In this work, nine fruits cultivated in the northern Amazon were studied: abiu, acerola, araçá, bacupari, biribá, caçarí, fruta-do-conde, graviola and taperebá, with the objective of carrying out a bromatological and nutritional study of the pulps of fruits studied. Of all of them, are the pulps of graviola (76.83 ± 0.02 kcal 100 g^-1) bacupari (76.83 ± 0.02 kcal 100 g^-1) and fruta-do-conde (46.66 ± 0.02 kcal 100 g^-1). Among the macronutrients, the high concentration of potassium stands out, especially in the graviola (541.16 ± 0.24 mg 100 g^-1) and the biribá (468.21 ± 0.13 mg 100 g^-1). Among the micronutrients, iron concentrations are representative for araçá pulp (3.04 ± 0.02 mg 100g-1), abiu is rich in zinc (3.71 ± 0.02 mg 100 g^-1) and manganese (6.61 ± 0.11 mg 100 g^-1). The presence of cobalt at the level of traces in some of the pulps studied stands out. The Pearson correlation coefficient was evaluated, as well as the statistical treatment by multivariate analysis PCA to establish the correlation between the variables studied.

Water quality forecasting is increasingly significant for agricultural management and environmental protection. Enormous amounts of water quality data are increasingly being collected by advanced sensors, which leads to an interest in using data-driven models for predicting trends in water quality. However, the unpredictable background noises introduced during water quality monitoring seriously degrade the performance of those models. Meanwhile, artificial neural networks (ANN) with feed-forward architecture lack the capability of maintaining and utilizing the accumulated temporal information, which leads to biased predictions in processing time series data. Hence, we propose a water quality predictive model based on a combination of Kernal Principal Component Analysis (kPCA) and Recurrent Neural Network (RNN) to forecast the trend of dissolved oxygen. Water quality variables are reconstructed based on kPCA method, which aims to reduce the noise from the raw sensory data and preserve actionable information. With the RNN's recurrent connections, our model can make use of the previous information in predicting the trend in the future. Data collected from Burnett River, Australia was applied to evaluate our kPCA-RNN model. The kPCA-RNN model achieved R2 scores up to 0.908, 0.823 and 0.671 for predicting the concentration of dissolved oxygen in the upcoming 1, 2 and 3 hours, respectively. Compared to current data-driven methods like ANN and SVR, the predictive accuracy of the kPCA-RNN model was at least 8 %, 17 % and 21 % better than the comparative models in these 3 cases. The study demonstrates the effectiveness of the kPAC-RNN modeling technique in predicting water quality variables with noisy sensory data.

The industry of the counterfeit goods is one of the largest underground business in the world and it is rapidly growing. Counterfeits can lead not only to loss of profit for honest producers but also have a negative impact on consumers who receive poor quality goods at an excessive price and may be exposed to health damages and safety issues. Perfume industry is constantly exposed to the problem of counterfeits with the fast developing parallel market of inspired perfumes being an important issue. It prompts for the identification of methods that classify the quality of this type of products. In this paper the application of proton nuclear magnetic resonance spectroscopy is employed for the authentication of perfumery products for the first time. Molecular composition of several types of authentic brand fragrances for women were compared with their cheaper inspired equivalents and fake products. Our approach offers the prospect of a fast and simple method for discrimination and counterfeit detection of perfumes using 1H NMR spectroscopy.

A method for classification is introduced in this article, and it is tested on ADNI database to diagnose alzheimer’s disease (AD). It is obvious that tunning the performance of a classification to get better results is a complicated problem, and when we want model’s accuracy or other peformance measurments higher than 90%, the problem will be more complicated. In this study, we tried and succeeded to discover a method to solve this problem. The final feature set can be used clustering too, because outgrowth feature set of the proposed method is invigorated. In the recent years, a lot of activities is done to develop computer aided systems (CAD) for alzheimer’s disease diagnosis. Most of these recently developed systems concenterated on extracting and combining features from MRI, PET, CSF, and …; in this article, we made attempt to do so and utilized one more technique to increase classification performance. Finding and producing the best features to solve three binary classification problems of AD vs. Normal Control (NC), Mild Cognitive Impairment (MCI) vs. NC, and MCI vs. AD are the purposes of this article. Experiments indicate performance and effectiveness rates of the proposed method, which are accuracies of 98.81%, 81.61%, and 81.40% for AD vs. NC, MCI vs. NC, and AD vs. MCI classification problems, respectively. As can be seen, using this method increased the performance of the three binary problems incredibly.

