Local Dataset Research Articles

Class-Hidden Client-Side Watermarking in Federated Learning

Federated learning consists of a central aggregator and multiple clients, forming a distributed structure that effectively protects data privacy. However, since all participants can access the global model, the risk of model leakage increases, especially when unreliable participants are involved. To safeguard model copyright while enhancing the robustness and secrecy of the watermark, this paper proposes a client-side watermarking scheme. Specifically, the proposed method introduces an additional watermark class, expanding the output layer of the client model into an N+1-class classifier. The client’s local model is then trained using both the watermark dataset and the local dataset. Notably, before uploading to the server, the parameters of the watermark class are removed from the output layer and stored locally. Additionally, the client uploads amplified parameters to address the potential weakening of the watermark during the aggregation. After aggregation, the global model is distributed to the clients for local training. Through multiple rounds of iteration, the saved watermark parameters are continuously updated until the global model converges. On the MNIST, CIFAR-100, and CIFAR-10 datasets, the watermark detection rates on VGG-16 and ResNet-18 reached 100%. Furthermore, extensive experiments demonstrate that this method has minimal impact on model performance and exhibits strong robustness against pruning and fine-tuning attacks.

Dynamic selectout and voting-based federated learning for enhanced medical image analysis

Abstract Federated Learning (FL) is a promising technique for training machine learning models on distributed, privacy-aware datasets. Nevertheless, FL faces difficulties with agent/client participation, model performance, and the heterogeneous nature of networked data sources when it comes to distributed healthcare systems. When these agents work together in the system, it is imperative to tackle the complexities of distributed deep learning. We suggest a novel approach that uses a voting mechanism and dynamic SelectOut inside the FL framework to address these problems. Local medical imaging datasets frequently show diversity in distribution and data imbalances. In certain situations, traditional FL techniques like FedProx and Federated Averaging (FedAvg), which depend on data size to weight contributions, might not be the optimal choice. We improve parameter aggregation and client selection unpredictability and increase the model's adaptability to imbalanced and heterogeneous datasets, our proposed FedVoteNet model introduces SelectOut techniques based on Voting methodology. Based on how much their local performance has improved from the last communication cycle, we arbitrarily remove clients. Additionally eliminated are clients whose model weights when combined with the global model adversely affect its performance. After each communication cycle, clients that improve both their local performance and their contribution to the global model are awarded higher voting values. This encourages more significant and effective contributions from clients by providing incentives for them to actively increase the diversity of their training data. We assess our approach on a dataset of medical images, including MRI scans, and find that the FL model performs noticeably better (F1 Score=0.968, Sensitivity=0.977, Specificity=0.945, and AUC=0.950). The voting system and the dynamic SelectOut algorithms improve the convergence of the FL model and successfully handle the difficulties presented by uneven and heterogeneous datasets. To sum up, our proposed approach uses voting and dynamic SelectOut techniques to improve FL performance on a variety of uneven, distributed, and varied datasets. This strategy has a lot of potential to improve FL across a range of applications, especially those that prioritize data privacy, diversity, and performance.

Two-layer homolog network approach for PFAS nontarget screening and retrospective data mining

The rapid increase of novel per- and polyfluoroalkyl substances (PFAS) raises concerns, while their identification remains challenging. Here, we develop a two-layer homolog network approach for PFAS nontarget screening using mass spectrometry. The first layer constructs networks between homologs, with evaluation showing that it filters 94% of false candidates. The second layer builds a network between classes to expedite the identification of PFAS. We detected 94 PFAS in twelve waterproof products and two related industrial sludges, including 36 novel PFAS not previously reported in any sample. A local dataset is constructed for retrospective analysis by re-analyzing our previous samples, revealing fifteen novel PFAS in samples collected in 2005. The retrieval of the public database MassIVE uncovers novel PFAS in samples from seven countries. Here, we reveal the historic and global presence of novel PFAS, providing guidance for the management and policy-making concerning persistent chemicals.

