Preprocessing Stage Research Articles

Predicting game-induced emotions using EEG, data mining and machine learning

BackgroundEmotion is a complex phenomenon that greatly affects human behavior and thinking in daily life. Electroencephalography (EEG), one of the human physiological signals, has been emphasized by most researchers in emotion recognition as its specific properties are closely associated with human emotion. However, the number of human emotion recognition studies using computer games as stimuli is still insufficient as there were no relevant publicly available datasets provided in the past decades. Most of the recent studies using the Gameemo public dataset have not clarified the relationship between the EEG signal’s changes and the emotion elicited using computer games. Thus, this paper is proposed to introduce the use of data mining techniques in investigating the relationships between the frequency changes of EEG signals and the human emotion elicited when playing different kinds of computer games. The data acquisition stage, data pre-processing, data annotation and feature extraction stage were designed and conducted in this paper to obtain and extract the EEG features from the Gameemo dataset. The cross-subject and subject-based experiments were conducted to evaluate the classifiers’ performance. The top 10 association rules generated by the RCAR classifier will be examined to determine the possible relationship between the EEG signal's frequency changes and game-induced emotions.ResultsThe RCAR classifier constructed for cross-subject experiment achieved highest accuracy, precision, recall and F1-score evaluated with over 90% in classifying the HAPV, HANV and LANV game-induced emotions. The 20 experiment cases’ results from subject-based experiments supported that the SVM classifier could accurately classify the 4 emotion states with a kappa value over 0.62, demonstrating the SVM-based algorithm’s capabilities in precisely determining the emotion label for each participant’s EEG features’ instance.ConclusionThe findings in this study fill the existing gap of game-induced emotion recognition field by providing an in-depth evaluation on the ruleset algorithm’s performance and feasibility of applying the generated rules on the game-induced EEG data for justifying the emotional state prediction result.

#176 System to detect vascular calcification through convolutional neural networks

Abstract Background and Aims Cardiovascular disease is the major cause of death in patients with chronic kidney disease (CKD). It is increasingly believed that the presence of vascular calcification (VC) increases the risk of mortality due to cardiovascular events. VC is very common in CKD, especially in hemodialysis patient. To improve the life expectancy of patients, it is important to detect VC in the early stages of the disease. There are several ways to discover VC such as computed tomography (CT) scans, echocardiography, and X-Ray among others. However, one of the best options is the usage of X-ray images because is an affordable and simpler process in comparison to using CT scans or echocardiography. However, the problem with X-ray images is the difficulty of finding a well-trained professional that can detect the VC after analyzing the images which leads to possible misleading diagnoses. The goal of this study is to develop a prototype that detects Vascular Calcification from X-ray images in an automated way using the power of Convolutional Neural Networks (CNN). Method To train a CNN to detect VC from lumbosacral spine X-ray images, a set of 700 DICOM files complemented with 236 JPEG images was available. After getting the images, they had to be anonymized, and data, such as name and age was removed from each image. Then, the images were divided into three datasets: Train, Test, and Validation. Given that the dataset was not balanced between the pictures with VC (Positive) and those with no VC (Negative), it had to be reduced (Table 1). The next step was implemented to clean the images and prepare them to feed the CNN. An example of preprocessing is that some X-ray images were of the entire body. For the analysis, the VC detection was assessed at the anterior and the posterior walls of the abdominal aorta adjacent to vertebrae L1-L4. In addition to that, the DICOM images were converted to JPEG. After having the dataset prepared, a CNN architecture had to be selected. To achieve it, a revision of modern CNN architectures was performed. Two items were considered when selecting these: the performance on the ImageNet Large Scale Visual Recognition Challenge (ILSVRC) and if there was an available pre-trained model. With that criteria, pre-trained EfficientNet-B0 and ResNet50 models were used to build the prototype due to the limited dataset. Results After having the dataset prepared, the architectures selected, and the pre-trained models, the Training, Validation and Test stages were executed. For Training, with EfficientNet-B0, the Accuracy was 86.9%, while for Validation, the Accuracy was 57.4% (Fig. 1). Using the RestNet50 model, the precision was 90.506% during Training, while for Validation, it was 54.8% (Fig. 2). Finally, during the Testing stage, ResNet50 had a better performance, because EfficientNetB0 detected 80% for the negative class and 60% for the positive class. After the CNN models were trained, an additional validation stage was performed to compare the model results with the analysis made by a radiologist and a nephrologist. This comparison determined that the model performed better with images having VC, while it was less accurate with images having no presence of VC. This contrast could have different reasons, one being possible false-positive images. Conclusion This study created a CNN model using pre-trained models on top of modern architectures such as EfficientNet-B0 and ResNet50. Tools were used to prepare the images as well as to train, validate and test the model. The Transfer Learning helped to build a model faster, giving high Accuracy rates with only a few images due to the limited dataset. However, the model could give better results if a wider and balanced set of images is available. Also, a better pre-processing stage can be executed to reduce false-positive images. For future work, this model can be used as the core of a system that not only accurately detects Vascular Calcification from X-Ray images, but also can determine the level of calcification using measures such as the Kauppila Index or Adragao.