Campylobacter infection affects more than 200,000 people every year in Europe and in the last four years trend shows an increase of campylobacteriosis. Main vehicle for transmission of the bacterium is contaminated food like meat, milk, fruit and vegetables. In this study, the aim was to find characteristic VOCs of C. jejuni in order to detect its presence with an array of MOX gas sensors. Using a starting concentration of 10³ CFU/mL, VOCs were analyzed using GC-MS with SPME technique at the initial time (T0) and after 20 hours (T20). It has been found that Campylobacter sample at T20 is characterized by a higher number of alcohol compounds that the one at T0 and this is due to sugar fermentation. Sensors results showed the ability of the system to follow bacteria curve growth from T0 to T20 using PCA. In particular, this results in a decrease of ΔR/R₀ value over time. For this reason, MOX sensors are a promising technology for the development of a rapid and sensitive system for C. jejuni.

Smoothed particle hydrodynamics (SPH) is regarded as a pure Lagrangian approach, which can solve fluid dynamics problems without the creation of mesh. In this paper, a paralleled SPH solver is developed to solve particle-based computational acoustics (PCA). The aim of this paper is to study the feasibility of using SPH to solve acoustic problems and to improve the efficiency of solving processes by paralleling some procedures on GPU during calculating. A stand SPH code running serially in a CPU is proposed to solve wave equation. This is a wave propagating in a two-dimensional domain. After finishing the computation, the results are compared with the theoretical solutions and they agree well. So its feasibility is verified. There are two main methods for searching neighbor particles: all-pair search method and linked-list search method. Both methods are used in different codes to simulate an identical problem and their runtimes are compared to investigate their searching efficiencies. The runtime results show that linked-list search method has a higher efficiency, which can save a lot of searching time when simulating problems with huge amounts of particles. Furthermore, the percentages of different procedures’ runtimes in a simulation are also discussed to find the most consuming one. Then, some codes are modified to run in different GPUs and their runtimes are compared with those of serial ones on a CPU. Runtime results show that the paralleled algorithm can be more than 80 times faster than the serial one. The result shows that GPU paralleled SPH computing can achieve desirable accuracy and speed in solving acoustic problems.

The concentration of total suspended particles (TSP) and nanoparticles (PM0.1) over Hat Yai city, Songkhla province, southern Thailand was measured in 2019. Organic carbon (OC) and elemental carbon (EC) were evaluated by carbon aerosol analyzer (IMPROVE-TOR) method. Thirteen trace elements including Al, Ba, K, Cu, Cr, Fe, Mg, Mn, Na, Ni, Ti, Pb, and Zn were evaluated by ICP-OES. Annual average TSP and PM0.1 mass concentrations were determined to be 58.3 ± 7.8 and 10.4 ± 1.2 µg/m3, respectively. The highest levels of PM occurred in the wet season with the corresponding values for the dry seasons being lower. The annual average OC/EC ratio ranged from 3.8 - 4.2 (TSP) and 2.5 - 2.7 (PM0.1). The char to soot ratios were constantly less than 1.0 for both TSP and PM0.1, indicating that land transportation is the main emission source. A principal component analysis (PCA) revealed that road transportation, industry, and biomass burning are the key sources of these particles. However, PM arising from Indonesian peatland fires causes an increase in the carbon and trace element concentrations in southern Thailand. The findings make useful information for air quality management and strategies for controlling this problem, based on a source apportionment analysis.

