Potential diagnostic application of a novel deep learning- based approach for COVID-19.

COVID-19 is a highly communicable respiratory illness caused by the novel coronavirus SARS-CoV-2, which has had a significant impact on global public health and the economy. Detecting COVID-19 patients during a pandemic with limited medical facilities can be challenging, resulting in errors and further complications. Therefore, this study aims to develop deep learning models to facilitate automated diagnosis of COVID-19 from CT scan records of patients. The study also introduced COVID-MAH-CT, a new dataset that contains 4442 CT scan images from 133 COVID-19 patients, as well as 133 CT scan 3D volumes. We proposed and evaluated six different transfer learning models for slide-level analysis that are responsible for detecting COVID-19 in multi-slice spiral CT. Additionally, multi-head attention squeeze and excitation residual (MASERes) neural network, a novel 3D deep model was developed for patient-level analysis, which analyzes all the CT slides of a given patient as a whole and can accurately diagnose COVID-19. The codes and dataset developed in this study are available at https://github.com/alrzsdgh/COVID . The proposed transfer learning models for slide-level analysis were able to detect COVID-19 CT slides with an accuracy of more than 99%, while MASERes was able to detect COVID-19 patients from 3D CT volumes with an accuracy of 100%. These achievements demonstrate that the proposed models in this study can be useful for automatically detecting COVID-19 in both slide-level and patient-level from patients' CT scan records, and can be applied for real-world utilization, particularly in diagnosing COVID-19 cases in areas with limited medical facilities.

Road networks and socio-demographic factors to explore COVID-19 infection during its different waves.

The COVID-19 pandemic triggered an unprecedented level of restrictive measures globally. Most countries resorted to lockdowns at some point to buy the much-needed time for flattening the curve and scaling up vaccination and treatment capacity. Although lockdowns, social distancing and business closures generally slowed the case growth, there is a growing concern about these restrictions' social, economic and psychological impact, especially on the disadvantaged and poorer segments of society. While we are all in this together, these segments often take the heavier toll of the pandemic and face harsher restrictions or get blamed for community transmission. This study proposes a road-network-based networked approach to model mobility patterns between localities during lockdown stages. It utilises a panel regression method to analyse the effects of mobility in transmitting COVID-19 in an Australian context, together with a close look at a suburban population's characteristics like their age, income and education. Firstly, we attempt to model how the local road networks between the neighbouring suburbs (i.e., neighbourhood measure) and current infection count affect the case growth and how they differ between delta and omicron variants. We use a geographic information system, population and infection data to measure road connections, mobility and transmission probability across the suburbs. We then looked at three socio-demographic variables: age, education and income and explored how they moderate independent and dependent variables (infection rates and neighbourhood measures). The result shows strong model performance to predict infection rate based on neighbourhood road connection. However, apart from age in the delta variant context, the other variables (income and education level) do not seem to moderate the relationship between infection rate and neighbourhood measure. The results indicate that suburbs with a more socio-economically disadvantaged population do not necessarily contribute to more community transmission. The study findings could be potentially helpful for stakeholders in tailoring any health decision for future pandemics.

Autonomous assessment of spontaneous retinal venous pulsations in fundus videos using a deep learning framework.

The presence or absence of spontaneous retinal venous pulsations (SVP) provides clinically significant insight into the hemodynamic status of the optic nerve head. Reduced SVP amplitudes have been linked to increased intracranial pressure and glaucoma progression. Currently, monitoring for the presence or absence of SVPs is performed subjectively and is highly dependent on trained clinicians. In this study, we developed a novel end-to-end deep model, called U3D-Net, to objectively classify SVPs as present or absent based on retinal fundus videos. The U3D-Net architecture consists of two distinct modules: an optic disc localizer and a classifier. First, a fast attention recurrent residual U-Net model is applied as the optic disc localizer. Then, the localized optic discs are passed on to a deep convolutional network for SVP classification. We trained and tested various time-series classifiers including 3D Inception, 3D Dense-ResNet, 3D ResNet, Long-term Recurrent Convolutional Network, and ConvLSTM. The optic disc localizer achieved a dice score of 95% for locating the optic disc in 30 milliseconds. Amongst the different tested models, the 3D Inception model achieved an accuracy, sensitivity, and F1-Score of 84 ± 5%, 90 ± 8%, and 81 ± 6% respectively, outperforming the other tested models in classifying SVPs. To the best of our knowledge, this research is the first study that utilizes a deep neural network for an autonomous and objective classification of SVPs using retinal fundus videos.

Ensemble Learning for Disease Prediction: A Review.

Machine learning models are used to create and enhance various disease prediction frameworks. Ensemble learning is a machine learning technique that combines multiple classifiers to improve performance by making more accurate predictions than a single classifier. Although numerous studies have employed ensemble approaches for disease prediction, there is a lack of thorough assessment of commonly used ensemble approaches against highly researched diseases. Consequently, this study aims to identify significant trends in the performance accuracies of ensemble techniques (i.e., bagging, boosting, stacking, and voting) against five hugely researched diseases (i.e., diabetes, skin disease, kidney disease, liver disease, and heart conditions). Using a well-defined search strategy, we first identified 45 articles from the current literature that applied two or more of the four ensemble approaches to any of these five diseases and were published in 2016-2023. Although stacking has been used the fewest number of times (23) compared with bagging (41) and boosting (37), it showed the most accurate performance the most times (19 out of 23). The voting approach is the second-best ensemble approach, as revealed in this review. Stacking always revealed the most accurate performance in the reviewed articles for skin disease and diabetes. Bagging demonstrated the best performance for kidney disease (five out of six times) and boosting for liver and diabetes (four out of six times). The results show that stacking has demonstrated greater accuracy in disease prediction than the other three candidate algorithms. Our study also demonstrates variability in the perceived performance of different ensemble approaches against frequently used disease datasets. The findings of this work will assist researchers in better understanding current trends and hotspots in disease prediction models that employ ensemble learning, as well as in determining a more suitable ensemble model for predictive disease analytics. This article also discusses variability in the perceived performance of different ensemble approaches against frequently used disease datasets.

Source Code for Optimized Parallel Inception: A Fast COVID-19 Screening Software.

COVID-19 and swine-origin influenza A (H1N1) are both pandemics that sparked significant concern worldwide. These two viruses have the same symptoms and occur at a collision timeline. Optimized Parallel Inception (OPI) presents a new strategy to screen the COVID-19 from H1N1 with use of only symptoms. In this paper, the process of preprocessing, screening, and specifying feature importance by OPI and particle swarm optimization is presented. Experimental results indicate 98.88 accuracy for screening COVID-19, H1N1, and Neither COVID-19 Nor H1N1.

Fast COVID-19 versus H1N1 screening using Optimized Parallel Inception.

COVID-19 and swine-origin influenza A (H1N1) are both pandemics that sparked significant concern worldwide. Since these two diseases have common symptoms, a fast COVID-19 versus H1N1 screening helps better manage patients at healthcare facilities. We present a novel deep model, called Optimized Parallel Inception, for fast screening of COVID-19 and H1N1 patients. We also present a Semi-supervised Generative Adversarial Network (SGAN) to address the problem related to the smaller size of the COVID-19 and H1N1 research data. To evaluate the proposed models, we have merged two separate COVID-19 and H1N1 data from different sources to build a new dataset. The created dataset includes 4,383 positive COVID-19 cases, 989 positive H1N1 cases, and 1,059 negative cases. We applied SGAN on this dataset to remove issues related to unequal class densities. The experimental results show that the proposed model's screening accuracy is 99.2% and 99.6% for COVID-19 and H1N1, respectively. According to our analysis, the most significant symptoms and underlying chronic diseases for COVID-19 versus H1N1 screening are dry cough, breathing problems, diabetes, and gastrointestinal.

EISI: Extended inter-spike interval for mental health patients clustering based on mental health services and medications utilisation.

Mental health is vital in all human life stages, and managing mental healthcare service resources is crucial for providers. This paper presents a new method, called Extended Inter-Spike Interval (EISI), on identifying the patients with a similar utilisation of mental health services and medications. The EISI measures the distance between the utilisation patterns of the patients. Then, the pairwise distances are given to a developed split-and-merge Partitioning Around Medoids (PAM) clustering algorithm to identify the patients with similar utilisation patterns. To evaluate the proposed method, we use two years (2013-2014) of the 10% publicly available sample of the Australian Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Scheme (PBS) administrative data. Results show that mental health patients can be grouped into ten clusters with distinct and interpretable utilisations patterns. The largest cluster comprises individuals who only visit general practitioners and take psycholeptics medications for a short time. The smallest group contains occasional visits with general practitioners and regularly utilises psycholeptics and psychoanaleptics medications over long periods. The proposed method provides insights on whom to target and how to structure services for different groups of individuals with mental health conditions.

SSP: Early prediction of sepsis using fully connected LSTM-CNN model.

BACKGROUND: Sepsis is a life-threatening condition that occurs due to the body's reaction to infections, and it is a leading cause of morbidity and mortality in hospitals. Early prediction of sepsis onset facilitates early interventions that promote the survival of suspected patients. However, reliable and intelligent systems for predicting sepsis are scarce. METHODS: This paper presents a novel technique called Smart Sepsis Predictor (SSP) to predict sepsis onset in patients admitted to an intensive care unit (ICU). SSP is a deep neural network architecture that encompasses long short-term memory (LSTM), convolutional, and fully connected layers to achieve early prediction of sepsis. SSP can work in two modes; Mode 1 uses demographic data and vital signs, and Mode 2 uses laboratory test results in addition to demographic data and vital signs. To evaluate SSP, we have used the 2019 PhysioNet/CinC Challenge dataset, which includes the records of 40,366 patients admitted to the ICU. RESULTS: To compare SSP with existing state-of-the-art methods, we have measured the accuracy of the SSP in 4-, 8-, and 12-h prediction windows using publicly available data. Our results show that the SSP performance in Mode 1 and Mode 2 is much higher than existing methods, achieving an area under the receiver operating characteristic curve (AUROC) of 0.89 and 0.92, 0.88 and 0.87, and 0.86 and 0.84 for 4 h, 8 h, and 12 h before sepsis onset, respectively. CONCLUSIONS: Using ICU data, sepsis onset can be predicted up to 12 h in advance. Our findings offer an early solution for mitigating the risk of sepsis onset.

Predicting diabetes second-line therapy initiation in the Australian population via time span-guided neural attention network.

INTRODUCTION: The first line of treatment for people with Diabetes mellitus is metformin. However, over the course of the disease metformin may fail to achieve appropriate glycemic control, and a second-line therapy may become necessary. In this paper we introduce Tangle, a time span-guided neural attention model that can accurately and timely predict the upcoming need for a second-line diabetes therapy from administrative data in the Australian adult population. The method is suitable for designing automatic therapy review recommendations for patients and their providers without the need to collect clinical measures. DATA: We analyzed seven years of de-identified records (2008-2014) of the 10% publicly available linked sample of Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Scheme (PBS) electronic databases of Australia. METHODS: By design, Tangle inherits the representational power of pre-trained word embedding, such as GloVe, to encode sequences of claims with the related MBS codes. Moreover, the proposed attention mechanism natively exploits the information hidden in the time span between two successive claims (measured in number of days). We compared the proposed method against state-of-the-art sequence classification methods. RESULTS: Tangle outperforms state-of-the-art recurrent neural networks, including attention-based models. In particular, when the proposed time span-guided attention strategy is coupled with pre-trained embedding methods, the model performance reaches an Area Under the ROC Curve of 90%, an improvement of almost 10 percentage points over an attentionless recurrent architecture. IMPLEMENTATION: Tangle is implemented in Python using Keras and it is hosted on GitHub at https://github.com/samuelefiorini/tangle.

Patterns and trends of potentially inappropriate high-density lipoprotein cholesterol testing in Australian adults at high risk of cardiovascular disease from 2008 to 2014: analysis of linked individual patient data from the Australian Medicare Benefits Schedule and Pharmaceutical Benefits Scheme.

OBJECTIVES: We examine the extent to which the adult Australian population on lipid-lowering medications receives the level of high-density lipoprotein cholesterol (HDL-C) testing recommended by national guidelines. DATA: We analysed records from 7 years (2008-2014) of the 10% publicly available sample of deidentified, individual level, linked Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Scheme (PBS) electronic databases of Australia. METHODS: The PBS data were used to identify individuals on stable prescriptions of lipid-lowering treatment. The MBS data were used to estimate the annual frequency of HDL-C testing. We developed a methodology to address the issue of 'episode coning' in the MBS data, which causes an undercounting of pathology tests. We used a published figure on the proportion of unreported HDL-C tests to correct for the undercounting and estimate the probability that an HDL-C test was performed. We judged appropriateness of testing frequency by comparing the HDL-C testing rate to guidelines' recommendations of annual testing for people at high risk for cardiovascular disease. RESULTS: We estimated that approximately 49% of the population on stable lipid-lowering treatment did not receive any HDL-C test in a given year. We also found that approximately 19% of the same population received two or more HDL-C tests within the year. These levels of underutilisation and overutilisation have been changing at an average rate of 2% and -4% a year, respectively, since 2009. The yearly expenditure associated with test overutilisation was approximately $A4.3 million during the study period, while the cost averted because of test underutilisation was approximately $A11.3 million a year. CONCLUSIONS: We found that approximately half of Australians on stable lipid-lowering treatment may be having fewer HDL-C testing than recommended by national guidelines, while nearly one-fifth are having more tests than recommended.