SPANN: annotating single-cell resolution spatial transcriptome data with scRNA-seq data.

MOTIVATION: The rapid development of spatial transcriptome technologies has enabled researchers to acquire single-cell-level spatial data at an affordable price. However, computational analysis tools, such as annotation tools, tailored for these data are still lacking. Recently, many computational frameworks have emerged to integrate single-cell RNA sequencing (scRNA-seq) and spatial transcriptomics datasets. While some frameworks can utilize well-annotated scRNA-seq data to annotate spatial expression patterns, they overlook critical aspects. First, existing tools do not explicitly consider cell type mapping when aligning the two modalities. Second, current frameworks lack the capability to detect novel cells, which remains a key interest for biologists. RESULTS: To address these problems, we propose an annotation method for spatial transcriptome data called SPANN. The main tasks of SPANN are to transfer cell-type labels from well-annotated scRNA-seq data to newly generated single-cell resolution spatial transcriptome data and discover novel cells from spatial data. The major innovations of SPANN come from two aspects: SPANN automatically detects novel cells from unseen cell types while maintaining high annotation accuracy over known cell types. SPANN finds a mapping between spatial transcriptome samples and RNA data prototypes and thus conducts cell-type-level alignment. Comprehensive experiments using datasets from various spatial platforms demonstrate SPANN's capabilities in annotating known cell types and discovering novel cell states within complex tissue contexts. AVAILABILITY: The source code of SPANN can be accessed at https://github.com/ddb-qiwang/SPANN-torch. CONTACT: dengmh@math.pku.edu.cn.

Interpretable feature extraction and dimensionality reduction in ESM2 for protein localization prediction.

As the application of large language models (LLMs) has broadened into the realm of biological predictions, leveraging their capacity for self-supervised learning to create feature representations of amino acid sequences, these models have set a new benchmark in tackling downstream challenges, such as subcellular localization. However, previous studies have primarily focused on either the structural design of models or differing strategies for fine-tuning, largely overlooking investigations into the nature of the features derived from LLMs. In this research, we propose different ESM2 representation extraction strategies, considering both the character type and position within the ESM2 input sequence. Using model dimensionality reduction, predictive analysis and interpretability techniques, we have illuminated potential associations between diverse feature types and specific subcellular localizations. Particularly, the prediction of Mitochondrion and Golgi apparatus prefer segments feature closer to the N-terminal, and phosphorylation site-based features could mirror phosphorylation properties. We also evaluate the prediction performance and interpretability robustness of Random Forest and Deep Neural Networks with varied feature inputs. This work offers novel insights into maximizing LLMs' utility, understanding their mechanisms, and extracting biological domain knowledge. Furthermore, we have made the code, feature extraction API, and all relevant materials available at https://github.com/yujuan-zhang/feature-representation-for-LLMs.

Generative models for protein sequence modeling: recent advances and future directions.

The widespread adoption of high-throughput omics technologies has exponentially increased the amount of protein sequence data involved in many salient disease pathways and their respective therapeutics and diagnostics. Despite the availability of large-scale sequence data, the lack of experimental fitness annotations underpins the need for self-supervised and unsupervised machine learning (ML) methods. These techniques leverage the meaningful features encoded in abundant unlabeled sequences to accomplish complex protein engineering tasks. Proficiency in the rapidly evolving fields of protein engineering and generative AI is required to realize the full potential of ML models as a tool for protein fitness landscape navigation. Here, we support this work by (i) providing an overview of the architecture and mathematical details of the most successful ML models applicable to sequence data (e.g. variational autoencoders, autoregressive models, generative adversarial neural networks, and diffusion models), (ii) guiding how to effectively implement these models on protein sequence data to predict fitness or generate high-fitness sequences and (iii) highlighting several successful studies that implement these techniques in protein engineering (from paratope regions and subcellular localization prediction to high-fitness sequences and protein design rules generation). By providing a comprehensive survey of model details, novel architecture developments, comparisons of model applications, and current challenges, this study intends to provide structured guidance and robust framework for delivering a prospective outlook in the ML-driven protein engineering field.

Molecular substructure tree generative model for de novo drug design.