Pulsed thermography is a commonly used non-destructive testing method, and is increasingly studied for advanced materials such as carbon fiber-reinforced polymer (CFRP) evaluation. Different processing approaches are proposed to detect and characterize anomalies that may be generated in structures during the manufacturing cycle or service period. In this study, we used a type of matrix decomposition using Robust-PCA via Inexact-ALM in our experiment. We investigate this method as a pre-and post-processing method on thermal data acquired by pulsed thermography. We employed state-of-the-art methods, i.e., PCT, PPT, and PLST, as the main process. The results indicate that pre-processing on thermal data can elevate the defect detectability while post-processing, in some cases, can deteriorate the results.

A regularization approach to model selection, within a generalized HJM framework, is introduced which learns the closest arbitrage-free model to a prespecified factor model. This optimization problem is represented as the limit of a one-parameter family of computationally tractable penalized model selection tasks. General theoretical results are derived and then specialized to affine term-structure models where new types of arbitrage-free machine learning models for the forward-rate curve are estimated numerically and compared to classical short-rate and the dynamic Nelson-Siegel factor models.

Fiber optic sensors cannot measure damage; for getting information about damage from strain measurements, additional strategies are needed, and several alternatives have been proposed. This paper discuss two independent concepts: the first one is based on detecting the new strains appearing around a damage spot; the structure does not need to be under loads; the technique is very robust, damage detectability is high, but it requires sensors to be located very close to the damage, so it is a local technique. The second approach offers a wider coverage of the structure, it is based on identifying the changes caused by the damage on the strains field in the whole structure for similar external loads. Damage location does not need to be known a priori, detectability is dependent upon the sensors network density, damage size and the external loads. Examples of application to real structures are given.

Climate change is a phenomenon that makes the climate system of a given region to be more unpredictable and increases the risk of water-related problems. GCMs under the new CMIP6 framework holds several climate models with many improvements as compared to past similar efforts. The improvements are mainly in the number of scenarios formulated, setup, parametrization, and resolution. In this study, 10 downscaled climate models from CMIP6 are evaluated by applying statistical and data mining tools and are ranked based on their capability to describe the historical observed series. The result of the analysis showed that the outputs of the MPI-ESM1-2-HR model have a good overall ranking among those 10 models. The output of this top-ranked model is used to understand future climate over UASB after properly bias-corrected using the QM method. Results of the bias correction step show that average annual precipitation has shown an increment of 6.5\% in the middle (SSP2-4.5) and 10.3\% in the worst (SSP5-8.5) case scenarios for the mid-century (2040 - 2069). Similarly, for the end of the century (2070 - 2099) an increment of 4.7\% and 17.5\% was predicted for the two scenarios respectively. Whereas average annual maximum temperature series showed an increment of 1.5 $\degree C$ for middle and 2.6 $\degree C$ for the worst case in the mid-century. At the same time, an increment of 2.2 $\degree C$ and 3.5 $\degree C$ were predicted for the end of the century similarly for those scenarios. Furthermore, it was predicted that the average annual minimum temperature series will have an increment of 2.6 $\degree C$ and 3.1 $\degree C$ for mid-century and 3.1 $\degree C$ and 4.7 $\degree C$ for the end century for the two scenarios respectively. An increase in precipitation with increased land degradation problems in the sub-basin increases the risk of flood events in the future.

The visualization of hyperspectral images in display devices, having RGB colour composition channels is quite difficult due to the high dimensionality of these images. Thus, principal component analysis has been used as a dimensionality reduction algorithm to reduce information loss, by creating uncorrelated features. To classify regions in the hyperspectral images, K-means clustering has been used to form clusters/regions. These two algorithms have been implemented on the three datasets imaged by AVIRIS and ROSIS sensors.