Factors affecting the fitting of a global geopotential model to local geodetic datasets over local areas in Egypt using multiple linear regression approach

Factors affecting the fitting of a global geopotential model to local geodetic datasets over local areas in Egypt using multiple linear regression approach

Development of an SDG-driven Convolutional Neural Network (CNN) model for multi-class classification of Tomato Leaf diseases using combined public and local datasets

Development of an SDG-driven Convolutional Neural Network (CNN) model for multi-class classification of Tomato Leaf diseases using combined public and local datasets

Towards Privacy-preserving and Fairness-aware Federated Learning Framework

Federated Learning (FL) enables the distributed training of a model across multiple data owners under the orchestration of a central server responsible for aggregating the models generated by the different clients. However, the original approach of FL has significant shortcomings related to privacy and fairness requirements. Specifically, the observation of the model updates may lead to privacy issues, such as membership inference attacks, while the use of imbalanced local datasets can introduce or amplify classification biases, especially for minority groups. In this work, we show that these biases can be exploited to increase the likelihood of privacy attacks against these groups. To do so, we propose a novel inference attack exploiting the knowledge of group fairness metrics during the training of the global model. Then to thwart this attack, we define a fairness-aware encrypted-domain aggregation algorithm that is differentially-private by design thanks to the approximate precision loss of the threshold multi-key CKKS homomorphic encryption scheme. Finally, we demonstrate the good performance of our proposal both in terms of fairness and privacy through experiments conducted over three real datasets.

Prediction and Detection of Scam Threats on Digital Platforms for Indonesian Users Using Machine Learning Models

Scam threats on digital platforms continue to rise alongside the rapid adoption of technology in Indonesia. The unique characteristics of Indonesian digital users, such as low digital literacy and high social media usage, make them particularly vulnerable to various forms of scams, including phishing, impersonation, and emotional manipulation. This study aims to develop a machine learning-based model for predicting and detecting scams by identifying threat patterns within a local context. The methodology involves collecting a survey-based dataset from Indonesian digital users, capturing language patterns and user interaction behaviors. The dataset was processed through text-cleaning techniques, tokenization, normalization, and representation using TF-IDF and Word Embeddings. The machine learning models employed in this study are Random Forest and Support Vector Machine (SVM), evaluated using accuracy, precision, recall, and F1-score metrics. Hyperparameter tuning was conducted to optimize model performance, while k-fold cross-validation was utilized to minimize the risk of overfitting. The results indicate that the Random Forest model achieved the best performance, with an accuracy of 92.5%, precision of 90.7%, recall of 94.1%, and F1-score of 92.4%. The use of local datasets improved detection accuracy by 7.8% compared to global datasets, highlighting the critical importance of contextual representation in identifying scam patterns specific to Indonesia. The model was also effective in recognizing unique threat patterns, such as the use of informal language and manipulative phrases in scam messages. This study makes a significant contribution to the field of digital security by providing an effective machine learning-based approach to detecting scam threats in Indonesia. Moreover, the findings underscore the importance of developing local datasets and educating users as part of a holistic solution to enhance digital security. These insights emphasize the necessity of incorporating cultural and contextual factors into technology-driven approaches for combating scams in developing countries like Indonesia

Germanium, indium, gallium and cadmium in zinc ores: a mineral system approach

Sphalerite is the main Zn ore mineral and is the primary source of Cd, Ge and In, and a minor source of Ga. Owing to a shift from fossil fuel to renewable energy sources, these four minor elements have progressively become more important to the economy. Despite this, resources of Cd, Ga, Ge and In are rarely reported, as these metals are not considered material to the economics of resource development. As a result, the distribution of these elements between and within deposits is poorly known, and national and international resources are largely unreported. Following previous studies, we have compiled analytical data for Cd, Ga, Ge and In from sphalerite and used global and local ore geochemical datasets to assess geochemical controls on the concentration of these elements in Zn deposits. Our results are similar to those of previous studies and suggest that lower-temperature deposits are enriched in Ge, whereas higher-temperature deposits are enriched in In. However, modelling of hydrothermal geochemistry indicates other factors are important in concentrating these metals. In particular, the oxidation state of the fluid (oxidised vs reduced) and the depositional mechanisms also have a strong influence in Ga, Ge and In enrichment. Reduction of oxidised fluids is particularly effective in depositing Ge, whereas cooling very effectively deposits In and, in some cases, Ge. As a consequence, some higher-temperature deposits (e.g. high sulfidation epithermal and some volcanic-hosted massive sulfide) can be Ge-enriched, and some lower-temperature deposits (e.g. siliciclastic–carbonate shale-hosted deposits) can be enriched in In. Using the existing ore geochemical data and calculated characteristic Ge/Zn and In/Zn ratios, indicative estimates have been made on the endowment of Australian Zn deposits of Ge and In. These estimates highlight the potential of the North Australian Zinc Belt for Ge and for VHMS deposits for In. Although there is a large amount of uncertainty in the estimates, they are indicative of the potential for these metals in Australia.