An automated geometry generation procedure for CWE applications in OpenFOAM: the case study of a thunderstorm downburst moving over Genoa city

AbstractSince a decade, Computational Fluid Dynamics (CFD) simulations for Wind Engineering (WE) on large-scale environments became a norm. Meso-$$\gamma$$ γ scales and microscale can be characterized by a large number of structures as well as complex terrain topographies. CFD simulations over such an area would commonly imply the involvement of a time-consuming geometry preparation in the pre-processing stage. This is also the innovative goal of the present paper for which an automated geometry generation procedure for Computational Wind Engineering (CWE) applications is developed in open-source environments. This paper further expands on the unsteady Reynolds-averaged Navier–Stokes (URANS) simulation of a thunderstorm downburst immersed in a background Atmospheric Boundary Layer (ABL) wind moving over Genoa city, Italy. Recommendations for geometry and grid generation of the present case study are also proposed. The study considers two simultaneous approaching wind conditions, a background ABL wind and a moving downburst (DB), the latter modeled as an impinging jet. The simulated results are compared with available full-scale LiDAR profiler data in terms of the vertical profiles of outflow velocity, showing relatively good agreement in terms of magnitude and profile shape.

DTDO: Driving Training Development Optimization enabled deep learning approach for brain tumour classification using MRI

ABSTRACT A brain tumour is an abnormal mass of tissue. Brain tumours vary in size, from tiny to large. Moreover, they display variations in location, shape, and size, which add complexity to their detection. The accurate delineation of tumour regions poses a challenge due to their irregular boundaries. In this research, these issues are overcome by introducing the DTDO-ZFNet for detection of brain tumour. The input Magnetic Resonance Imaging (MRI) image is fed to the pre-processing stage. Tumour areas are segmented by utilizing SegNet in which the factors of SegNet are biased using DTDO. The image augmentation is carried out using eminent techniques, such as geometric transformation and colour space transformation. Here, features such as GIST descriptor, PCA-NGIST, statistical feature and Haralick features, SLBT feature, and CNN features are extricated. Finally, the categorization of the tumour is accomplished based on ZFNet, which is trained by utilizing DTDO. The devised DTDO is a consolidation of DTBO and CDDO. The comparison of proposed DTDO-ZFNet with the existing methods, which results in highest accuracy of 0.944, a positive predictive value (PPV) of 0.936, a true positive rate (TPR) of 0.939, a negative predictive value (NPV) of 0.937, and a minimal false-negative rate (FNR) of 0.061%.

Enhance micro-Doppler signatures-based human activity classification accuracy of FMCW radar using the threshold method

Nowadays, radar-based human activity classification is being widely adopted in healthcare systems due to its benefits in terms of personal privacy compliance, non-contact sensing, and being unaffected by weather conditions. This study proposes a threshold method in the pre-processing stage to improve human activity classification accuracy by determining the region of meaningful information (RMI) on the spectrogram. Initially, a mask function, which is created by a certain threshold value, is applied to the input spectrogram to highlight the RMI from the micro-Doppler (m-D) signatures. Only the highlighted RMI on the spectrogram is retained as input to the classifiers. Then, five Convolutional Neural Networks (CNNs) of varying complexity are employed to extract features, identify activities, and assess the effectiveness of the suggested approach. The experimental results demonstrate that the suggested approach has enhanced classification accuracy by up to 11% when compared to the original unprocessed dataset.