Hyperspectral imaging provides vital information about the objects and elements present inside the image. That’s why they are very useful in satellite imagery as well as image forensics. Hyperspectral document analysis (HSDI) can be used for document authentication using ink analysis which can provide sufficient information about the composition and type of ink. In this project, we have implemented HSDI based ink classification technique using Principle Component Analysis for dimensionality reduction and K-means clustering for ink classification. This is unsupervised learning approach and it is very simple and efficient in order to classify limited number of bands. We have used this technique to classify 33 different bands of ink.

Floods and droughts are driven, in part, by spatial patterns of extreme rainfall. Heat waves are driven by spatial patterns of extreme temperature. It is therefore of interest to design statistical methodologies that allow the identification of likely patterns of extreme rain or temperature from observed historical data. The standard work-horse for identifying patterns of climate variability in historical data is Principal Component Analysis (PCA) and its variants. But PCA optimizes for variance not spatial extremes, and so there is no particular reason why the first PCA spatial pattern should identify, or even approximate, the types of patterns that may drive these phenomena, even if the linear assumptions underlying PCA are correct. We present an alternative pattern identification algorithm that makes the same linear assumptions as PCA, but which can be used to explicitly optimize for spatial extremes. We call the method Directional Component Analysis (DCA), since it involves introducing a preferred direction, or metric, such as `sum of all points in the spatial field'. We compare the first PCA and DCA spatial patterns for US rainfall anomalies on a 6 month timescale, using the sum metric for the definition of DCA in order to focus on total rainfall anomaly over the domain, and find that they are somewhat different. The definitions of PCA and DCA result in the first PCA spatial pattern having the larger explained variance of the two patterns, while the first DCA spatial pattern, when scaled appropriately, has a higher likelihood and greater total rainfall anomaly, and indeed is the pattern with the highest total rainfall anomaly for any given likelihood. In combination these two patterns yield more insight into rainfall variability and extremes than either pattern on its own.

Dayaoshan, as an important metal ore producing area in China, is faced with the dilemma of resource depletion due to long-term exploitation. In this paper, remote sensing method is used to circle the favorable metallogenic areas and find new ore points for Gulong. Firstly, vegetation interference bas been removed by using mixed pixel decomposition method with hyperplane and genetic algorithm (GA) optimization; then, altered mineral distribution information has been extracted based on principal component analysis (PCA) and support vector machine (SVM) method; Thirdly, the favorable areas of gold mining in Gulong has been delineated by using ant colony algorithm (ACA) optimization SVM model to remove false altered minerals; Lastly, field survey verified that the extracted alteration mineralization information is correct and effective. The results show that the mineral alteration extraction method proposed in this paper has certain guiding significance for metallogenic prediction by remote sensing.

Associating the most influential parameters with the product distribution is of uttermost importance in complex catalytic processes such as fluid catalytic cracking (FCC). These correlations can lead to the information-driven catalyst screening, kinetic modeling and reactor design. In this work, a dataset of 104 uncorrelated experiments, with 64 variables, has been obtained in an FCC simulator using 6 types of feedstock (vacuum gasoil, polyethylene pyrolysis waxes, scrap tire pyrolysis oil, dissolved polyethylene and blends of the previous), 36 possible sets of conditions (varying contact time, temperature and catalyst/oil ratio) and 3 industrial catalysts. Principal component analysis (PCA) has been applied over the dataset, showing that the main components are associated with feed composition (27.41% variance); operational conditions (19.09%) and catalyst properties (12.72%). The variables of each component have been correlated with the indexes and yields of the products: conversion, octane number, aromatics, olefins (propylene) or coke, among others.

Human footprint is having a unique set of ridges unmatched by any other human being, and therefore it can be used in different identity documents for example birth certificate, Indian biometric identification system AADHAR card, driving license, PAN card, and passport. There are many instances of the crime scene where an accused must walk around and left the footwear impressions as well as barefoot prints and therefore it is very crucial to recovering the footprints to identify the criminals. Footprint-based biometric is a considerably newer technique for personal identification. Fingerprints, retina, iris and face recognition are the methods most useful for attendance record of the person. This time world is facing the problem of global terrorism. It is challenging to identify the terrorist because they are living as regular as the citizens do. Their soft target includes the industries of special interests such as defense, silicon and nanotechnology chip manufacturing units, pharmacy sectors. They pretend themselves as religious persons, so temples and other holy places, even in markets is in their targets. These are the places where one can obtain their footprints easily. The gait itself is sufficient to predict the behaviour of the suspects. The present research is driven to identify the usefulness of footprint and gait as an alternative to personal identification.

Traditional Filuferru is an ancient spirit from Sardinia, Italy, usually obtained from the distillation of wine or grape marc. In this contribution, the results of the first chemical characterization of a wide number of craft Filuferru samples has been accomplished in terms of evaluation of the alcoholic strength, qualitative and quantitative GC-MS analysis of the volatile composition of the distillate, and its trace element composition by means a ICP-MS method. Both instrumental methods have been validated and applied on 21 craft samples of Filuferru, whereas one sample of commercial distillate has been analyzed for comparison purposes. Alcoholic strength ranged between 41.0 and 62.4% (v/v). Sixty volatile compounds were identified and ten of them have been quantified. Analogies and differences with Grappa (i.e. the Italian distilled spirit most close to Filuferru) have been highlighted in the qualitative and quantitative profile of this matrix. Often meaningful amounts of acetaldehyde, ethyl acetate, dietyl acetal and acetic acid were measured. Elemental analysis, performed on toxic, non-toxic elements and oligoelements, 18 in total, revealed a wide variability of concentrations in both analytes and samples. High concentrations of Cu are sometimes evidenced, likely caused by losses from the distillation apparatus. The principal components analysis (PCA) allowed the differentiation of the ten volatile compounds quantified in two groups: the former, described mainly by PC1, constituted by acetic acid, ethyl acetate, dietyl acetal and acetaldehyde, and the second, described by PC2, constituted by 1-propanol, 2-methyl-1-propanol, the two coeluiting isomers 2-methyl-1-butanol and 3-methyl-1-butanol,1-hexanol, 2-phenylethanol and 2,3-butanediol. Data obtained may be useful in order to establish a regulation for the production of high-quality traditional Filuferru from Sardinia.

Malware complexity is rapidly increasing, causing catastrophic impacts on computer systems. Memory dump malware is gaining increased attention due to its ability to expose plaintext passwords or key encryption files. This paper presents an enhanced classification model based on One class SVM (OCSVM) classifier that can identify any deviation from the normal memory dump file patterns and detect it as malware. The proposed model integrates OCSVM and Principal Component Analysis (PCA) for increased model sensitivity and efficiency. An up-to-date dataset known as “MALMEMANALYSIS-2022” was utilized during the evaluation phase of this study. The accuracy achieved by the traditional one-class classification (TOCC) model was 55%, compared to 99.4% in the one-class classification with PCA (OCC-PCA) model. Such results have confirmed the increased performance achieved by the proposed model.

China's implementation of public-private partnerships projects has been quite effective, involving infrastructure and other livelihood projects, a total of 19 industries, and an investment of nearly 1.5 trillion yuan. The characteristics of PPP projects such as long construction period and large investment amount determine the risks of PPP projects are also great, and the PPP projects of comprehensive environmental governance are also the same. The government and social capital use the PPP model to cooperate, and use the principal component analysis method to assess the risks of the entire life cycle of the comprehensive environmental governance PPP project. Therefore, it plays an important role in ensuring the smooth implementation of projects and reducing the losses caused by risks. According to the risk factors of the whole life cycle of the comprehensive environmental governance PPP project, an indicator system of 5 first-level indicators, 18 second-level indicators, and 43 third-level indicators has been established. Principal component analysis is used to analyze the influence weight of risk factors at each stage. The analysis shows that among the four stages, environmental pollution risk, project approval delay risk, completion risk, interest rate and financial fluctuation risk, and franchise life risk are the most influential risks in the implementation of PPP projects. Therefore, suggestions are made through the risk factors of each stage in the comprehensive environmental governance PPP project. For example, strengthen the response to the external environment risks of the comprehensive environmental governance PPP project, standardize the bidding and procurement of the comprehensive environmental governance PPP project, and strengthen the subsequent management of the transfer of the comprehensive environmental governance PPP project. In this way, the ability to resist risks of comprehensive environmental governance PPP projects is improved; the smooth implementation of the project is guaranteed, and the long-term development of comprehensive environmental governance PPP projects is promoted.