Impact of image formation factors on material discrimination in spectral CT.

The accuracy of material decomposition in spectral CT depends on the information quality captured in image acquisition, a factor that cannot be adequately assessed using conventional image quality metrologies due to the multi-energy nature of spectral CT. This work used metrologies specific to spectral CT to evaluate the impact of acquisition conditions on the quality of spectral CT images and accuracy of material decomposition techniques.
Approach: Computational phantoms were created with cylindrical shapes and variable sizes (20-40 cm), containing inserts of iodine and gadolinium (1-8 mg/mL). The phantoms were imaged using a validated CT simulator modeling a clinical photon-counting CT scanner. The acquisitions were done at different detector energy thresholds (50-90 keV) and tube currents (25-250 mAs). The images were used to develop and train a data-driven material identification and quantification algorithm. Two spectral metrologies, multivariate contrast-to-noise ratio (CNR) and separability index, were used to characterize the impact of energy threshold, tube current, phantom size, and material concentration on signal quality. The results were interpreted in terms of figures of merit of accuracy for classification and mean absolute error (MAE) and root mean squared error (RMSE) for regression. 
Main results: Signal quality for iodine and gadolinium was maximized with a low energy threshold, high tube current, and small phantom size. While conventional CNR terms predicted variable image quality for two-thirds of all conditions, multivariate CNR was above 10 for half of those. Separability index showed that for a phantom size greater than 30 cm, a minimum of 75-110 mAs is required to separate 2 mg/mL of iodine and gadolinium. For both classification and regression tasks, a random forest model with a local statistics dataset provided the best performance. Across conditions, classification performance was 0.66-0.99 for I accuracy, 0.72-0.99 for Gd accuracy. Regression performance was 0.02-0.91 mg/mL I and 0.02 - 0.59 mg/mL Gd for MAE and 0.11-1.08 mg/mL I and 0.07-0.76 mg/mL Gd for RMSE.
Significance: Multivariate CNR and separability index metrologies can predict material decomposition performance. Theses metrics demonstrated that the decomposition of iodine and gadolinium have higher separability when the acquisition is done at a lower energy threshold, with a higher tube current, and when the imaged object has a smaller size. Object size had the largest impact on metrics and decomposition performance.&#xD.

A novel method of cognitive overload assessment based on a fusion feature selection using EEG signals.

Cognitive overload, as an overload state of cognitive workload, negatively impacts individuals' task performance and mental health. Cognitive overload assessment models based on Electroencephalography (EEG) can effectively prevent the occurrence of overload through early warning, thereby enhancing task execution efficiency and safeguarding individuals' mental health. Although existing EEG-based cognitive load assessment methods have achieved significant research outcomes, evaluating cognitive overload remains an ongoing challenge. Current research aims to develop an effective cognitive overload assessment model and enhance its efficacy through feature selection methods.
Approach. In the cognitive overload assessment model, we firstly employ Variational Mode Decomposition (VMD) to adaptively decompose the signal from each channel into four sub-band signals to capture valuable time-frequency information. Subsequently, frequency domain features are extracted from each sub-band, and an effective feature selection method based on Mutual Information (MI) and Neighborhood Component Analysis (NCA) was applied for feature selection, which optimizes the distribution of the feature space while considering feature correlations, making the selected features more representative. Finally, traditional machine learning methods are utilized for classification, and the effectiveness of the proposed method is tested using both offline and online classification results.
Main results. The average accuracy of offline cognitive overload assessment using the proposed method on local and open datasets is 82.44 ± 1.59% and 78.24 ± 1.43%, respectively. The average classification accuracy of its online cognitive overload assessment is about 79.90 ± 2.53%. This indicates that the proposed method can effectively assess cognitive overload under both offline and online conditions. Furthermore, we found that higher-frequency sub-bands are more advantageous for cognitive overload assessment.
Significance. EEG signals can be used for effectively cognitive overload assessment, and the integration of feature selection methods enhances the accuracy of the evaluation, providing reliable methodological support for future cognitive overload monitoring in human-computer interaction systems.&#xD.