Feature selection method for banknote dirtiness recognition based on mathematical functions driven slime mould algorithm

In most data dimensionality reduction tasks, feature selection (FS) is an essential preprocessing stage. The slime mould algorithm (SMA) is a novel metaheuristic optimization algorithm. This paper proposes a feature selection method for banknote dirtiness recognition based on mathematical functions driven slime mould algorithm. This method is applied to the feature selection of banknote dirtiness recognition, and the Lévy flight operator is used to replace the control parameter vc in SMA, and a series of mathematical functions are added to replace the control parameter vb in the slime mould algorithm to enhance the global search capability. Test results show that this method achieves the highest accuracy of 76.85 % in recognizing the banknote dirtiness. Additionally, by applying the same method to feature selection and recognition of dirtiness for the left middle white area images of banknotes, the highest accuracy reaches 74.99 %. This suggests that the left middle white area can essentially replace the whole banknote. To verify the performance of the proposed feature selection method, comparative tests with four other optimization algorithms are carried out, and the results demonstrate that three improvement strategies can effectively improve the performance of SMA while maintaining a balance between exploration and exploitation. Tests have revealed that these three algorithms have significant advantages over the past SMA in terms of recognition accuracy and convergence speed when it comes to identifying the dirtiness of banknotes. Therefore, this method is considered to be one of the most effective methods to solve the feature selection problem of banknote dirtiness.

Evaluating distributed-learning on real-world obstetrics data: comparing distributed, centralized and local models

This study focused on comparing distributed learning models with centralized and local models, assessing their efficacy in predicting specific delivery and patient-related outcomes in obstetrics using real-world data. The predictions focus on key moments in the obstetric care process, including discharge and various stages of hospitalization. Our analysis: using 6 different machine learning methods like Decision Trees, Bayesian methods, Stochastic Gradient Descent, K-nearest neighbors, AdaBoost, and Multi-layer Perceptron and 19 different variables with various distributions and types, revealed that distributed models were at least equal, and often superior, to centralized versions and local versions. We also describe thoroughly the preprocessing stage in order to help others implement this method in real-world scenarios. The preprocessing steps included cleaning and harmonizing missing values, handling missing data and encoding categorical variables with multisite logic. Even though the type of machine learning model and the distribution of the outcome variable can impact the result, we reached results of 66% being superior to the centralized and local counterpart and 77% being better than the centralized with AdaBoost. Our experiments also shed light in the preprocessing steps required to implement distributed models in a real-world scenario. Our results advocate for distributed learning as a promising tool for applying machine learning in clinical settings, particularly when privacy and data security are paramount, thus offering a robust solution for privacy-concerned clinical applications.

Lane detection using hough transformation and Yolov8

Autonomous vehicles necessitate the integration of advanced technologies such as computer vision and deep learning to comprehend and navigate their surroundings. A crucial yet challenging component of this integration is the accurate detection of lanes, which can be influenced by a multitude of varying lane characteristics and conditions. This research undertakes a comparative analysis of lane detection methodologies, explicitly focusing on traditional image processing techniques and Convolutional Neural Networks (CNNs). The evaluation utilized a sample of 500 images from the CULane dataset, which encompasses a diverse range of traffic scenarios. Initially, a method incorporating Gaussian blurring, Canny edge detection, and Hough line transformation was examined. Despite its efficiency, operating at 30 frames per second, this approach exhibited a high error rate (average Mean Squared Error (MSE) of 0.537), which is attributable to the loss of critical image details during the preprocessing stage. Subsequently, the performance of a fine-tuned YOLOv8 model, trained on a reformatted version of the CULane dataset was assessed. The combination of object detection and subsequent Hough transformation yielded high accuracy, demonstrating the model’s ability to learn and identify relevant lane features. The deep CNNs demonstrated superior performance over classical image processing techniques in terms of lane detection accuracy, thereby underscoring their potential applicability within the realm of autonomous vehicle technology

An enhanced Convolutional Neural Network (CNN) model for the detection of lung cancer using X-Ray image