The COVID-19 pandemic situation, disease intensity, weak healthcare facilities, unawareness, and misinformation led people to fear and anxiety in Bangladesh. This study intended to get peoples’ perception on psychosocial, socio-economic and environmental crisis amidst the pandemi. An online questionnaire was surveyed nationwide (respondents no.1066). Datasets were analyzed through the Principal Component Analysis (PCA), hierarchical Cluster Analysis (CA), Pearson’s correlation matrix (PCM), Linear regression analysis (LRA), and psychometric characteristics were included in the Classical Test Theory (CTT) analysis. There were good associations among the psychosocial, socio-economic and environmental parameters. A significant association between fear of COVID-19 with struggling healthcare system (p<0.05) was found. Also, negative association between fragile health system and government’s ability to deal with the pandemic (p<0.05) revealing poor governance. Again, a positive association of shutdown and social distancing with fear of losing life, and due to lack of health treatment (p<0.05) reveals that shut down hampers normal activities which may lead to mental and economic stress. However, a positive association of socio-economic impact of the shutdown with poor people’s suffering, the price hike of basic need, hamper of formal education (p<0.05) may lead to severe socio-economic and health crisis. There is a possibility of climate-induced disaster during/after the pandemic, which will create severe food insecurity (p<0.01). Daily wage earners and poors will suffer most by food and nutritional deficiency, and the country may face huge economic burden. Proper risk assessment and communications is needed to alleviate fear and anxiety. Thus, financial support and mental boosting is required.

Study of fungal antibiotics in their competitive interactions with arthropods may lead to development novel biorational insecticides. Extracts of Alternaria tenuissima MFP253011 obtained by various methods showed a wide range of biological activity, including entomotoxic properties. Analysis of their composition and bioactivity allowed to reveal several known mycotoxins and unidentified compounds that may be involved in entomotoxic activity of the extracts. Among them, tenuazonic acid (TeA), which was the major component of the A. tenuissima extracts, was found the most likely to have larvicidal activity against Galleria mellonella. In the intrahaemocoel injection bioassay, TeA was toxic to G. mellonella and of Zophobas morio with LT50 6 and 2 days, respectively, at the level of 50 µg/larva. Administered orally, TeA inhibited growth of G. mellonella larvae and caused mortality of Acheta domesticus imagines (LT50 7 days) at a concentration of 250 µg/g of feed. TeA showed weak contact-intestinal activity against the two phytophages, Tetranychus urticae and Schizaphis graminum, causing the 12 and 40% of mortality at a concentration of 1 mg/mL. TeA was cytotoxic to Sf9 cell line (IC50 25 µg/mL). Thus, model insect G. mellonella and cell line Sf9 could be used for a further toxicological characterization of TeA.

Lung cancer is a deadly disease if not diagnosed in its early stages. However, early detection of lung cancer is a challenging task due to the shape and size of its nodules. Radiologists need support from automated tools for precise opinion. Automated detection of the affected lungs nodule is difficult because of the shape similarity among healthy tissues. Over the years, several expert systems have been developed that help radiologists to diagnose lung cancer. In this article, we propose a framework to precisely detect lungs cancer by classifying it between benign and malignant nodules. The framework is tested using the subset of the publicly available at the Lung Image Database Consortium image collection (LIDC-IDRI). Multiple techniques including filtering and noise removing are applied for pre-processing. Subsequently, the OTSU and the semantic segmentation are used to accurately detect the unhealthy lungs nodules. In total, 13 nodules features were extracted using Principal Components Analysis (PCA) algorithm. Four optimal features are selected based on the classification performance. In the classification phase, 9 different classifiers are used along with two types of validation schemes i.e. train test holdout validation with 70-30 data split and 10 fold cross-validation. Our experiments show that the proposed system provides 99.23\% accuracy using logic boost classifier.

