Artificial intelligence automates the characterization of reversibly actuating planar-flow-casted NiTi shape memory alloy foil.

Nickel-Titanium (NiTi) shape memory alloys (SMAs) are smart materials able to recover their original shape under thermal stimulus. Near-net-shape NiTi SMA foils of 2 meters in length and width of 30 mm have been successfully produced by a planar flow casting facility at CSIRO, opening possibilities of wider applications of SMA foils. The study also focuses on establishing a fully automated experimental system for the characterisation of their reversible actuation, significantly improving SMA foils adaptation into real applications. Artificial Intelligence involving Computer Vision and Machine Learning based methods were successfully employed in the development of the automation SMA characterization process. The study finds that an Extreme Gradient Boosting (XGBoost) Regression model based predictive system experimented with over 175,000 video samples could achieve 99% overall prediction accuracy. Generalisation capability of the proposed system makes a significant contribution towards the efficient optimisation of the material design to produce high quality 30 mm SMA foils.

Development of a physiological-based model that uses standard polysomnography and clinical data to predict oral appliance treatment outcomes in obstructive sleep apnea.

STUDY OBJECTIVES: Oral appliance (OA) therapy is a well-tolerated alternative to continuous positive airway pressure. However, it is less efficacious. A major unresolved clinical challenge is the inability to accurately predict who will respond to OA therapy. We recently developed a model to estimate obstructive sleep apnea pathophysiological endotypes. This study aimed to apply this physiological-based model to predict OA treatment responses. METHODS: Sixty-two men and women with obstructive sleep apnea (aged 29-71 years) were studied to investigate the efficacy of a novel OA device. An in-laboratory diagnostic followed by an OA treatment efficacy polysomnography were performed. Seven polysomnography variables from the diagnostic study plus age and body mass index were included in our machine-learning-based model to predict OA therapy response according to standard apnea-hypopnea index (AHI) definitions. Initially, the model was trained on data from the first 45 participants using 10-fold cross-validation. A blinded independent validation was then performed for the remaining 17 participants. RESULTS: Mean accuracy of the trained model to predict OA therapy responders vs nonresponders (AHI < 5 events/h) using 10-fold cross-validation was 91% ± 8%. In the independent blinded validation, 100% (AHI < 5 events/h); 59% (AHI < 10 events/h); 71% (50% reduction in AHI); and 82% (50% reduction in AHI to < 20 events/h) of the 17 participants were correctly classified for each of the treatment outcome definitions respectively. CONCLUSIONS: While further evaluation in larger clinical data sets is required, these findings highlight the potential to use routinely collected sleep study and clinical data with machine learning-based approaches underpinned by obstructive sleep apnea endotype concepts to help predict treatment outcomes to OA therapy for people with obstructive sleep apnea. CITATION: Dutta R, Tong BK, Eckert DJ. Development of a physiological-based model that uses standard polysomnography and clinical data to predict oral appliance treatment outcomes in obstructive sleep apnea. J Clin Sleep Med. 2022;18(3):861-870.

Vision based supervised restricted Boltzmann machine helps to actuate novel shape memory alloy accurately.

Extraordinary shape recovery capabilities of shape memory alloys (SMAs) have made them a crucial building block for the development of next-generation soft robotic systems and associated cognitive robotic controllers. In this study we desired to determine whether combining video data analysis techniques with machine learning techniques could develop a computer vision based predictive system to accurately predict force generated by the movement of a SMA body that is capable of a multi-point actuation performance. We identified that rapid video capture of the bending movements of a SMA body while undergoing external electrical excitements and adapting that characterisation using computer vision approach into a machine learning model, can accurately predict the amount of actuation force generated by the body. This is a fundamental area for achieving a superior control of the actuation of SMA bodies. We demonstrate that a supervised machine learning framework trained with Restricted Boltzmann Machine (RBM) inspired features extracted from 45,000 digital thermal infrared video frames captured during excitement of various SMA shapes, is capable to estimate and predict force and stress with 93% global accuracy with very low false negatives and high level of predictive generalisation.

A Novel Model to Estimate Key Obstructive Sleep Apnea Endotypes from Standard Polysomnography and Clinical Data and Their Contribution to Obstructive Sleep Apnea Severity.