Deep learning shortens the cycle of the drug discovery for its success in extracting features of molecules and proteins. Generating new molecules with deep learning methods could enlarge the molecule space and obtain molecules with specific properties. However, it is also a challenging task considering that the connections between atoms are constrained by chemical rules. Aiming at generating and optimizing new valid molecules, this article proposed Molecular Substructure Tree Generative Model, in which the molecule is generated by adding substructure gradually. The proposed model is based on the Variational Auto-Encoder architecture, which uses the encoder to map molecules to the latent vector space, and then builds an autoregressive generative model as a decoder to generate new molecules from Gaussian distribution. At the same time, for the molecular optimization task, a molecular optimization model based on CycleGAN was constructed. Experiments showed that the model could generate valid and novel molecules, and the optimized model effectively improves the molecular properties.

The Disproportionate Impact of Coronavirus Disease 2019 (COVID-19) Pandemic on Healthcare-Associated Infections in Community Hospitals: Need for Expanding the Infectious Disease Workforce.

BACKGROUND: The coronavirus disease 2019 (COVID-19) pandemic had a considerable impact on US healthcare systems, straining hospital resources, staff, and operations. However, a comprehensive assessment of the impact on healthcare-associated infections (HAIs) across different hospitals with varying level of infectious disease (ID) physician expertise, resources, and infrastructure is lacking. METHODS: This retrospective longitudinal multicenter cohort study included central-line-associated bloodstream infections (CLABSIs), catheter-associated urinary tract infections (CAUTIs), Clostridioides difficile infections (CDIs), and ventilator-associated events (VAEs) from 53 hospitals (academic and community) in Southeastern United States from 1 January 2018 to 31 March 2021. Segmented negative binomial regression generalized estimating equations models estimated changes in monthly incidence rates in the baseline (01/2018-02/2020) compared to the pandemic period (03/2020-03/2021, further divided into three pandemic phases). RESULTS: CLABSIs and VAEs increased by 24% and 34%, respectively, during the pandemic period. VAEs increased in all phases of the pandemic, while CLABSIs increased in later phases of the pandemic. CDI trend increased by 4.2% per month in the pandemic period. On stratifying the analysis by hospital characteristics, the impact of the pandemic on healthcare-associated infections was more significant in smaller sized and community hospitals. CAUTIs did not change significantly during the pandemic across all hospital types. CONCLUSIONS: CLABSIs, VAEs, and CDIs increased significantly during the pandemic, especially in smaller community hospitals, most of which lack ID physician expertise. Future efforts should focus on better understanding challenges faced by community hospitals, strengthening the infection prevention infrastructure, and expanding the ID workforce, particularly to community hospitals.

A denoising method of ECG signal based on variational autoencoder and masked convolution.

Wearable electrocardiogram (ECG) equipment can realize continuous monitoring of cardiovascular diseases, but these devices are more susceptible to interference from various noises, which will seriously reduce the diagnostic correctness. In this work, a novel noise reduction model for ECG signals is proposed based on variational autoencoder and masked convolution. The variational Bayesian inference is conducted to capture the global features of the ECG signals by encouraging the approximate posterior of the latent variables to fit the prior distribution, and we use the skip connection and feature concatenation to realize the information interaction across the channels. To strengthen the connection of local features of the ECG signals, the masked convolution module is used to extract local feature information, which supplement the global features and the noise reduction performance of whole model can be greatly improved. Experiments are carried out on the MIT-BIH arrythmia database, and the results display that the performance metrics of signal-to-noise ratio (SNR) and root mean square error (RMSE) are significantly improved compared with other approaches while causing less signal distortion.

Application of Variational AutoEncoder (VAE) Model and Image Processing Approaches in Game Design.

In recent decades, the Variational AutoEncoder (VAE) model has shown good potential and capability in image generation and dimensionality reduction. The combination of VAE and various machine learning frameworks has also worked effectively in different daily life applications, however its possible use and effectiveness in modern game design has seldom been explored nor assessed. The use of its feature extractor for data clustering has also been minimally discussed in the literature neither. This study first attempts to explore different mathematical properties of the VAE model, in particular, the theoretical framework of the encoding and decoding processes, the possible achievable lower bound and loss functions of different applications; then applies the established VAE model to generate new game levels based on two well-known game settings; and to validate the effectiveness of its data clustering mechanism with the aid of the Modified National Institute of Standards and Technology (MNIST) database. Respective statistical metrics and assessments are also utilized to evaluate the performance of the proposed VAE model in aforementioned case studies. Based on the statistical and graphical results, several potential deficiencies, for example, difficulties in handling high-dimensional and vast datasets, as well as insufficient clarity of outputs are discussed; then measures of future enhancement, such as tokenization and the combination of VAE and GAN models, are also outlined. Hopefully, this can ultimately maximize the strengths and advantages of VAE for future game design tasks and relevant industrial missions.