Lung Cancer is a life-threatening disease which can be diagnosed by Medical Imaging such as X-Ray, MRI, CT Scan etc. This research presented an enhanced model using Convolutional Neural Network (CNN) to detect lung cancer using X-Ray image. Medical image processing relies heavily on the diagnosis of lung cancer images. It aids doctors in determining the correct diagnosis and management. For many patients, lung cancer ranks among the mostdeadly diseases. Many lives can be saved if cancerous growth is diagnosed early. The purported model was predominantly built on Convolutional Neural Network (CNN) architecture and the model was built with enhanced features such as Image Enhancement, Segmenting ROI (Region of Interest), Features Extraction and Nodule Classification. In preprocessing stage, the AMF (Adaptive Median Filter) filtering method was applied to eliminate noise in X-Ray image of the dataset, and quality of X-Ray image was improved with the support of CLAHE (Contrast Limited Adaptive Histogram Equalization). Secondly, K-means Clustering algorithm was used to extract the relevant feature or Region of Interest (ROI) of the lung field automatically i.e. the model was effectively trained to identify and crop the exact location of the lung field automatically. The model was able to classify the cancer nodule as either Cancerous or Non-Cancerous. The framework worked on C# platform, and used EMGU for detection of the tumour in the lung xray image. Experimental result showed that the developed system was able to detect Lung Cancer with 90.77% accuracy, 86.65% precision and 95.31% Recall/Sensitivity.

TemporalAugmenter: An Ensemble Recurrent Based Deep Learning Approach for Signal Classification

Ensemble modeling has been widely used to solve complex problems as it helps to improve overall performance and generalization. In this paper, we propose a novel TemporalAugmenter approach based on ensemble modeling for augmenting the temporal information capturing for long-term and short-term dependencies in data integration of two variations of recurrent neural networks in two learning streams to obtain the maximum possible temporal extraction. Thus, the proposed model augments the extraction of temporal dependencies. In addition, the proposed approach reduces the preprocessing and prior stages of feature extraction, which reduces the required energy to process the models built upon the proposed TemporalAugmenter approach, contributing towards green AI. Moreover, the proposed model can be simply integrated into various domains including industrial, medical, and human-computer interaction applications. Our proposed approach empirically evaluated the speech emotion recognition, electrocardiogram signal, and signal quality examination tasks as three different signals with varying complexity and different temporal dependency features.

A Foam Line Position Detection Algorithm for A/O Pool Based on YOLOv5

During the biochemical pretreatment process of leachate in urban landfill sites, if the foam in the A/O pool is not promptly addressed, it can lead to overflow, posing hazards to the surrounding environment and personnel. Therefore, a real-time foam line detection algorithm based on YOLOv5x was proposed, which enhances feature information and improves anchor box regression prediction to accurately detect the position of foam lines. Firstly, in the preprocessing stage, employing a rectangular box to simultaneously label the foam line and the edge of the A/O pool within the same region, enhances the feature information of the foam line. Then, the C3NAM module was proposed, which applies weight sparse penalties to attention modules in the feature extraction section, to enhance the capability of extracting foam line features. Subsequently, a B-SPPCSPC module was proposed to enhance the fusion of shallow and deep feature information, addressing the issue of susceptibility to background interference during foam line detection. Next, the Focal_EIOU was introduced to ameliorate the issue of class imbalance in detection, providing more accurate bounding box predictions. Lastly, optimizing the detection layer scale improves the detection performance for smaller targets. The experimental results demonstrate that the accuracy of this algorithm reaches 98.9%, and the recall reaches 88.1%, with a detection frame rate of 26.2 frames per second, which can meet the actual detection requirements of real-world application scenarios.

Named Entity Recognition Using Conditional Random Fields for Flood Detection In Gerbang Kertosusila Based Twitter Data

The national strategic area Gerbang Kertosusila East Java should be aware of floods. One of the existing efforts is to place flood sensors at several flood-prone points. However, that way is constrained by the need for more equipment to handle the many needy areas. So it is necessary to develop technology for the dissemination of flood information. Dissemination of flood information was quickly obtained from social media Twitter. One way is to use Twitter's text data source for a Named Entity Recognition model to help detect flood events and their locations. The Named Entity Recognition (NER) model was constructed using the Conditional Random Fields (CRFs) method to achieve research objectives. This research adds slang word handling at the preprocessing stage to improve model performance and the use of the BIO format in the labeling process and POS Tagging in the Feature Extraction process. Evaluation results with five Kfolds, 80% training data, and 20% test data show that the NER CRFs model performs excellently with a Precision of 0.981, Recall of 0.926, and f-measure of 0.950 so that these results can help the community and government regarding the information on the distribution of floods.

