Search

Abstract This study explores the effectiveness of Explainable Artificial Intelligence (XAI) for predicting suicide risk from medical tabular data. Given the common challenge of limited datasets in health-related Machine Learning (ML) applications, we use data augmentation in tandem with ML to enhance the identification of individuals at high risk of suicide. We use SHapley Additive exPlanations (SHAP) for XAI and traditional correlation analysis to rank feature importance, pinpointing primary factors influencing suicide risk and preventive measures. Experimental results show the Random Forest (RF) model is excelling in accuracy, F1 score, and AUC (>97% across metrics). According to SHAP, anger issues, depression, and social isolation emerge as top predictors of suicide risk, while individuals with high incomes, esteemed professions, and higher education present the lowest risk. Our findings underscore the effectiveness of ML and XAI in suicide risk assessment, offering valuable insights for psychiatrists and facilitating informed clinical decisions.

Ensuring the security and safety of passengers and cargo through effective baggage screening presents a critical challenge in high-traffic environments like airports, where traditional manual processes are affected by high error rates, fatigue among security personnel, and privacy concerns. These issues underscore the urgent need for sophisticated automated systems capable of accurately detecting concealed threats. Advancements in deep learning have introduced more refined approaches to baggage screening, yet they struggle with capturing the global context of threats and challenges presented by clutter, concealment, and occlusion. Addressing these issues, this paper presents the Multi-Scale Hierarchical Transformer for X-ray Detection (MHT-X), a novel framework that surpasses conventional limitations by integrating a multi-scale contour mapping technique with a hierarchical feature extraction process using visual Transformers. MHT-X is distinctive in its ability to generate multi-level contour maps that unveil complex features of threat objects, facilitating their accurate identification. It employs a top-down approach for hierarchical feature processing, capturing vital spatial characteristics essential for precise threat localization through an efficient encoder-decoder mechanism. MHT-X represents a significant advancement in the detection of concealed and obstructed threat objects in X-ray baggage scans. By combining multi-scale contour maps with a hierarchical feature extraction process, it significantly enhances detection accuracy and outperforms traditional Convolutional Neural Network (CNN) models by effectively capturing the global context of illegal items with advanced feature representation. The proposed framework is rigorously validated on two public datasets, dubbed CLCXray, and PIDray, where it outperforms the state-of-the-art (SOTA) methods by achieving the mean average precision score of 65.26%, and 78.19%, respectively.

Maize (Zea mays L.) has been shown to be sensitive to temperature deviations, influencing its yield potential. The development of new maize hybrids resilient to unfavourable weather is a desirable aim for crop breeders. In this paper, we showcase the development of a multimodal deep learning model using RGB images, phenotypic, and weather data under temporal effects to predict the yield potential of maize before or during anthesis and silking stages. The main objective of this study was to assess if the inclusion of historical weather data, maize growth captured through imagery, and important phenotypic traits would improve the predictive power of an established multimodal deep learning model. Evaluation of the model performance when training from scratch showed its ability to accurately predict ~89% of hybrids with high-yield potential and demonstrated enhanced explanatory power compared with previously published models. Shapley Additive explanations (SHAP) analysis indicated the top influential features include plant density, hybrid placement in the field, date to anthesis, parental line, temperature, humidity, and solar radiation. Including weather historical data was important for model performance, significantly enhancing the predictive and explanatory power of the model. For future research, the use of the model can move beyond maize yield prediction by fine-tuning the model on other crop data, serving as a potential decision-making tool for crop breeders to determine high-performing individuals from diverse crop types.

Weeds pose a significant threat to agricultural production, leading to substantial yield losses and increased herbicide usage, with severe economic and environmental implications. This paper uses deep learning to explore a novel approach via targeted segmentation mapping of crop plants rather than weeds, focusing on canola (Brassica napus) as the target crop. Multiple deep learning architectures (ResNet-18, ResNet-34, and VGG-16) were trained for the pixel-wise segmentation of canola plants in the presence of other plant species, assuming all non-canola plants are weeds. Three distinct datasets (T1_miling, T2_miling, and YC) containing 3799 images of canola plants in varying field conditions alongside other plant species were collected with handheld devices at 1.5 m. The top performing model, ResNet-34, achieved an average precision of 0.84, a recall of 0.87, a Jaccard index (IoU) of 0.77, and a Macro F1 score of 0.85, with some variations between datasets. This approach offers increased feature variety for model learning, making it applicable to the identification of a wide range of weed species growing among canola plants, without the need for separate weed datasets. Furthermore, it highlights the importance of accounting for the growth stage and positioning of plants in field conditions when developing weed detection models. The study contributes to the growing field of precision agriculture and offers a promising alternative strategy for weed detection in diverse field environments, with implications for the development of innovative weed control techniques.

Long non-coding ribonucleic acids (lncRNAs) have been shown to play an important role in plant gene regulation, involving both epigenetic and transcript regulation. LncRNAs are transcripts longer than 200 nucleotides that are not translated into functional proteins but can be translated into small peptides. Machine learning models have predominantly used transcriptome data with manually defined features to detect lncRNAs, however, they often underrepresent the abundance of lncRNAs and can be biased in their detection. Here we present a study using Natural Language Processing (NLP) models to identify plant lncRNAs from genomic sequences rather than transcriptomic data. The NLP models were trained to predict lncRNAs for seven model and crop species (Zea mays, Arabidopsis thaliana, Brassica napus, Brassica oleracea, Brassica rapa, Glycine max and Oryza sativa) using publicly available genomic references. We demonstrated that lncRNAs can be accurately predicted from genomic sequences with the highest accuracy of 83.4% for Z. mays and the lowest accuracy of 57.9% for B. rapa, revealing that genome assembly quality might affect the accuracy of lncRNA identification. Furthermore, we demonstrated the potential of using NLP models for cross-species prediction with an average of 63.1% accuracy using target species not previously seen by the model. As more species are incorporated into the training datasets, we expect the accuracy to increase, becoming a more reliable tool for uncovering novel lncRNAs. Finally, we show that the models can be interpreted using explainable artificial intelligence to identify motifs important to lncRNA prediction and that these motifs frequently flanked the lncRNA sequence.

Recently, transformers have achieved great success in a number of computer vision tasks due to their excellent ability to capture long-range feature dependencies. In contrast, convolutional neural networks (CNNs) are good at extracting local features. Given that the capture of short- and long-range band dependencies are both important for hyperspectral data classification, we propose MCE-ST, a convolution-transformer (conformer) based framework capable of exploiting the complementary strengths of transformers and CNNs. In contrast to the conventional transformer, which uses a linear projection for tokenization, the proposed MCE-ST uses a convolution-based tokenization method to extract local dependency between spectral bands. Moreover, since different hyperspectral samples may have different spans of local relationships, a multiscale conformer encoder (MCE) comprising two separate branches of depth-wise dilated convolution with different kernel sizes is used to extract the different spans of the local interactions between tokens. We conducted experiments on four salt stress datasets and one cassava disease dataset. The results show that the proposed MCE-ST outperforms the state-of-the-art techniques for crop stress classification using hyperspectral data. The code for MCE-ST is publicly available at https://github.com/Weejaa04/MCE-ST-GitHub.

Abstract Recent growth in crop genomic and trait data have opened opportunities for the application of novel approaches to accelerate crop improvement. Machine learning and deep learning are at the forefront of prediction-based data analysis. However, few approaches for genotype to phenotype prediction compare machine learning with deep learning and further interpret the models that support the predictions. This study uses genome wide molecular markers and traits across 1110 soybean individuals to develop accurate prediction models. For 13/14 sets of predictions, XGBoost or random forest outperformed deep learning models in prediction performance. Top ranked SNPs by F-score were identified from XGBoost, and with further investigation found overlap with significantly associated loci identified from GWAS and previous literature. Feature importance rankings were used to reduce marker input by up to 90%, and subsequent models maintained or improved their prediction performance. These findings support interpretable machine learning as an approach for genomic based prediction of traits in soybean and other crops.

Atrial fibrillation arises mainly due to abnormalities in the cardiac conduction system and is associated with anatomical remodeling of the atria and the pulmonary veins. Cardiovascular imaging techniques, such as echocardiography, computed tomography, and magnetic resonance imaging, are crucial in the management of atrial fibrillation, as they not only provide anatomical context to evaluate structural alterations but also help in determining treatment strategies. However, interpreting these images requires significant human expertise. The potential of artificial intelligence in analyzing these images has been repeatedly suggested due to its ability to automate the process with precision comparable to human experts. This review summarizes the benefits of artificial intelligence in enhancing the clinical care of patients with atrial fibrillation through cardiovascular image analysis. It provides a detailed overview of the two most critical steps in image-guided AF management, namely, segmentation and classification. For segmentation, the state-of-the-art artificial intelligence methodologies and the factors influencing the segmentation performance are discussed. For classification, the applications of artificial intelligence in the diagnosis and prognosis of atrial fibrillation are provided. Finally, this review also scrutinizes the current challenges hindering the clinical applicability of these methods, with the aim of guiding future research toward more effective integration into clinical practice.

IntroductionNatural language processing (NLP) uses various computational methods to analyse and understand human language, and has been applied to data acquired at Emergency Department (ED) triage to predict various outcomes. The objective of this scoping review is to evaluate how NLP has been applied to data acquired at ED triage, assess if NLP based models outperform humans or current risk stratification techniques when predicting outcomes, and assess if incorporating free-text improve predictive performance of models when compared to predictive models that use only structured data.MethodsAll English language peer-reviewed research that applied an NLP technique to free-text obtained at ED triage was eligible for inclusion. We excluded studies focusing solely on disease surveillance, and studies that used information obtained after triage. We searched the electronic databases MEDLINE, Embase, Cochrane Database of Systematic Reviews, Web of Science, and Scopus for medical subject headings and text keywords related to NLP and triage. Databases were last searched on 01/01/2022. Risk of bias in studies was assessed using the Prediction model Risk of Bias Assessment Tool (PROBAST). Due to the high level of heterogeneity between studies and high risk of bias, a metanalysis was not conducted. Instead, a narrative synthesis is provided.ResultsIn total, 3730 studies were screened, and 20 studies were included. The population size varied greatly between studies ranging from 1.8 million patients to 598 triage notes. The most common outcomes assessed were prediction of triage score, prediction of admission, and prediction of critical illness. NLP models achieved high accuracy in predicting need for admission, triage score, critical illness, and mapping free-text chief complaints to structured fields. Incorporating both structured data and free-text data improved results when compared to models that used only structured data. However, the majority of studies (80%) were assessed to have a high risk of bias, and only one study reported the deployment of an NLP model into clinical practice.ConclusionUnstructured free-text triage notes have been used by NLP models to predict clinically relevant outcomes. However, the majority of studies have a high risk of bias, most research is retrospective, and there are few examples of implementation into clinical practice. Future work is needed to prospectively assess if applying NLP to data acquired at ED triage improves ED outcomes when compared to usual clinical practice.