Semi-Supervised Variational Autoencoders for Out-of-Distribution Generation.

Humans are able to quickly adapt to new situations, learn effectively with limited data, and create unique combinations of basic concepts. In contrast, generalizing out-of-distribution (OOD) data and achieving combinatorial generalizations are fundamental challenges for machine learning models. Moreover, obtaining high-quality labeled examples can be very time-consuming and expensive, particularly when specialized skills are required for labeling. To address these issues, we propose BtVAE, a method that utilizes conditional VAE models to achieve combinatorial generalization in certain scenarios and consequently to generate out-of-distribution (OOD) data in a semi-supervised manner. Unlike previous approaches that use new factors of variation during testing, our method uses only existing attributes from the training data but in ways that were not seen during training (e.g., small objects of a specific shape during training and large objects of the same shape during testing).

Image-level trajectory inference of tau pathology using variational autoencoder for Flortaucipir PET.

PURPOSE: Alzheimer's disease (AD) studies revealed that abnormal deposition of tau spreads in a specific spatial pattern, namely Braak stage. However, Braak staging is based on post mortem brains, each of which represents the cross section of the tau trajectory in disease progression, and numerous studies were reported that do not conform to that model. This study thus aimed to identify the tau trajectory and quantify the tau progression in a data-driven approach with the continuous latent space learned by variational autoencoder (VAE). METHODS: A total of 1080 [18F]Flortaucipir brain positron emission tomography (PET) images were collected from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. VAE was built to compress the hidden features from tau images in latent space. Hierarchical agglomerative clustering and minimum spanning tree (MST) were applied to organize the features and calibrate them to the tau progression, thus deriving pseudo-time. The image-level tau trajectory was inferred by continuously sampling across the calibrated latent features. We assessed the pseudo-time with regard to tau standardized uptake value ratio (SUVr) in AD-vulnerable regions, amyloid deposit, glucose metabolism, cognitive scores, and clinical diagnosis. RESULTS: We identified four clusters that plausibly capture certain stages of AD and organized the clusters in the latent space. The inferred tau trajectory agreed with the Braak staging. According to the derived pseudo-time, tau first deposits in the parahippocampal and amygdala, and then spreads to the fusiform, inferior temporal lobe, and posterior cingulate. Prior to the regional tau deposition, amyloid accumulates first. CONCLUSION: The spatiotemporal trajectory of tau progression inferred in this study was consistent with Braak staging. The profile of other biomarkers in disease progression agreed well with previous findings. We addressed that this approach additionally has the potential to quantify tau progression as a continuous variable by taking a whole-brain tau image into account.

An efficient deep neural network framework for COVID-19 lung infection segmentation.

Since the outbreak of Coronavirus Disease 2019 (COVID-19) in 2020, it has significantly affected the global health system. The use of deep learning technology to automatically segment pneumonia lesions from Computed Tomography (CT) images can greatly reduce the workload of physicians and expand traditional diagnostic methods. However, there are still some challenges to tackle the task, including obtaining high-quality annotations and subtle differences between classes. In the present study, a novel deep neural network based on Resnet architecture is proposed to automatically segment infected areas from CT images. To reduce the annotation cost, a Vector Quantized Variational AutoEncoder (VQ-VAE) branch is added to reconstruct the input images for purpose of regularizing the shared decoder and the latent maps of the VQ-VAE are utilized to further improve the feature representation. Moreover, a novel proportions loss is presented for mitigating class imbalance and enhance the generalization ability of the model. In addition, a semi-supervised mechanism based on adversarial learning to the network has been proposed, which can utilize the information of the trusted region in unlabeled images to further regularize the network. Extensive experiments on the COVID-SemiSeg are performed to verify the superiority of the proposed method, and the results are in line with expectations.

DA-LSTM-VAE: Dual-Stage Attention-Based LSTM-VAE for KPI Anomaly Detection.