Effect of Phase Shifting on Real-Time Detection and Classification of Power Quality Disturbances

Power quality improvement and Power quality disturbance (PQD) detection are two significant concerns that must be addressed to ensure an efficient power distribution within the utility grid. When the process to analyze PQD is migrated to real-time platforms, the possible occurrence of a phase mismatch can affect the algorithm’s accuracy; this paper evaluates phase shifting as an additional stage in signal acquisition for detecting and classifying eight types of single power quality disturbances. According to their mathematical models, a set of disturbances was generated using an arbitrary waveform generator BK Precision 4064. The acquisition, detection, and classification stages were embedded into a BeagleBone Black. The detection stage was performed using multiresolution analysis. The feature vectors of the acquired signals were obtained from the combination of Shannon entropy and log-energy entropy. For classification purposes, four types of classifiers were trained: multilayer perceptron, K-nearest neighbors, probabilistic neural network, and decision tree. The results show that incorporating a phase-shifting stage as a preprocessing stage significantly improves the classification accuracy in all cases.

A popular topic detection method based on microblog images and short text information

Popular topic detection is a topic identification by the information of documents posted by users in social networking platforms. In a large body of research literature, most popular topic detection methods identify the distribution of unknown topics by integrating information from documents based on social networking platforms. However, among these popular topic detection methods, most of them have a low accuracy in topic detection due to the short text content and the abundance of useless punctuation marks and emoticons. Image information in short texts has also been overlooked, while this information may contain the real topic matter of the user's posted content. In order to solve the above problems and improve the quality of topic detection, this paper proposes a popular topic detection method based on microblog images and short text information. The method uses an image description model to obtain more information about short texts, identifies hot words by a new word discovery algorithm in the preprocessing stage, and uses a PTM model to improve the quality and effectiveness of topic detection during topic detection and aggregation. The experimental results show that the topic detection method in this paper improves the values of evaluation indicators compared with the other three topic detection methods. In conclusion, the popular topic detection method proposed in this paper can improve the performance of topic detection by integrating microblog images and short text information, and outperforms other topic detection methods selected in this paper.

Computer vision classification detection of chicken parts based on optimized Swin-Transformer

ABSTRACT In order to achieve real-time classification and detection of various chicken parts, this study introduces an optimized Swin-Transformer method for the classification and detection of multiple chicken parts. It initially leverages the Transformer’s self-attention structure to capture more comprehensive high-level visual semantic information from chicken part images. The image enhancement technique was applied to the image in the preprocessing stage to enhance the feature information of the image, and the migration learning method was used to train and optimize the Swin-Transformer model on the enhanced chicken parts dataset for classification and detection of chicken parts. Furthermore, this model was compared to four commonly used models in object target detection tasks: YOLOV3-Darknet53, YOLOV3-MobileNetv3, SSD-MobileNetv3, and SSD-VGG16. The results indicated that the Swin-Transformer model outperforms these models with a higher mAP value by 1.62%, 2.13%, 5.26%, and 4.48%, accompanied by a reduction in detection time by 16.18 ms, 5.08 ms, 9.38 ms, and 23.48 ms, respectively. The method of this study fulfills the production line requirements while exhibiting superior performance and greater robustness compared to existing conventional methods.

Zero-One Matrix on Tudung Saji Pattern via Image Processing

Pattern recognition is a method for classifying or describing the quantitative measurement of features of any object, data, or source. It involves any input that depends on the context of the study, such as biometric data, sensor data, signals, images, and so on. Prior research had been conducted on tudung saji, with a particular emphasis placed on ethnographic research that was related to group theory. In order to advance the study of the relationships between tudung saji and group theory, additional analysis is undertaken to examine the pattern of tudung saji within certain finite matrix rings. This analysis involves studying the zero-one matrix that is obtained through image processing. It is important to understand the process of pattern recognition to produce the output as a classification result from the feature extraction of the input. The study focuses on the stages of image preprocessing, which include scaling, grayscale conversion, and thresholding. The binarization image is classified through feature extraction using the image of tudung saji patterns. The binarization image produces a binary value. The threshold value is calculated by using the Otsu’s Method. This method evaluates the binary value by comparing the threshold value with the grayscale value. The result assists researchers in the subsequent classification of tudung saji patterns into group theory by means of analysing the finite zero-one matrix. Hence, the classification of tudung saji patterns can be achieved by utilising finite zero-one matrices, which allows for a more comprehensive analysis of these patterns from the standpoint of group theory. In light of these considerations, a mathematical analysis is conducted on the patterns of tudung saji under the field of pattern recognition. In this analysis, the image of the tudung saji pattern serves as the input for data processing, resulting in the production of a finite zero-one matrix.