Hajj, the Muslim pilgrimage, is a large mass gathering event that involves performing rituals at several sites on specific days and times in a fixed order, thereby requiring transport of pilgrims between sites. For the past two decades, Hajj transport has relied on conventional and shuttle buses, train services, and pilgrims walking along pedestrian routes that link these sites. To ensure smooth and efficient transport during Hajj, specific groups of pilgrims are allocated with the cooperation of Hajj authorities to specific time windows, modes, and routes. However, the large number of pilgrims, delays and changes in bus schedules/timetables, and occasional lack of coordination between transport modes have often caused congestion or delays in pilgrim transfer between sites, with a cascading effect on transport management. This study focuses on modelling and simulating the transport of pilgrims between the sites using a discrete event simulation tool called "ExtendSim". Three transport modules were validated, and different scenarios were developed. These scenarios consider changes in the percentages of pilgrims allocated to each transport mode and the scheduling of various modes. The results can aid authorities to make informed decisions regarding transport strategies for managing the transport infrastructure and fleets. The proposed solutions could be implemented with judicious allocation of resources, through pre-event planning and real-time monitoring during the event.

Introduction Surveys conducted internationally have found widespread interest in artificial intelligence (AI) amongst medical students. No similar surveys have been conducted in Western Australia (WA) and it is not known how medical students in WA feel about the use of AI in healthcare or their understanding of AI. We aim to assess WA medical students’ attitudes towards AI in general, AI in healthcare, and the inclusion of AI education in the medical curriculum. Methods A digital survey instrument was developed based on a review of available literature and consultation with subject matter experts. The survey was piloted with a group of medical students and refined based on their feedback. We then sent this anonymous digital survey to all medical students in WA (approximately 1539 students). Responses were open from the 7th of September 2021 to the 7th of November 2021. Students’ categorical responses were qualitatively analysed, and free text comments from the survey were qualitatively analysed using open coding techniques. Results Overall, 134 students answered one or more questions (8.9% response rate). The majority of students (82.0%) were 20–29 years old, studying medicine as a postgraduate degree (77.6%), and had started clinical rotations (62.7%). Students were interested in AI (82.6%), self-reported having a basic understanding of AI (84.8%), but few agreed that they had an understanding of the basic computational principles of AI (33.3%) or the limitations of AI (46.2%). Most students (87.5%) had not received teaching in AI. The majority of students (58.6%) agreed that AI should be part of medical training and most (72.7%) wanted more teaching focusing on AI in medicine. Medical students appeared optimistic regarding the role of AI in medicine, with most (74.4%) agreeing with the statement that AI will improve medicine in general. The majority (56.6%) of medical students were not concerned about the impact of AI on their job security as a doctor. Students selected radiology (72.6%), pathology (58.2%), and medical administration (44.8%) as the specialties most likely to be impacted by AI, and psychiatry (61.2%), palliative care (48.5%), and obstetrics and gynaecology (41.0%) as the specialties least likely to be impacted by AI. Qualitative analysis of free text comments identified the use of AI as a tool, and that doctors will not be replaced as common themes. Conclusion Medical students in WA appear to be interested in AI. However, they have not received education about AI and do not feel they understand its basic computational principles or limitations. AI appears to be a current deficit in the medical curriculum in WA, and most students surveyed were supportive of its introduction. These results are consistent with previous surveys conducted internationally.

Abstract Our aim was to investigate the usefulness of machine learning approaches on linked administrative health data at the population level in predicting older patients’ one-year risk of acute coronary syndrome and death following the use of non-steroidal anti-inflammatory drugs (NSAIDs). Patients from a Western Australian cardiovascular population who were supplied with NSAIDs between 1 Jan 2003 and 31 Dec 2004 were identified from Pharmaceutical Benefits Scheme data. Comorbidities from linked hospital admissions data and medication history were inputs. Admissions for acute coronary syndrome or death within one year from the first supply date were outputs. Machine learning classification methods were used to build models to predict ACS and death. Model performance was measured by the area under the receiver operating characteristic curve (AUC-ROC), sensitivity and specificity. There were 68,889 patients in the NSAIDs cohort with mean age 76 years and 54% were female. 1882 patients were admitted for acute coronary syndrome and 5405 patients died within one year after their first supply of NSAIDs. The multi-layer neural network, gradient boosting machine and support vector machine were applied to build various classification models. The gradient boosting machine achieved the best performance with an average AUC-ROC of 0.72 predicting ACS and 0.84 predicting death. Machine learning models applied to linked administrative data can potentially improve adverse outcome risk prediction. Further investigation of additional data and approaches are required to improve the performance for adverse outcome risk prediction.