DLLabelsCT: Annotation tool using deep transfer learning to assist in creating new datasets from abdominal computed tomography scans, case study: Pancreas.

The utilization of artificial intelligence (AI) is expanding significantly within medical research and, to some extent, in clinical practice. Deep learning (DL) applications, which use large convolutional neural networks (CNN), hold considerable potential, especially in optimizing radiological evaluations. However, training DL algorithms to clinical standards requires extensive datasets, and their processing is labor-intensive. In this study, we developed an annotation tool named DLLabelsCT that utilizes CNN models to accelerate the image analysis process. To validate DLLabelsCT, we trained a CNN model with a ResNet34 encoder and a UNet decoder to segment the pancreas on an open-access dataset and used the DL model to assist in annotating a local dataset, which was further used to refine the model. DLLabelsCT was also tested on two external testing datasets. The tool accelerates annotation by 3.4 times compared to a completely manual annotation method. Out of 3,715 CT scan slices in the testing datasets, 50% did not require editing when reviewing the segmentations made by the ResNet34-UNet model, and the mean and standard deviation of the Dice similarity coefficient was 0.82±0.24. DLLabelsCT is highly accurate and significantly saves time and resources. Furthermore, it can be easily modified to support other deep learning models for other organs, making it an efficient tool for future research involving larger datasets.

BiG-Fed: Bilevel Optimization Enhanced Graph-Aided Federated Learning

In federated learning (FL), due to the non-i.i.d. nature of distributedly owned local datasets, personalization is an important design goal. In this paper, we investigate FL scenarios in which data owners are related by a network topology (e.g., traffic prediction based on sensor networks). Existing personalized FL approaches cannot take this information into account. To address this limitation, we propose the Bilevel Optimization enhanced Graph-aided Federated Learning (BiG-Fed) approach. The inner weights enable local tasks to evolve towards personalization, and the outer shared weights on the server side target the non-i.i.d problem enabling individual tasks to evolve towards a global constraint space. To the best of our knowledge, BiG-Fed is the first bilevel optimization technique to enable FL approaches to cope with two nested optimization tasks at the FL server and FL clients simultaneously. Theoretical analysis shows that BiG-Fed is guaranteed to converge in an efficient manner. Extensive experiments on both synthetic and real-world data demonstrate significant superior performance of BiG-Fed over seven state-of-the-art methods.

An Experimental Study of Byzantine-Robust Aggregation Schemes in Federated Learning

Byzantine-robust federated learning aims at mitigating Byzantine failures during the federated training process, where malicious participants may upload arbitrary local updates to the central server to degrade the performance of the global model. In recent years, several robust aggregation schemes have been proposed to defend against malicious updates from Byzantine clients and improve the robustness of federated learning. These solutions were claimed to be Byzantine-robust, under certain assumptions. Other than that, new attack strategies are emerging, striving to circumvent the defense schemes. However, there is a lack of systematic comparison and empirical study thereof. In this paper, we conduct an experimental study of Byzantine-robust aggregation schemes under different attacks using two popular algorithms in federated learning, FedSGD and FedAvg . We first survey existing Byzantine attack strategies and Byzantine-robust aggregation schemes that aim to defend against Byzantine attacks. We also propose a new scheme, ClippedClustering , to enhance the robustness of a clustering-based scheme by automatically clipping the updates. Then we provide an experimental evaluation of eight aggregation schemes in the scenario of five different Byzantine attacks. Our results show that these aggregation schemes sustain relatively high accuracy in some cases but are ineffective in others. In particular, our proposed ClippedClustering successfully defends against most attacks under independent and IID local datasets. However, when the local datasets are Non-IID, the performance of all the aggregation schemes significantly decreases. With Non-IID data, some of these aggregation schemes fail even in the complete absence of Byzantine clients. We conclude that the robustness of all the aggregation schemes is limited, highlighting the need for new defense strategies, in particular for Non-IID datasets.