Tracking and co-location of global point clouds for large-area indoor environments

Extended reality (XR) experiences are on the verge of becoming widely adopted in diverse application domains. An essential part of the technology is accurate tracking and localization of the headset to create an immersive experience. A subset of the applications require perfect co-location between the real and the virtual world, where virtual objects are aligned with real-world counterparts. Current headsets support co-location for small areas, but suffer from drift when scaling up to larger ones such as buildings or factories. This paper proposes tools and solutions for this challenge by splitting up the simultaneous localization and mapping (SLAM) into separate mapping and localization stages. In the pre-processing stage, a feature map is built for the entire tracking area. A global optimizer is applied to correct the deformations caused by drift, guided by a sparse set of ground truth markers in the point cloud of a laser scan. Optionally, further refinement is applied by matching features between the ground truth keyframe images and their rendered-out SLAM estimates of the point cloud. In the second, real-time stage, the rectified feature map is used to perform localization and sensor fusion between the global tracking and the headset. The results show that the approach achieves robust co-location between the virtual and the real 3D environment for large and complex tracking environments.

Enhanced Fault Detection in Bearings Using Machine Learning and Raw Accelerometer Data: A Case Study Using the Case Western Reserve University Dataset

This study introduces a novel approach for fault classification in bearing components utilizing raw accelerometer data. By employing various neural network models, including deep learning architectures, we bypass the traditional preprocessing and feature-extraction stages, streamlining the classification process. Utilizing the Case Western Reserve University (CWRU) bearing dataset, our methodology demonstrates remarkable accuracy, particularly in deep learning networks such as the three variant convolutional neural networks (CNNs), achieving above 98% accuracy across various loading levels, establishing a new benchmark in fault-detection efficiency. Notably, data exploration through principal component analysis (PCA) and t-distributed stochastic neighbor embedding (t-SNE) provided valuable insights into feature relationships and patterns, aiding in effective fault detection. This research not only proves the efficacy of neural network classifiers in handling raw data but also opens avenues for more straightforward yet effective diagnostic methods in machinery health monitoring. These findings suggest significant potential for real-world applications, offering a faster yet reliable alternative to conventional fault-classification techniques.

Abstract PO4-07-03: Machine learning breast cancer risk prediction using sequential past mammograms - a pilot study

Abstract Background and Aim Some studies have reported the use of mammogram AI deep learning algorithms to accurately predict the risk of future breast cancer development in women. In these studies, random and unrelated mammograms were independently used for training of the AI model. We propose a novel deep learning system using sequential past mammograms instead of random studies and aimed to determine if it can further improve risk prediction in a pilot study. Methods Pre-training of the backbone of a Swin-transformer machine learning model was performed by applying region-of-interest lesional masks of benign and malignant on a training set of 1644 mammograms derived from the open-access Curated Breast Imaging Subset of Digital Database for Screening Mammography dataset. To further develop the model, deep learning training and validation were performed on the Cohort of Screen-age Women (CSAW) open-access screening mammogram set which was derived from the long term data of 873 women who developed breast cancer for the first time and 7850 non-cancer women. We then compared the diagnostic accuracy of a model using sequential past mammograms for training against the standard, non-sequential model. For the sequential model, imaging segmentation of the potential cancer site on mammography was performed on all past, pre-cancer mammograms obtained within 60 to 2555 days of cancer diagnosis in the pre-processing stage for breast cancer cases. AI training was then carried out by combining the pre-cancer mammograms belonging to the same woman in a temporal sequence using a transformer encoder. The learning rate started from 1e-6 and decayed 10 times every 10 epochs. For the non-sequential model, we removed the temporal encoder and used random, unrelated mammogram studies for training. Validation was performed on a dataset consisting of 75 pre-cancer mammograms from breast cancer cases and 69 mammograms from non-cancer cases. Results Sensitivity, specificity and overall accuracy on the validation set were 77.3% (58/75), 68.1% (47/69) and (105/144) 72.9% respectively for the sequential model compared to 69.3% (52/75), 69.6% (48/69) and 69.4% (100/144) respectively for the non-sequential model. