Automated histopathological evaluation of pterygium using artificial intelligence

PurposeThis study aimed to evaluate the efficacy of a new automated method for the evaluation of histopathological images of pterygium using artificial intelligence.MethodsAn in-house software for automated grading of histopathological images was developed. Histopathological images of pterygium (400 images from 40 patients) were analysed using our newly developed software. Manual grading (I–IV), labelled based on an established scoring system, served as the ground truth for training the four-grade classification models. Region of interest segmentation was performed before the classification of grades, which was achieved by the combination of expectation-maximisation and k-nearest neighbours. Fifty-five radiomic features extracted from each image were analysed with feature selection methods to examine the significant features. Five classifiers were evaluated for their ability to predict quantitative grading.ResultsAmong the classifier models applied for automated grading in this study, the bagging tree showed the best performance, with a 75.9% true positive rate (TPR) and 75.8% positive predictive value (PPV) in internal validation. In external validation, the method also demonstrated reproducibility, with an 81.3% TPR and 82.0% PPV for the average of four classification grades.ConclusionsOur newly developed automated method for quantitative grading of histopathological images of pterygium may be a reliable method for quantitative analysis of histopathological evaluation of pterygium.

Frozen Hybrid Kernelised Expectation Maximisation For Bremsstrahlung SPECT Reconstruction

Background: Selective internal radiation therapy with Yttrium-90 microspheres is an effective therapy for liver cancer and liver metastases. Yttrium-90 is mainly a high-energy beta particle emitter. These beta particles emit Bremsstrahlung radiation during their interaction with tissue making post-therapy imaging of the radioactivity distribution feasible. Nevertheless, image quality and quantification is difficult due to the continuous energy spectrum which makes resolution modelling, and attenuation and scatter estimation challenging. Methods: In this study, a modified hybrid kernelised expectation maximisation is used to improve resolution and contrast and reduce noise. The iterative part of the kernel was frozen at the 72nd sub-iteration to avoid over-fitting of noise and background. A NEMA phantom with spherical inserts was used for the optimisation and validation of the algorithm, and data from 5 patients treated with Selective internal radiation therapy were used as proof of clinical relevance of the method. Results: The results suggest a maximum improvement of 56% for region of interest mean recovery coefficient at fixed coefficient of variation and better identification of the hot volumes in the NEMA phantom. Similar improvements were achieved with patient data, showing 47% mean value improvement over the gold standard used in hospitals. Conclusions: Such quantitative improvements could facilitate improved dosimetry calculations with SPECT when treating patients with Selective internal radiation therapy, as well as provide a more visible position of the cancerous lesions in the liver.

Performance profiling of the unit trust funds in Malaysia with data mining techniques

Background: Millennials are exposed to many investment opportunities, and they have shown their interest in gaining more income via investments. One popular investment avenue is unit trusts. However, analysing unit trusts’ financial data and gaining valuable insights may not be as simple because not everyone has the required financial knowledge and adequate time to perform in-depth analytics on the numerous financial data. Furthermore, it is not easy to compile the performance of each unit trust available in Malaysia. The primary objective of this research is to identify unit trust funds that provide higher returns than their average peers via performance profiling. Methods: This research proposed a performance profiling on Malaysia unit trust funds using the two data mining techniques, i.e., Expectation Maximisation (EM) and Apriori, to assist amateur retail investors to choose the right unit trust based on their risk tolerance. EM clustered the unit trust funds in Malaysia into several groups based on their annual financial performances. This was then followed by finding the rules associated with each cluster by applying Apriori. The resulted rules shall serve the purpose of profiling the clustered unit trust funds. Retail investors can then select their preferred unit trust funds based on the performance profile of the clusters. Results: The yearly average total return of the financial year 2018 and 2019 was used to evaluate unit trust funds’ performance in the clusters. The evaluation results indicated that the profiling could provide valuable and insightful information to retail investors with varying risk appetites. Conclusions: This research has demonstrated that the financial performance profiling of unit trust funds could be acquired via data mining approaches. This valuable information is crucial to unit trust investors for selecting suitable funds in investment.

Imputing Biomarker Status from RWE Datasets—A Comparative Study

Missing data is a universal problem in analysing Real-World Evidence (RWE) datasets. In RWE datasets, there is a need to understand which features best correlate with clinical outcomes. In this context, the missing status of several biomarkers may appear as gaps in the dataset that hide meaningful values for analysis. Imputation methods are general strategies that replace missing values with plausible values. Using the Flatiron NSCLC dataset, including more than 35,000 subjects, we compare the imputation performance of six such methods on missing data: predictive mean matching, expectation-maximisation, factorial analysis, random forest, generative adversarial networks and multivariate imputations with tabular networks. We also conduct extensive synthetic data experiments with structural causal models. Statistical learning from incomplete datasets should select an appropriate imputation algorithm accounting for the nature of missingness, the impact of missing data, and the distribution shift induced by the imputation algorithm. For our synthetic data experiments, tabular networks had the best overall performance. Methods using neural networks are promising for complex datasets with non-linearities. However, conventional methods such as predictive mean matching work well for the Flatiron NSCLC biomarker dataset.

Learned Primal Dual Reconstruction for PET

We have adapted, implemented and trained the Learned Primal Dual algorithm suggested by Adler and Öktem and evaluated its performance in reconstructing projection data from our PET scanner. Learned Primal Dual reconstructions are compared to Maximum Likelihood Expectation Maximisation (MLEM) reconstructions. Different strategies for training are also compared. Whenever the noise level of the data to reconstruct is sufficiently represented in the training set, the Learned Primal Dual algorithm performs well on the recovery of the activity concentrations and on noise reduction as compared to MLEM. The algorithm is also shown to be robust against the appearance of artefacts, even when the images that are to be reconstructed present features were not present in the training set. Once trained, the algorithm reconstructs images in few seconds or less.

High-Performance Work Practices And Employee Resilience In New Zealand's Public Services: A Conditional Indirect Effect Model Of Leadership Styles

<p><b>This study considers the relationship between high involvement work practices (HIWP) and employee resilience, as moderated by leadership style. It offers an empirical test of a structural model exploring these relationships in New Zealand’s core public sector.</b></p> <p>The 2016 Workplace Dynamics Survey, undertaken by the New Zealand Public Service Association (PSA) and Victoria University of Wellington, gathered information on the psychological outcomes of workers' job experiences and the organisations for which they worked. All participants were PSA members and were asked questions regarding their jobs, workplaces, and personal lives. The original sample included 14125 participants. For the current study, the sample was subsequently narrowed downed to core public sector members both managerial and non-managerial employees were selected, and items with missing values were imputed using the Expectation-Maximisation logarithm –the imputation resulting in 7326 unique replies for this study.</p> <p>Confirmatory factor analysis (CFA) was used to test the measurement model, and structural equation modelling (SEM) was used to explain the relationship between, on the one hand, the four organizational elements as comprising HIWP –power, information, knowledge, and rewards (PIRK) –and employee resilience, on the other. The hypothesized structural model was then tested, parameters were estimated, and moderators added to see if they could explain variation (heterogeneity) in the effect sizes.</p> <p>According to the model, HIWP positively affects individual employee's self-reported resilience and that this relationship is moderated by and perceived through management style. The model investigates the relationship between a second-order latent variable encompassing the combined effect of the PIRK attributes on a first-order latent variable measuring employee resilience. The model also posits that this relationship is direct and indirect, through two first-order latent variables measuring constructive and laissez-faire leadership styles.</p> <p>This study adds theoretical and practical knowledge by demonstrating that leadership style matters in the relationship between human resource management and the capacity of employees to positively cope, adapt and even thrive in dynamic environments. This research's key finding is that HIWP is positively related to employee resilience and that leadership style mediates that relationship. The results of this study further indicate that, while an individual’s level of education moderates the relationship between HIWP and employee resilience, the employee’s ethnicity and tenure on the job do not. Finally, this study offers proposals for future research, including data collection and recommendations for practitioners.</p>