Applications of cluster-based transfer learning in image and localization tasks

Transfer learning can address the issue of insufficient labels in machine learning. Using knowledge in a labeled domain (source domain) can assist in acquiring and learning knowledge in a domain (target domain) that lacks some or all labels. In this paper, we propose a new cluster-based semi-supervised transfer learning (CBSSTL) under a new assumption that samples in the target domain are unlabeled but contain cluster information. Furthermore, we propose a new transfer learning framework and a method for fine-tuning parameters. We tested and compared the proposed method with other unsupervised and semi-supervised transfer learning methods on well-known image datasets. The experimental results demonstrate the effectiveness of the proposed method. Additionally, we created a localization dataset for transfer learning. Finally, we tested and analyzed the proposed method on this dataset. Its particularly challenging nature makes it difficult for our method to work effectively.

Flexible Vertical Federated Learning With Heterogeneous Parties.

We propose flexible vertical federated learning (Flex-VFL), a distributed machine algorithm that trains a smooth, nonconvex function in a distributed system with vertically partitioned data. We consider a system with several parties that wish to collaboratively learn a global function. Each party holds a local dataset; the datasets have different features but share the same sample ID space. The parties are heterogeneous in nature: the parties' operating speeds, local model architectures, and optimizers may be different from one another and, further, they may change over time. To train a global model in such a system, Flex-VFL utilizes a form of parallel block coordinate descent (P-BCD), where parties train a partition of the global model via stochastic coordinate descent. We provide theoretical convergence analysis for Flex-VFL and show that the convergence rate is constrained by the party speeds and local optimizer parameters. We apply this analysis and extend our algorithm to adapt party learning rates in response to changing speeds and local optimizer parameters. Finally, we compare the convergence time of Flex-VFL against synchronous and asynchronous VFL algorithms, as well as illustrate the effectiveness of our adaptive extension.

Exploring parental factors influencing low birth weight on the 2022 CDC natality dataset

Background and aimsLow birth weight (LBW), known as the condition of a newborn weighing less than 2500 g, is a growing concern in the United States (US). Previous studies have identified several contributing factors, but many have analyzed these variables in isolation, limiting their ability to capture the combined influence of multiple factors. Moreover, past research has predominantly focused on maternal health, demographics, and socioeconomic conditions, often neglecting paternal factors such as age, educational level, and ethnicity. Additionally, most studies have utilized localized datasets, which may not reflect the diversity of the US population. To address these gaps, this study leverages machine learning to analyze the 2022 Centers for Disease Control and Prevention’s National Natality Dataset, identifying the most significant factors contributing to LBW across the US.MethodsWe combined anthropometric, socioeconomic, maternal, and paternal factors to train logistic regression, random forest, XGBoost, conditional inference tree, and attention mechanism models to predict LBW and normal birth weight (NBW) outcomes. These models were interpreted using odds ratio analysis, feature importance, partial dependence plots (PDP), and Shapley Additive Explanations (SHAP) to identify the factors most strongly associated with LBW.ResultsAcross all five models, the most consistently associated factors with birth weight were maternal height, pre-pregnancy weight, weight gain during pregnancy, and parental ethnicity. Other pregnancy-related factors, such as prenatal visits and avoiding smoking, also significantly influenced birth weight.ConclusionThe relevance of maternal anthropometric factors, pregnancy weight gain, and parental ethnicity can help explain the current differences in LBW and NBW rates among various ethnic groups in the US. Ethnicities with shorter average statures, such as Asians and Hispanics, are more likely to have newborns below the World Health Organization’s 2500-gram threshold. Additionally, ethnic groups with historical challenges in accessing nutrition and perinatal care face a higher risk of delivering LBW infants.

KL-FedDis: A federated learning approach with distribution information sharing using Kullback-Leibler divergence for non-IID data

Data Heterogeneity or Non-IID (non-independent and identically distributed) data identification is one of the prominent challenges in Federated Learning (FL). In Non-IID data, clients have their own local data, which may not be independently and identically distributed. This arises because clients involved in federated learning typically have their own unique, local datasets that vary significantly due to factors like geographical location, user behaviors, or specific contexts. Model divergence is another critical challenge where the local models trained on different clients, data may diverge significantly but making it difficult for the global model to converge. To identify the non-IID data, few federated learning models have been introduced as FedDis, FedProx and FedAvg, but their accuracy is too low. To address the clients Non-IID data along with ensuring privacy, federated learning emerged with appropriate distribution mechanism is an effective solution. In this paper, a modified FedDis learning method called KL-FedDis is proposed, which incorporates Kullback-Leibler (KL) divergence as the regularization technique. KL-FedDis improves accuracy and computation time over the FedDis and FedAvg technique by successfully maintaining the distribution information and encouraging improved collaboration among the local models by utilizing KL divergence.

A Privacy Preserving System for Movie Recommendations Using Federated Learning

Recommender systems have become ubiquitous in the past years. They solve the tyranny of choice problem faced by many users, and are utilized by many online businesses to drive engagement and sales. Besides other criticisms, like creating filter bubbles within social networks, recommender systems are often reproved for collecting considerable amounts of personal data. However, to personalize recommendations, personal information is fundamentally required. A recent distributed learning scheme called federated learning has made it possible to learn from personal user data without its central collection. Consequently, we present a recommender system for movie recommendations, which provides privacy and thus trustworthiness on multiple levels: First and foremost, it is trained using federated learning and thus, by its very nature, privacy-preserving, while still enabling users to benefit from global insights. Furthermore, a novel federated learning scheme, called FedQ, is employed, which not only addresses the problem of non-i.i.d.-ness and small local datasets, but also prevents input data reconstruction attacks by aggregating client updates early. Finally, to reduce the communication overhead, compression is applied, which significantly compresses the exchanged neural network parametrizations to a fraction of their original size. We conjecture that this may also improve data privacy through its lossy quantization stage.

New Continual Federated Learning System for Intrusion Detection in SDN‐Based Edge Computing

ABSTRACTSoftware Defined Networking (SDN) is an open network approach that has been proposed to address some of the main problems with traditional networks. However, SDN faces cybersecurity issues. To provide a network defense against attacks, an Intrusion Detection System (IDS) needs to be updated and included into the SDN architecture on a regular basis. Machine learning methods have proved effective in detecting intrusions in SDN. Moreover, these techniques pose the problem of significant computational overload and the absence of regular updates when new cyber‐attacks appear. To address these issues, we propose a new SDN‐based cloud intrusion detection system called Continual Federated Learning (CFL). In CFL, we modify the classical federated learning process by granting a more important and dynamic role to each participating client. On the one hand, it can trigger this process whenever a new type of intrusion is detected. On the other hand, once the new model has been identified, the customer can decide whether or not to deploy it in his network. In addition, to verify the accuracy of the CFL system, we have formally specified it by a communication protocol. This specification organizes the exchanges between the different communicating entities involved in the CFL. To verify the accuracy of this specification, we described it using the PROMELA language and checked with the associated SPIN tool. On the experimental side, we deployed this specification of the CFL system in an SDN computing environment. We defined different scenarios, and we proposed that each client decides locally to deploy or not the newly obtained intrusion detection model. The decision is based on a modified metric where we integrate the severity of the intrusions. Experimental results using private local datasets show that the proposed CFL system can efficiently and accurately detect new types of intrusions while preserving client confidentiality. Thus, it can be considered a promising system for SDN‐based edge computing.

Integrating Remote Photoplethysmography and Machine Learning on Multimodal Dataset for Noninvasive Heart Rate Monitoring.

Non-contact heart monitoring is crucial in advancing telemedicine, fitness tracking, and mass screening. Remote photoplethysmography (rPPG) is a non-contact technique to obtain information about heart pulse by analyzing the changes in the light intensity reflected or absorbed by the skin during the blood circulation cycle. However, this technique is sensitive to environmental lightning and different skin pigmentation, resulting in unreliable results. This research presents a multimodal approach to non-contact heart rate estimation by combining facial video and physical attributes, including age, gender, weight, height, and body mass index (BMI). For this purpose, we collected local datasets from 60 individuals containing a 1 min facial video and physical attributes such as age, gender, weight, and height, and we derived the BMI variable from the weight and height. We compare the performance of two machine learning models, support vector regression (SVR) and random forest regression on the multimodal dataset. The experimental results demonstrate that incorporating a multimodal approach enhances model performance, with the random forest model achieving superior results, yielding a mean absolute error (MAE) of 3.057 bpm, a root mean squared error (RMSE) of 10.532 bpm, and a mean absolute percentage error (MAPE) of 4.2% that outperforms the state-of-the-art rPPG methods. These findings highlight the potential for interpretable, non-contact, real-time heart rate measurement systems to contribute effectively to applications in telemedicine and mass screening.