To ensure the normal operation of the system, the enterprise's operations engineer will monitor the system through the KPI (key performance indicator). For example, web page visits, server memory utilization, etc. KPI anomaly detection is a core technology, which is of great significance for rapid fault detection and repair. This paper proposes a novel dual-stage attention-based LSTM-VAE (DA-LSTM-VAE) model for KPI anomaly detection. Firstly, in order to capture time correlation in KPI data, long-short-term memory (LSTM) units are used to replace traditional neurons in the variational autoencoder (VAE). Then, in order to improve the effect of KPI anomaly detection, an attention mechanism is introduced into the input stage of the encoder and decoder, respectively. During the input stage of the encoder, a time attention mechanism is adopted to assign different weights to different time points, which can adaptively select important input sequences to avoid the influence of noise in the data. During the input stage of the decoder, a feature attention mechanism is adopted to adaptively select important latent variable representations, which can capture the long-term dependence of time series better. In addition, this paper proposes an adaptive threshold method based on anomaly scores measured by reconstruction probability, which can minimize false positives and false negatives and avoid adjustment of the threshold manually. Experimental results in a public dataset show that the proposed method in this paper outperforms other baseline methods.

A robust variational autoencoder using beta divergence.

The presence of outliers can severely degrade learned representations and performance of deep learning methods and hence disproportionately affect the training process, leading to incorrect conclusions about the data. For example, anomaly detection using deep generative models is typically only possible when similar anomalies (or outliers) are not present in the training data. Here we focus on variational autoencoders (VAEs). While the VAE is a popular framework for anomaly detection tasks, we observe that the VAE is unable to detect outliers when the training data contains anomalies that have the same distribution as those in test data. In this paper we focus on robustness to outliers in training data in VAE settings using concepts from robust statistics. We propose a variational lower bound that leads to a robust VAE model that has the same computational complexity as the standard VAE and contains a single automatically-adjusted tuning parameter to control the degree of robustness. We present mathematical formulations for robust variational autoencoders (RVAEs) for Bernoulli, Gaussian and categorical variables. The RVAE model is based on beta-divergence rather than the standard Kullback-Leibler (KL) divergence. We demonstrate the performance of our proposed ß-divergence-based autoencoder for a variety of image and categorical datasets showing improved robustness to outliers both qualitatively and quantitatively. We also illustrate the use of our robust VAE for detection of lesions in brain images, formulated as an anomaly detection task. Finally, we suggest a method to tune the hyperparameter of RVAE which makes our model completely unsupervised.

A Visual and VAE Based Hierarchical Indoor Localization Method.

Precise localization and pose estimation in indoor environments are commonly employed in a wide range of applications, including robotics, augmented reality, and navigation and positioning services. Such applications can be solved via visual-based localization using a pre-built 3D model. The increase in searching space associated with large scenes can be overcome by retrieving images in advance and subsequently estimating the pose. The majority of current deep learning-based image retrieval methods require labeled data, which increase data annotation costs and complicate the acquisition of data. In this paper, we propose an unsupervised hierarchical indoor localization framework that integrates an unsupervised network variational autoencoder (VAE) with a visual-based Structure-from-Motion (SfM) approach in order to extract global and local features. During the localization process, global features are applied for the image retrieval at the level of the scene map in order to obtain candidate images, and are subsequently used to estimate the pose from 2D-3D matches between query and candidate images. RGB images only are used as the input of the proposed localization system, which is both convenient and challenging. Experimental results reveal that the proposed method can localize images within 0.16 m and 4° in the 7-Scenes data sets and 32.8% within 5 m and 20° in the Baidu data set. Furthermore, our proposed method achieves a higher precision compared to advanced methods.

Attention Autoencoder for Generative Latent Representational Learning in Anomaly Detection.

Today, accurate and automated abnormality diagnosis and identification have become of paramount importance as they are involved in many critical and life-saving scenarios. To accomplish such frontiers, we propose three artificial intelligence models through the application of deep learning algorithms to analyze and detect anomalies in human heartbeat signals. The three proposed models include an attention autoencoder that maps input data to a lower-dimensional latent representation with maximum feature retention, and a reconstruction decoder with minimum remodeling loss. The autoencoder has an embedded attention module at the bottleneck to learn the salient activations of the encoded distribution. Additionally, a variational autoencoder (VAE) and a long short-term memory (LSTM) network is designed to learn the Gaussian distribution of the generative reconstruction and time-series sequential data analysis. The three proposed models displayed outstanding ability to detect anomalies on the evaluated five thousand electrocardiogram (ECG5000) signals with 99% accuracy and 99.3% precision score in detecting healthy heartbeats from patients with severe congestive heart failure.

Log Sequence Anomaly Detection Method Based on Contrastive Adversarial Training and Dual Feature Extraction.

The log messages generated in the system reflect the state of the system at all times. The realization of autonomous detection of abnormalities in log messages can help operators find abnormalities in time and provide a basis for analyzing the causes of abnormalities. First, this paper proposes a log sequence anomaly detection method based on contrastive adversarial training and dual feature extraction. This method uses BERT (Bidirectional Encoder Representations from Transformers) and VAE (Variational Auto-Encoder) to extract the semantic features and statistical features of the log sequence, respectively, and the dual features are combined to perform anomaly detection on the log sequence, with a novel contrastive adversarial training method also used to train the model. In addition, this paper introduces the method of obtaining statistical features of log sequence and the method of combining semantic features with statistical features. Furthermore, the specific process of contrastive adversarial training is described. Finally, an experimental comparison is carried out, and the experimental results show that the method in this paper is better than the contrasted log sequence anomaly detection method.

Characterisation of the online public debate od MMR vaccine against measles, mumps and rubella on the Polish Internet.

INTRODUCTION: MMR vaccine is a controversial topic of public debate. The controversies include such issues as autism, adjuvants or ethical questions related to the culturing of the rubella virus on human cell lines. OBJECTIVE: The objective was to characterise the public debate on MMR vaccination on the Polish Internet between January 2018 and June 2020. MATERIAL AND METHODS: Quantitative and qualitative analysis of Polish-language online content between 1 January 2018 and 30 June 2020 related to MMR vaccination. The quantitative analysis comprised all available mentions of MMR vaccination in postings (n=14,632), while qualitative analysis relied on a systematic sample of 819 mentions. RESULTS: Quantitative study: 79.6% of MMR vaccine-related postings were published on Facebook, 6.9% on Twitter, and the remaining 14.6% appeared on other websites. There were two surges in posting count in November 2018 and March 2019. Qualitative study: 48% of postings expressed anti-vaccination sentiment, 33% were pro-vaccination and 19% were neutral. CONCLUSIONS: The social media play a significant role in the dissemination of untrue medical claims regarding MMR vaccination. A substantial part of the discussion about MMR vaccination in Poland takes place on Facebook. Despite the general availability of research results stating the absence of a link between autism and vaccination, this is an ongoing most frequent topics in the MMR debate. At the same time, more postings on that topic expressed pro-vaccination rather than anti-vaccination sentiment.

Prevalence, Clinical Characteristics, and Outcomes Related to Ventilator-Associated Events in Neurocritically Ill Patients.

BACKGROUND: The prevalence, characteristics, and outcomes related to the ventilator-associated event(s) (VAE) in neurocritically ill patients are unknown and examined in this study. METHODS: A retrospective study was performed on neurocritically ill patients at a 413-bed level 1 trauma and stroke center who received three or more days of mechanical ventilation to describe rates of VAE, describe characteristics of patients with VAE, and examine the association of VAE on ventilator days, mortality, length of stay, and discharge to home. RESULTS: Over a 5-year period from 2014 through 2018, 855 neurocritically ill patients requiring mechanical ventilation were identified. A total of 147 VAEs occurred in 130 (15.2%) patients with an overall VAE rate of 13 per 1000 ventilator days and occurred across age, sex, BMI, and admission Glasgow Coma Scores. The average time from the start of ventilation to a VAE was 5 (range 3-48) days after initiation of mechanical ventilation. Using Centers for Disease Control and Prevention definitions, VAEs met criteria for a ventilator-associated condition in 58% of events (n = 85), infection-related VAE in 22% of events (n = 33), and possible ventilator-associated pneumonia in 20% of events (n = 29). A most common trigger for VAE was an increase in positive end-expiratory pressure (84%). Presence of a VAE was associated with an increase in duration of mechanical ventilation (17.4[IQR 20.5] vs. 7.9[8.9] days, p < 0.001, 95% CI 7.86-13.92), intensive care unit (ICU) length of stay (20.2[1.1] vs. 12.5[0.4] days, p < 0.001 95% CI 5.3-10.02), but not associated with in-patient mortality (34.1 vs. 31.3%. 95% CI 0.76-1.69) or discharge to home (12.7% vs. 16.3%, 95% 0.47-1.29). CONCLUSIONS: VAE are prevalent in the neurocritically ill. They result in an increased duration of mechanical ventilation and ICU length of stay, but may not be associated with in-hospital mortality or discharge to home.

Breast papillary lesions diagnosed and treated using ultrasound-guided vacuum-assisted excision.

BACKGROUND: The management of papillary lesions of the breast remains controversial, and thus, we assessed the value of vacuum-assisted excision (VAE)-guided ultrasound in the diagnosis and treatment of breast papillary lesions. METHODS: We retrospectively reviewed the data of 108 patients with papillary lesions diagnosed using VAE between August 2014 and January 2019. Cases without postoperative breast imaging in the follow-up were excluded, and 85 cases were eligible for the study. The follow-up period ranged from 6 to 53 months, with 38 months on average. All the papillary lesions were located away from the skin or nipple with a size less than or equal to 30 mm, and the lesions categorized as C2-4b were completely excised using VAE. All VAEs were performed using an 8-gauge vacuum-assisted biopsy needle under the guidance of ultrasound using a 10 MHz linear probe. RESULTS: Most patients with breast papillary lesions were asymptomatic (56.5%), and when the size of the breast papillary lesion was more than 20 mm on ultrasound imaging, atypical hyperplasia may have been concomitant. Breast lesions might have been pathologically diagnosed as papilloma after biopsy when they were categorized as BI-RADS 4a on ultrasound images. The rate of underestimation was 7.7% in papillary lesions diagnosed with VAE, and the recurrence rate of papilloma after VAE was low. CONCLUSIONS: Breast papilloma was a common lesion on ultrasonographic screening, and VAE was applicable for completely excising small papillomas, even papillomas with atypical hyperplasia, to obtain an accurate diagnosis with a low rate of underestimation and recurrence. We believe that papilloma diagnosed by VAE might not require immediate excision, and imaging follow-up may be safe for at least 3 years.

LSTM-Based VAE-GAN for Time-Series Anomaly Detection.

Time series anomaly detection is widely used to monitor the equipment sates through the data collected in the form of time series. At present, the deep learning method based on generative adversarial networks (GAN) has emerged for time series anomaly detection. However, this method needs to find the best mapping from real-time space to the latent space at the anomaly detection stage, which brings new errors and takes a long time. In this paper, we propose a long short-term memory-based variational autoencoder generation adversarial networks (LSTM-based VAE-GAN) method for time series anomaly detection, which effectively solves the above problems. Our method jointly trains the encoder, the generator and the discriminator to take advantage of the mapping ability of the encoder and the discrimination ability of the discriminator simultaneously. The long short-term memory (LSTM) networks are used as the encoder, the generator and the discriminator. At the anomaly detection stage, anomalies are detected based on reconstruction difference and discrimination results. Experimental results show that the proposed method can quickly and accurately detect anomalies.

Detection and Analysis of Heartbeats in Seismocardiogram Signals.

This paper presents an unsupervised methodology to analyze SeismoCardioGram (SCG) signals. Starting from raw accelerometric data, heartbeat complexes are extracted and annotated, using a two-step procedure. An unsupervised calibration procedure is added to better adapt to different user patterns. Results show that the performance scores achieved by the proposed methodology improve over related literature: on average, 98.5% sensitivity and 98.6% precision are achieved in beat detection, whereas RMS (Root Mean Square) error in heartbeat interval estimation is as low as 4.6 ms. This allows SCG heartbeat complexes to be reliably extracted. Then, the morphological information of such waveforms is further processed by means of a modular Convolutional Variational AutoEncoder network, aiming at extracting compressed, meaningful representation. After unsupervised training, the VAE network is able to recognize different signal morphologies, associating each user to its specific patterns with high accuracy, as indicated by specific performance metrics (including adjusted random and mutual information score, completeness, and homogeneity). Finally, a Linear Model is used to interpret the results of clustering in the learned latent space, highlighting the impact of different VAE architectural parameters (i.e., number of stacked convolutional units and dimension of latent space).