Narrow-leafed lupin (Lupinus angustifolius) is an important dryland crop, providing a protein source in global grain markets. While agronomic practices have successfully controlled many dicot weeds among narrow-leafed lupins, the closely related sandplain lupin (Lupinus cosentinii) has proven difficult to control, reducing yield and harvest quality. Here, we successfully trained a segmentation model to detect sandplain lupins and differentiate them from narrow-leafed lupins under field conditions. The deep learning model was trained using 9171 images collected from a field site in the Western Australian grain belt. Images were collected using an unoccupied aerial vehicle at heights of 4, 10, and 20 m. The dataset was supplemented with images sourced from the WeedAI database, which were collected at 1.5 m. The resultant model had an average precision of 0.86, intersection over union of 0.60, and F1 score of 0.70 for segmenting the narrow-leafed and sandplain lupins across the multiple datasets. Images collected at a closer range and showing plants at an early developmental stage had significantly higher precision and recall scores (p-value < 0.05), indicating image collection methods and plant developmental stages play a substantial role in the model performance. Nonetheless, the model identified 80.3% of the sandplain lupins on average, with a low variation (±6.13%) in performance across the 5 datasets. The results presented in this study contribute to the development of precision weed management systems within morphologically similar crops, particularly for sandplain lupin detection, supporting future narrow-leafed lupin grain yield and quality.

Generating textual descriptions of images has been an important topic in computer vision and natural language processing. A number of techniques based on deep learning have been proposed on this topic. These techniques use human-annotated images for training and testing the models. These models require a large number of training data to perform at their full potential. Collecting human generated images with associative captions is expensive and time-consuming. In this paper, we propose an image captioning method that uses both real and synthetic data for training and testing the model. We use a Generative Adversarial Network (GAN) based text to image generator to generate synthetic images. We use an attention-based image captioning method trained on both real and synthetic images to generate the captions. We demonstrate the results of our models using both qualitative and quantitative analysis on popularly used evaluation metrics. We show that our experimental results achieve two fold benefits of our proposed work: i) it demonstrates the effectiveness of image captioning for synthetic images, and ii) it further improves the quality of the generated captions for real images, understandably because we use additional images for training.

Hajj is a mass gathering event that takes place annually in Makkah, Saudi Arabia. Typically, around three million people participate in the event and perform rituals that involve their movements within strict space and time restrictions. Despite efforts by the Hajj organisers, such massive crowd gathering and movement cause overcrowding related problems at the Hajj sites. Several previous simulation studies on Hajj focused on the rituals individually. Tawaf, followed by Sayee, are two important rituals that are performed by all the pilgrims at the same venue on the same day. These events have a strong potential for crowd buildup and related problems. As opposed the previous works in the literature, in this paper we study these two events jointly, rather than separately. We use ExtendSim, a Discrete Event Simulation tool, to integrate the Tawaf and Sayee rituals into one model. The validated model was applied to a wide range of scenarios where different percentages of pilgrims were allocated to the various Tawaf and Sayee areas. The effect of such allocations on the time to complete Tawaf and Sayee indicate strategies for managing these two key Hajj rituals.

Genomic prediction tools support crop breeding based on statistical methods, such as the genomic best linear unbiased prediction (GBLUP). However, these tools are not designed to capture non-linear relationships within multi-dimensional datasets, or deal with high dimension datasets such as imagery collected by unmanned aerial vehicles. Machine learning (ML) algorithms have the potential to surpass the prediction accuracy of current tools used for genotype to phenotype prediction, due to their capacity to autonomously extract data features and represent their relationships at multiple levels of abstraction. This review addresses the challenges of applying statistical and machine learning methods for predicting phenotypic traits based on genetic markers, environment data, and imagery for crop breeding. We present the advantages and disadvantages of explainable model structures, discuss the potential of machine learning models for genotype to phenotype prediction in crop breeding, and the challenges, including the scarcity of high-quality datasets, inconsistent metadata annotation and the requirements of ML models.

Weakly supervised semantic segmentation with image-level labels is of great significance since it alleviates the dependency on dense annotations. However, as it relies on image classification networks that are only capable of producing sparse object localization maps, its performance is far behind that of fully supervised semantic segmentation models. Inspired by the successful use of multi-scale features for an improved performance in a wide range of visual tasks, we propose a Scale-Aware Feature Network (SAFN) for generating object localization maps. The proposed SAFN uses an attention module to learn the relative weights of multi-scale features in a modified fully convolutional network with dilated convolutions. This approach leads to efficient enlargements of the receptive fields of view and produces dense object localization maps. Our approach achieves mIoUs of 62.3% and 66.5% on the PASCAL VOC 2012 test set using VGG16 based and ResNet based segmentation models, respectively, outperforming other state-of-the-art methods for the weakly supervised semantic segmentation task.

Deep learning has demonstrated state-of-the-art performance for a variety of challenging computer vision tasks. On one hand, this has enabled deep visual models to pave the way for a plethora of critical applications like disease prognostics and smart surveillance. On the other, deep learning has also been found vulnerable to adversarial attacks, which calls for new techniques to defend deep models against these attacks. Among the attack algorithms, the black-box schemes are of serious practical concern since they only need publicly available knowledge of the targeted model. We carefully analyze the inherent weakness of deep models in black-box settings where the attacker may develop the attack using a model similar to the targeted model. Based on our analysis, we introduce a novel gradient regularization scheme that encourages the internal representation of a deep model to be orthogonal to another, even if the architectures of the two models are similar. Our unique constraint allows a model to concomitantly endeavour for higher accuracy while maintaining near orthogonal alignment of gradients with respect to a reference model. Detailed empirical study verifies that controlled misalignment of gradients under our orthogonality objective significantly boosts a model's robustness against transferable black-box adversarial attacks. In comparison to regular models, the orthogonal models are significantly more robust to a range of lp norm bounded perturbations. We verify the effectiveness of our technique on a variety of large-scale models.

This paper proposes a novel machine learning approaches to predict the outcome of facial rejuvenation prior to a cosmetic procedure. This is achieved by estimating the required amount of dermal filler volume that needs to be applied on the face by learning the underlying structural mapping from the pretreatment and posttreatment 3D face images. We develop and train our proposed deep neural network, called Rejuv3DNet, designed specifically to predict the dermal filler volume. We also propose the kernel regression (KR)-based model to validate and improve our volume estimation results using regression. Our other contributions include the development of the first 3D face cosmetic dataset, which consists of real-world pretreatment and posttreatment 3D face images and a novel technique for the generation of synthetic cosmetic treatment 3D face images. Our experimental results show that the proposed Rejuv3DNet and the KR model achieve 62.5% and 66.67%, respectively, on real-world data, while these techniques achieve a prediction accuracy of 75.2% and 89.5%, and 77.2% and 90.1% on our two different synthetic datasets. Our proposed techniques have been found to be computationally efficient, achieving near real-time prediction performance. The reported accuracies are our preliminary results for proof of concept, which can be improved with more data. The proposed approach has the potential for further investigation in the cosmetic surgery domain.

Biotic and abiotic plant stress (e.g., frost, fungi, diseases) can significantly impact crop production. It is thus essential to detect such stress at an early stage before visual symptoms and damage become apparent. To this end, this paper proposes a novel deep learning method, called Spectral Convolution and Channel Attention Network (SC-CAN), which exploits the difference in spectral responses of healthy and stressed crops. The proposed SC-CAN method comprises two main modules: (i) a spectral convolution module, which consists of dilated causal convolutional layers stacked in a residual manner to capture the spectral features; (ii) a channel attention module, which consists of a global pooling layer and fully connected layers that compute inter-relationship between feature map channels before scaling them based on their importance level (attention score). Unlike standard convolution, which focuses on learning local features, the dilated convolution layers can learn both local and global features. These layers also have long receptive fields, making them suitable for capturing long dependency patterns in hyperspectral data. However, because not all feature maps produced by the dilated convolutional layers are important, we propose a channel attention module that weights the feature maps according to their importance level. We used SC-CAN to classify salt stress (i.e., abiotic stress) on four datasets (Chinese Spring (CS), Aegilops columnaris (co(CS)), Ae. speltoides auchery (sp(CS)), and Kharchia datasets) and Fusarium head blight disease (i.e., biotic stress) on Fusarium dataset. Reported experimental results show that the proposed method outperforms existing state-of-the-art techniques with an overall accuracy of 83.08%, 88.90%, 82.44%, 82.10%, and 82.78% on CS, co(CS), sp(CS), Kharchia, and Fusarium datasets, respectively.

Gene models are regions of the genome that can be transcribed into RNA and translated to proteins, or belong to a class of non-coding RNA genes. The prediction of gene models is a complex process that can be unreliable, leading to false positive annotations. To help support the calling of confident conserved gene models and minimize false positives arising during gene model prediction we have developed Truegene, a machine learning approach to classify potential low confidence gene models using 14 gene and 41 protein-based characteristics. Amino acid and nucleotide sequence-based features were calculated for conserved (high confidence) and non-conserved (low confidence) annotated genes from the published Pisum sativum Cameor genome. These features were used to train eXtreme Gradient Boost (XGBoost) classifier models to predict whether a gene model is likely to be real. The optimized models demonstrated a prediction accuracy ranging from 87% to 90% and an F-1 score of 0.91–0.94. We used SHapley Additive exPlanations (SHAP) and feature importance plots to identify the features that contribute to the model predictions, and we show that protein and gene-based features can be used to build accurate models for gene prediction that have applications in supporting future gene annotation processes.

Lung cancer is a life-threatening disease and its diagnosis is of great significance. Data scarcity and unavailability of datasets is a major bottleneck in lung cancer research. In this paper, we introduce a dataset of pulmonary lesions for designing the computer-aided diagnosis (CAD) systems. The dataset has fine contour annotations and nine attribute annotations. We define the structure of the dataset in detail, and then discuss the relationship of the attributes and pathology, and the correlation between the nine attributes with the chi-square test. To demonstrate the contribution of our dataset to computer-aided system design, we define four tasks that can be developed using our dataset. Then, we use our dataset to model multi-attribute classification tasks. We discuss the performance in 2D, 2.5D, and 3D input modes of the classification model. To improve performance, we introduce two attention mechanisms and verify the principles of the attention mechanisms through visualization. Experimental results show the relationship between different models and different levels of attributes.

BackgroundChest pain is amongst the most common reason for presentation to the emergency department (ED). There are many causes of chest pain, and it is important for the emergency physician to quickly and accurately diagnose life threatening causes such as acute myocardial infarction (AMI). Multiple clinical decision tools have been developed to assist clinicians in risk stratifying patients with chest. There is growing recognition that machine learning (ML) will have a significant impact on the practice of medicine in the near future and may assist with diagnosis and risk stratification. This systematic review aims to evaluate how ML has been applied to adults presenting to the ED with undifferentiated chest pain and assess if ML models show improved performance when compared to physicians or current risk stratification techniques.Methods and findingsWe conducted a systematic review of journal articles that applied a ML technique to an adult patient presenting to an emergency department with undifferentiated chest pain. Multiple databases were searched from inception through to November 2020. In total, 3361 articles were screened, and 23 articles were included. We did not conduct a metanalysis due to a high level of heterogeneity between studies in both their methods, and reporting. The most common primary outcomes assessed were diagnosis of acute myocardial infarction (AMI) (12 studies), and prognosis of major adverse cardiovascular event (MACE) (6 studies). There were 14 retrospective studies and 5 prospective studies. Four studies reported the development of a machine learning model retrospectively then tested it prospectively. The most common machine learning methods used were artificial neural networks (14 studies), random forest (6 studies), support vector machine (5 studies), and gradient boosting (2 studies). Multiple studies achieved high accuracy in both the diagnosis of AMI in the ED setting, and in predicting mortality and composite outcomes over various timeframes. ML outperformed existing risk stratification scores in all cases, and physicians in three out of four cases. The majority of studies were single centre, retrospective, and without prospective or external validation. There were only 3 studies that were considered low risk of bias and had low applicability concerns. Two studies reported integrating the ML model into clinical practice.ConclusionsResearch on applications of ML for undifferentiated chest pain in the ED has been ongoing for decades. ML has been reported to outperform emergency physicians and current risk stratification tools to diagnose AMI and predict MACE but has rarely been integrated into practice. Many studies assessing the use of ML in undifferentiated chest pain in the ED have a high risk of bias. It is important that future studies make use of recently developed standardised ML reporting guidelines, register their protocols, and share their datasets and code. Future work is required to assess the impact of ML model implementation on clinical decision making, patient orientated outcomes, and patient and physician acceptability.Trial registrationInternational Prospective Register of Systematic Reviews registration number: CRD42020184977.

3-D object recognition is a challenging task for many applications including autonomous robot navigation and scene understanding. Accurate recognition relies on the selection/learning of discriminative features that are in turn used to uniquely characterize the objects. This paper proposes a novel evolutionary feature learning (EFL) technique for 3-D object recognition. The proposed novel automatic feature learning approach can operate directly on 3-D raw data, alleviating the need for data pre-processing, human expertise and/or defining a large set of parameters. EFL offers smart search strategy to learn the best features in a huge feature space to achieve superior recognition performance. The proposed technique has been extensively evaluated for the task of 3-D object recognition on four popular data sets including Washington RGB-D (low resolution 3-D Video), CIN 2D3D, Willow 2D3D and ETH-80 object data set. Reported experimental results and evaluation against existing state-of-the-art methods (e.g., unsupervised dictionary learning and deep networks) show that the proposed EFL consistently achieves superior performance on all these data sets.

Deep convolutional neural networks (DCNNs) have been driving significant advances in semantic image segmentation due to their powerful feature representation for recognition. However, their performance in preserving object boundaries is still not satisfactory. Visual mechanism theory indicates that image segmentation tasks require not only recognition, like DCNNs, but also local visual attention capability. Considering that superpixel is good at grasping detailed local structure, we propose a probabilistic superpixel-based dense conditional random field model (PSP-CRF) to refine label assignments as a postprocessing optimization method. First, the well-known fully convolutional networks (FCN) and Deeplab-ResNet are employed to produce coarse prediction probabilistic maps at each pixel. Second, we construct a fully connected CRF model based on the PSP generated by the simple linear iterative clustering algorithm. In our approach, an effective refining algorithm with entropy is developed to convert the pixel-level appearance and position features to the normalized PSP, which works well for CRF. Third, our method optimizes the PSP-CRF to obtain the final label assignment results by employing a highly efficient mean field inference algorithm and some quadratic programming relaxation related algorithms. The experiments on the PASCAL VOC segmentation dataset demonstrate the effectiveness of our methods which can improve the segmentation performance of DCNNs to 82% in mIoU while increasing the computational efficiency by 47%.

Person re-identification (re-ID), which aims to identify the same individual from a gallery collected with different cameras, has attracted increasing attention in the multimedia retrieval community. Current deep learning methods for person re-ID focus on learning classification models on training identities to obtain an ID-discriminative embedding (IDE) extractor, which is used to extract features from testing images for re-ID. The IDE features of the testing identities might not be discriminative due to that the training identities are different from the testing identities. In this paper, we introduce a new ID-adaptation network (ID-AdaptNet), which aims to improve the discriminative power of the IDE features of the testing identities for better person re-ID. The main idea of the ID-AdaptNet is to transform the IDE features to a common discriminative latent space, where the representations of the “seen”training identities are enforced to adapt to those of the “unseen”training identities. More specifically, the ID-AdaptNet is trained by simultaneously minimizing the classification cross-entropy and the discrepancy between the “seen”and the “unseen”training identities in the hidden space. To calculate the discrepancy, we represent their probability distributions as moment sequences and calculate their distance using their central moments. We further propose a stacking ID-AdaptNet that jointly trains multiple ID-AdaptNets with a regularization method for better re-ID. Experiments show that the ID-AdaptNet and stacking ID-AdaptNet effectively improve the discriminative power of IDE features.

Assessing crop production in the field often requires breeders to wait until the end of the season to collect yield-related measurements, limiting the pace of the breeding cycle. Early prediction of crop performance can reduce this constraint by allowing breeders more time to focus on the highest-performing varieties. Here, we present a multimodal deep learning model for predicting the performance of maize (Zea mays) at an early developmental stage, offering the potential to accelerate crop breeding. We employed multispectral images and eight vegetation indices, collected by an uncrewed aerial vehicle approximately 60 days after sowing, over three consecutive growing cycles (2017, 2018 and 2019). The multimodal deep learning approach was used to integrate field management and genotype information with the multispectral data, providing context to the conditions that the plants experienced during the trial. Model performance was assessed using holdout data, in which the model accurately predicted the yield (RMSE 1.07 t/ha, a relative RMSE of 7.60% of 16 t/ha, and R2 score 0.73) and identified the majority of high-yielding varieties, outperforming previously published models for early yield prediction. The inclusion of vegetation indices was important for model performance, with a normalized difference vegetation index and green with normalized difference vegetation index contributing the most to model performance. The model provides a decision support tool, identifying promising lines early in the field trial.

We present DataDeps.jl: a julia package for the reproducible handling of static datasets to enhance the repeatability of scripts used in the data and computational sciences. It is used to automate the data setup part of running software which accompanies a paper to replicate a result. This step is commonly done manually, which expends time and allows for confusion. This functionality is also useful for other packages which require data to function (e.g. a trained machine learning based model). DataDeps.jl simplifies extending research software by automatically managing the dependencies and makes it easier to run another author’s code, thus enhancing the reproducibility of data science research.

BackgroundThe prediction of readmission or death after a hospital discharge for heart failure (HF) remains a major challenge. Modern healthcare systems, electronic health records, and machine learning (ML) techniques allow us to mine data to select the most significant variables (allowing for reduction in the number of variables) without compromising the performance of models used for prediction of readmission and death. Moreover, ML methods based on transformation of variables may potentially further improve the performance.ObjectiveTo use ML techniques to determine the most relevant and also transform variables for the prediction of 30-day readmission or death in HF patients.MethodsWe identified all Western Australian patients aged 65 years and above admitted for HF between 2003-2008 in linked administrative data. We evaluated variables associated with HF readmission or death using standard statistical and ML based selection techniques. We also tested the new variables produced by transformation of the original variables. We developed multi-layer perceptron prediction models and compared their predictive performance using metrics such as Area Under the receiver operating characteristic Curve (AUC), sensitivity and specificity.ResultsFollowing hospital discharge, the proportion of 30-day readmissions or death was 23.7% in our cohort of 10,757 HF patients. The prediction model developed by us using a smaller set of variables (n = 8) had comparable performance (AUC 0.62) to the traditional model (n = 47, AUC 0.62). Transformation of the original 47 variables further improved (pConclusionsA small set of variables selected using ML matched the performance of the model that used the full set of 47 variables for predicting 30-day readmission or death in HF patients. Model performance can be further significantly improved by transforming the original variables using ML methods.

Screening baggage against potential threats has become one of the prime aviation security concerns all over the world, where manual detection of prohibited items is a time-consuming and hectic process. Many researchers have developed autonomous systems to recognize baggage threats using security X-ray scans. However, all of these frameworks are vulnerable against screening cluttered and concealed contraband items. Furthermore, to the best of our knowledge, no framework possesses the capacity to recognize baggage threats across multiple scanner specifications without an explicit retraining process. To overcome this, we present a novel meta-transfer learning-driven tensor-shot detector that decomposes the candidate scan into dual-energy tensors and employs a meta-one-shot classification backbone to recognize and localize the cluttered baggage threats. In addition, the proposed detection framework can be well-generalized to multiple scanner specifications due to its capacity to generate object proposals from the unified tensor maps rather than diversified raw scans. We have rigorously evaluated the proposed tensor-shot detector on the publicly available SIXray and GDXray datasets (containing a cumulative of 1,067,381 grayscale and colored baggage X-ray scans). On the SIXray dataset, the proposed framework achieved a mean average precision (mAP) of 0.6457, and on the GDXray dataset, it achieved the precision and F1 score of 0.9441 and 0.9598, respectively. Furthermore, it outperforms state-of-the-art frameworks by 8.03% in terms of mAP, 1.49% in terms of precision, and 0.573% in terms of F1 on the SIXray and GDXray dataset, respectively.

In this paper, we propose a high performance Two-Stream spectral-spatial Residual Network (TSRN) for hyperspectral image classification. The first spectral residual network (sRN) stream is used to extract spectral characteristics, and the second spatial residual network (saRN) stream is concurrently used to extract spatial features. The sRN uses 1D convolutional layers to fit the spectral data structure, while the saRN uses 2D convolutional layers to match the hyperspectral spatial data structure. Furthermore, each convolutional layer is preceded by a Batch Normalization (BN) layer that works as a regularizer to speed up the training process and to improve the accuracy. We conducted experiments on three well-known hyperspectral datasets, and we compare our results with five contemporary methods across various sizes of training samples. The experimental results show that the proposed architecture can be trained with small size datasets and outperforms the state-of-the-art methods in terms of the Overall Accuracy, Average Accuracy, Kappa Value, and training time.

Across the globe, remote image data is rapidly being collected for the assessment of benthic communities from shallow to extremely deep waters on continental slopes to the abyssal seas. Exploiting this data is presently limited by the time it takes for experts to identify organisms found in these images. With this limitation in mind, a large effort has been made globally to introduce automation and machine learning algorithms to accelerate both classification and assessment of marine benthic biota. One major issue lies with organisms that move with swell and currents, such as kelps. This paper presents an automatic hierarchical classification method local binary classification as opposed to the conventional flat classification to classify kelps in images collected by autonomous underwater vehicles. The proposed kelp classification approach exploits learned feature representations extracted from deep residual networks. We show that these generic features outperform the traditional off-the-shelf CNN features and the conventional hand-crafted features. Experiments also demonstrate that the hierarchical classification method outperforms the traditional parallel multi-class classifications by a significant margin (90.0% vs. 57.6% and 77.2% vs. 59.0%) on Benthoz15 and Rottnest datasets respectively. Furthermore, we compare different hierarchical classification approaches and experimentally show that the sibling hierarchical training approach outperforms the inclusive hierarchical approach by a significant margin. We also report an application of our proposed method to study the change in kelp cover over time for annually repeated AUV surveys.

Recently, digital wax-up is proposed as a tool to aid prosthodontic planning. However, there are no data about the effect of prosthodontic planning on lateral occlusion scheme. Objective : This study aims to evaluate the impact of conventional and digital prosthodontic planning on lateral occlusion scheme. Material and Methods : Dental models of 10 patients were collected. All models had Angle Class I occlusion and were undergoing prosthodontic treatment that would influence the lateral occlusion scheme. Each set of models had received both conventional wax-up and digital wax-up. In relation to the lateral occlusion scheme, the following variables were evaluated: the prevalence of the different lateral occlusion scheme, number of contacting teeth and percentage of each contacting tooth. Four excursive positions on the working side were included: 0.5, 1.0, 2.0 and 3.0 mm from the maximal intercuspation position. Results : The lateral occlusion scheme of the two wax-up models was subjected to alterations following excursion. There was a tendency for the prevalence of canine-guided occlusion to increase and for the prevalence of group function occlusion to decrease with increasing excursion. The number of contacting teeth was decreasing with the increasing magnitude of excursion. For the 0.5 mm and 1.0 mm positions, the two wax-ups had significantly greater contacts than the pre-treatment models, while at the 2.0 mm and 3.0 mm positions, all the models were similar. For all models, canines were the most commonly contacting teeth, followed by the teeth adjacent to them. No difference was observed between the two wax-ups in relation to the number of contacting teeth. Conclusion : Although the prosthodontic planning had influenced the pattern of the lateral occlusion scheme and contacts, there was no difference between the conventional and digital prosthodontic planning.

In prosthodontics, conventional methods of fabrication of oral and facial prostheses have been considered the gold standard for many years. The development of computer-aided manufacturing and the medical application of this industrial technology have provided an alternative way of fabricating oral and facial prostheses. This narrative review aims to evaluate the different streams of computer-aided manufacturing in prosthodontics. To date, there are two streams: the subtractive and the additive approaches. The differences reside in the processing protocols, materials used, and their respective accuracy. In general, there is a tendency for the subtractive method to provide more homogeneous objects with acceptable accuracy that may be more suitable for the production of intraoral prostheses where high occlusal forces are anticipated. Additive manufacturing methods have the ability to produce large workpieces with significant surface variation and competitive accuracy. Such advantages make them ideal for the fabrication of facial prostheses.