Deep learning classification is the state-of-the-art of machine learning approach. Earlier work proves that the deep convolutional neural network has successfully and brilliantly in different applications such as images or video data. Recognizing and clarifying the remote sensing aspect of the earth's surface and exploit land cover and land use (LCLU). First, this article summarized the remote sensing emerging application and challenges for deep learning methods. Second, we propose four approaches to learn efficient and effective CNNs to transfer image representation on the ImageNet dataset to recognize LCLU datasets. We use VGG16, Inception-ResNet-V2, Inception-V3, and DenseNet201 models to extract features from the EACC dataset. We use pre-trained CNNs on ImageNet to extract features. For feature selection we proposed principal component analysis (PCA) to improve accuracy and speed up the model. We train our model by multi-layer perceptron (MLP) as a classifier. Lastly, we apply the multi-granularity encoding ensemble model. We achieve an overall accuracy of 92.3% for the nine-class classification problem. This work will help remote sensing scientists understand deep learning tools and apply them in large-scale remote sensing challenges

Exercise-based cardiac rehabilitation requires patients to perform a set of certain prescribed exercises a specific number of times. Local muscular endurance (LME) exercises are an important part of the rehabilitation program. Automatic exercise recognition and repetition counting, from wearable sensor data is an important technology to enable patients to perform exercises independently in remote settings, e.g. their own home. In this paper we first report on a comparison of traditional approaches to exercise recognition and repetition counting, corresponding to supervised machine learning and peak detection from inertial sensing signals respectively, with more recent machine learning approaches, specifically Convolutional Neural Networks (CNNs). We investigated two different types of CNN: one using the AlexNet architecture, the other using time-series array. We found that the performance of CNN based approaches were better than the traditional approaches. For exercise recognition task, we found that the AlexNet based single CNN model outperformed other methods with an overall 97.18% F1-score measure. For exercise repetition counting , again the AlexNet architecture based single CNN model outperformed other methods by correctly counting repetitions in 90% of the performed exercise sets within an error of ±1. To the best of our knowledge, our approach of using a single CNN method for both recognition and repetition counting is novel. In addition to reporting our findings, we also make the dataset we created, the INSIGHT-LME dataset, publicly available to encourage further research.

A method for generating fluoroscopic (time-varying) volumetric images using patient-specific motion models derived from 4-dimensional cone-beam CT (4D-CBCT) images is developed. 4D-CBCT images acquired immediately prior to treatment have the potential to accurately represent patient anatomy and respiration during treatment. Fluoroscopic 3D image estimation is done in two steps: 1) deriving motion models and 2) optimization. To derive motion models, every phase in a 4D-CBCT set is registered to a reference phase chosen from the same set using deformable image registration (DIR). Principal components analysis (PCA) is used to reduce the dimensionality of the displacement vector fields (DVFs) resulting from DIR into a few vectors representing organ motion found in the DVFs. The PCA motion models are optimized iteratively by comparing a cone-beam CT (CBCT) projection to a simulated projection computed from both the motion model and a reference 4D-CBCT phase, resulting in a sequence of fluoroscopic 3D images. Patient datasets were used to evaluate the method by estimating the tumor location in the generated images compared to manually defined ground truth positions. Experimental results showed that the average tumor mean absolute error (MAE) along the superior-inferior (SI) direction and the 95th percentile in two patient datasets were (2.29 mm and 5.79 mm) for patient 1 and (1.89 mm and 4.82 mm) for patient 2. This study has demonstrated the feasibility of deriving 4D-CBCT-based PCA motion models that have the potential to account for the 3D non-rigid patient motion and localize tumors and other patient anatomical structures on the day of treatment.