Rationale: There are at least four key pathophysiological endotypes that contribute to obstructive sleep apnea (OSA) pathophysiology. These include 1) upper-airway collapsibility (Pcrit); 2) arousal threshold; 3) loop gain; and 4) pharyngeal muscle responsiveness. However, an easily interpretable model to examine the different ways and the extent to which these OSA endotypes contribute to conventional polysomnography-defined OSA severity (i.e., the apnea-hypopnea index) has not been investigated. In addition, clinically deployable approaches to estimate OSA endotypes to advance knowledge on OSA pathogenesis and targeted therapy at scale are not currently available.Objectives: Develop an interpretable data-driven model to 1) determine the different ways and the extent to which the four key OSA endotypes contribute to polysomnography-defined OSA severity and 2) gain insight into how standard polysomnographic and clinical variables contribute to OSA endotypes and whether they can be used to predict OSA endotypes.Methods: Age, body mass index, and eight polysomnography parameters from a standard diagnostic study were collected. OSA endotypes were also quantified in 52 participants (43 participants with OSA and nine control subjects) using gold-standard physiologic methodology on a separate night. Unsupervised multivariate principal component analyses and data-driven supervised machine learning (decision tree learner) were used to develop a predictive algorithm to address the study objectives.Results: Maximum predictive performance accuracy of the trained model to identify standard polysomnography-defined OSA severity levels (no OSA, mild to moderate, or severe) using the using the four OSA endotypes was approximately twice that of chance. Similarly, performance accuracy to predict OSA endotype categories ("good," "moderate," or "bad") from standard polysomnographic and clinical variables was approximately twice that of chance for Pcrit and slightly lower for arousal threshold.Conclusions: This novel approach provides new insights into the different ways in which OSA endotypes can contribute to polysomnography-defined OSA severity. Although further validation work is required, these findings also highlight the potential for routine sleep study and clinical data to estimate at least two of the key OSA endotypes using data-driven predictive analysis methodology as part of a clinical decision support system to inform scalable research studies to advance OSA pathophysiology and targeted therapy for OSA.

Big data integration shows Australian bush-fire frequency is increasing significantly.

Increasing Australian bush-fire frequencies over the last decade has indicated a major climatic change in coming future. Understanding such climatic change for Australian bush-fire is limited and there is an urgent need of scientific research, which is capable enough to contribute to Australian society. Frequency of bush-fire carries information on spatial, temporal and climatic aspects of bush-fire events and provides contextual information to model various climate data for accurately predicting future bush-fire hot spots. In this study, we develop an ensemble method based on a two-layered machine learning model to establish relationship between fire incidence and climatic data. In a 336 week data trial, we demonstrate that the model provides highly accurate bush-fire incidence hot-spot estimation (91% global accuracy) from the weekly climatic surfaces. Our analysis also indicates that Australian weekly bush-fire frequencies increased by 40% over the last 5 years, particularly during summer months, implicating a serious climatic shift.

Deep cognitive imaging systems enable estimation of continental-scale fire incidence from climate data.

Unplanned fire is a major control on the nature of terrestrial ecosystems and causes substantial losses of life and property. Given the substantial influence of climatic conditions on fire incidence, climate change is expected to substantially change fire regimes in many parts of the world. We wished to determine whether it was possible to develop a deep neural network process for accurately estimating continental fire incidence from publicly available climate data. We show that deep recurrent Elman neural network was the best performed out of ten artificial neural networks (ANN) based cognitive imaging systems for determining the relationship between fire incidence and climate. In a decennium data experiment using this ANN we show that it is possible to develop highly accurate estimations of fire incidence from monthly climatic data surfaces. Our estimations for the continent of Australia had over 90% global accuracy and a very low level of false negatives. The technique is thus appropriate for use in estimating the spatial consequences of climate scenarios on the monthly incidence of wildfire at the landscape scale.

Electronic-nose technology using sputum samples in diagnosis of patients with tuberculosis.

We investigated the potential of two different electronic noses (EN; code named "Rob" and "Walter") to differentiate between sputum headspace samples from tuberculosis (TB) patients and non-TB patients. Only samples from Ziehl-Neelsen stain (ZN)- and Mycobacterium tuberculosis culture-positive (TBPOS) sputum samples and ZN- and culture-negative (TBNEG) samples were used for headspace analysis; with EN Rob, we used 284 samples from TB suspects (56 TBPOS and 228 TBNEG samples), and with EN Walter, we used 323 samples from TB suspects (80 TBPOS and 243 TBNEG samples). The best results were obtained using advanced data extraction and linear discriminant function analysis, resulting in a sensitivity of 68%, a specificity of 69%, and an accuracy of 69% for EN Rob; for EN Walter, the results were 75%, 67%, and 69%, respectively. Further research is still required to improve the sensitivity and specificity by choosing more selective sensors and type of sampling technique.

Intelligent Bayes Classifier (IBC) for ENT infection classification in hospital environment.

Electronic Nose based ENT bacteria identification in hospital environment is a classical and challenging problem of classification. In this paper an electronic nose (e-nose), comprising a hybrid array of 12 tin oxide sensors (SnO2) and 6 conducting polymer sensors has been used to identify three species of bacteria, Escherichia coli (E. coli), Staphylococcus aureus (S. aureus), and Pseudomonas aeruginosa (P. aeruginosa) responsible for ear nose and throat (ENT) infections when collected as swab sample from infected patients and kept in ISO agar solution in the hospital environment. In the next stage a sub-classification technique has been developed for the classification of two different species of S. aureus, namely Methicillin-Resistant S. aureus (MRSA) and Methicillin Susceptible S. aureus (MSSA). An innovative Intelligent Bayes Classifier (IBC) based on "Baye's theorem" and "maximum probability rule" was developed and investigated for these three main groups of ENT bacteria. Along with the IBC three other supervised classifiers (namely, Multilayer Perceptron (MLP), Probabilistic neural network (PNN), and Radial Basis Function Network (RBFN)) were used to classify the three main bacteria classes. A comparative evaluation of the classifiers was conducted for this application. IBC outperformed MLP, PNN and RBFN. The best results suggest that we are able to identify and classify three bacteria main classes with up to 100% accuracy rate using IBC. We have also achieved 100% classification accuracy for the classification of MRSA and MSSA samples with IBC. We can conclude that this study proves that IBC based e-nose can provide very strong and rapid solution for the identification of ENT infections in hospital environment.

Comparison of neural network predictors in the classification of tracheal-bronchial breath sounds by respiratory auscultation.

Despite extensive research in the area of identification and discrimination of tracheal-bronchial breath sounds by computer analysis, the process of identifying auscultated sounds is still subject to high estimation uncertainties. Here we assess the performance of the relatively new constructive probabilistic neural network (CPNN) against the more common classifiers, namely the multilayer perceptron (MLP) and radial basis function network (RBFN), in classifying a broad range of tracheal-bronchial breath sounds. We present our data as signal estimation models of the tracheal-bronchial frequency spectra. We have examined the trained structure of the CPNN with respect to the other architectures and conclude that this architecture offers an attractive means with which to analyse this type of data. This is based partly on the classification accuracies attained by the CPNN, MLP and RBFN which were 97.8, 77.8 and 96.2%, respectively. We concluded that CPNN and RBFN networks are capable of working successfully with this data, with these architectures being acceptable in terms of topological size and computational overhead requirements. We further believe that the CPNN is an attractive classification mechanism for auscultated data analysis due to its optimal data model generation properties and computationally lightweight architecture.

Electronic nose based tea quality standardization.

In this paper we have used a metal oxide sensor (MOS) based electronic nose (EN) to analyze five tea samples with different qualities, namely, drier month, drier month again over-fired, well fermented normal fired in oven, well fermented overfired in oven, and under fermented normal fired in oven. The flavour of tea is determined mainly by its taste and smell, which is generated by hundreds of Volatile Organic Compounds (VOCs) and Non-Volatile Organic Compounds present in tea. These VOCs are present in different ratios and determine the quality of the tea. For example Assamica (Sri Lanka and Assam Tea) and Assamica Sinesis (Darjeeling and Japanese Tea) are two different species of tea giving different flavour notes. Tea flavour is traditionally measured through the use of a combination of conventional analytical instrumentation and human or ganoleptic profiling panels. These methods are expensive in terms of time and labour and also inaccurate because of a lack of either sensitivity or quantitative information. In this paper an investigation has been made to determine the flavours of different tea samples using an EN and to explore the possibility of replacing existing analytical and profiling panel methods. The technique uses an array of 4 MOSs, each of, which has an electrical resistance that has partial sensitivity to the headspace of tea. The signals from the sensor array are then conditioned by suitable interface circuitry. The data were processed using Principal Components Analysis (PCA), Fuzzy C Means algorithm (FCM). We also explored the use of a Self-Organizing Map (SOM) method along with a Radial Basis Function network (RBF) and a Probabilistic Neural Network classifier. Using FCM and SOM feature extraction techniques along with RBF neural network we achieved 100% correct classification for the five different tea samples with different qualities. These results prove that our EN is capable of discriminating between the flavours of teas manufactured under different processing conditions, viz. over-fermented, over-fired, under fermented, etc.

Bacteria classification using Cyranose 320 electronic nose.

BACKGROUND: An electronic nose (e-nose), the Cyrano Sciences' Cyranose 320, comprising an array of thirty-two polymer carbon black composite sensors has been used to identify six species of bacteria responsible for eye infections when present at a range of concentrations in saline solutions. Readings were taken from the headspace of the samples by manually introducing the portable e-nose system into a sterile glass containing a fixed volume of bacteria in suspension. Gathered data were a very complex mixture of different chemical compounds. METHOD: Linear Principal Component Analysis (PCA) method was able to classify four classes of bacteria out of six classes though in reality other two classes were not better evident from PCA analysis and we got 74% classification accuracy from PCA. An innovative data clustering approach was investigated for these bacteria data by combining the 3-dimensional scatter plot, Fuzzy C Means (FCM) and Self Organizing Map (SOM) network. Using these three data clustering algorithms simultaneously better 'classification' of six eye bacteria classes were represented. Then three supervised classifiers, namely Multi Layer Perceptron (MLP), Probabilistic Neural network (PNN) and Radial basis function network (RBF), were used to classify the six bacteria classes. RESULTS: A [6 x 1] SOM network gave 96% accuracy for bacteria classification which was best accuracy. A comparative evaluation of the classifiers was conducted for this application. The best results suggest that we are able to predict six classes of bacteria with up to 98% accuracy with the application of the RBF network. CONCLUSION: This type of bacteria data analysis and feature extraction is very difficult. But we can conclude that this combined use of three nonlinear methods can solve the feature extraction problem with very complex data and enhance the performance of Cyranose 320.