High-Performance Work Practices And Employee Resilience In New Zealand's Public Services: A Conditional Indirect Effect Model Of Leadership Styles

<p><b>This study considers the relationship between high involvement work practices (HIWP) and employee resilience, as moderated by leadership style. It offers an empirical test of a structural model exploring these relationships in New Zealand’s core public sector.</b></p> <p>The 2016 Workplace Dynamics Survey, undertaken by the New Zealand Public Service Association (PSA) and Victoria University of Wellington, gathered information on the psychological outcomes of workers' job experiences and the organisations for which they worked. All participants were PSA members and were asked questions regarding their jobs, workplaces, and personal lives. The original sample included 14125 participants. For the current study, the sample was subsequently narrowed downed to core public sector members both managerial and non-managerial employees were selected, and items with missing values were imputed using the Expectation-Maximisation logarithm –the imputation resulting in 7326 unique replies for this study.</p> <p>Confirmatory factor analysis (CFA) was used to test the measurement model, and structural equation modelling (SEM) was used to explain the relationship between, on the one hand, the four organizational elements as comprising HIWP –power, information, knowledge, and rewards (PIRK) –and employee resilience, on the other. The hypothesized structural model was then tested, parameters were estimated, and moderators added to see if they could explain variation (heterogeneity) in the effect sizes.</p> <p>According to the model, HIWP positively affects individual employee's self-reported resilience and that this relationship is moderated by and perceived through management style. The model investigates the relationship between a second-order latent variable encompassing the combined effect of the PIRK attributes on a first-order latent variable measuring employee resilience. The model also posits that this relationship is direct and indirect, through two first-order latent variables measuring constructive and laissez-faire leadership styles.</p> <p>This study adds theoretical and practical knowledge by demonstrating that leadership style matters in the relationship between human resource management and the capacity of employees to positively cope, adapt and even thrive in dynamic environments. This research's key finding is that HIWP is positively related to employee resilience and that leadership style mediates that relationship. The results of this study further indicate that, while an individual’s level of education moderates the relationship between HIWP and employee resilience, the employee’s ethnicity and tenure on the job do not. Finally, this study offers proposals for future research, including data collection and recommendations for practitioners.</p>

A Comparative Study of Several EOF Based Imputation Methods for Long Gap Missing Values in a Single-Site Temporal Time Dependent (SSTTD) Air Quality (PM10) Dataset

Missing values are often a major problem in many scientific fields of environmental research, leading to prediction inaccuracy and biased analysis results. This study compares the performance of existing Empirical Orthogonal Functions (EOF) based imputation methods. The EOF mean centred approach (EOF-mean) with several proposed EOF based methods, which include the EOF-median, EOF-trimmean and the newly applied Regularised Expectation-Maximisation Principal Component Analysis based method, namely R-EMPCA in estimating missing values for long gap sequence of missing values problem that exists in a Single Site Temporal Time-Dependent (SSTTD) multivariate structure air quality (PM10) data set. The study was conducted using real PM10 data set from the Klang air quality monitoring station. Performance assessment and evaluation of the methods were conducted via a simulation plan which was carried out according to four percentages (5, 10, 20 and 30) of missing values with respect to several long gap sequences (12, 24, 168 and 720) of missing points (hours). Based on several performance indicators such as RMSE, MAE, R-Square and AI, the results have shown that R-EMPCA outperformed the other methods. The results also conclude that the proposed EOF-median and EOF-trimmean have better performance than the existing EOF-mean based method in which EOF-trimmean is the best among the three. The methodology and findings of this study contribute as a solution to the problem of missing values with long gap sequences for the SSTTD data set.