop a hierarchical ternary classification region proposal network (HTRPN) to localize the potential unlabeled novel objects and assign them new objectness labels to distinguish these objects from the base training dataset classes. Our improved hierarchical sampling strategy for the region proposal network (RPN) also boosts the perception ability of the object detection model for large objects. We test our approach and COCO and PASCAL VOC baselines that are commonly used in FSOD literature. Our experimental results indicate that our method is effective and outperforms the existing state-of-the-art (SOTA) FSOD methods. Our implementation is provided as a supplement to support reproducibility of the results https://github.com/zshanggu/HTRPN.1.

PMID:39243514 | DOI:10.1016/j.neunet.2024.106699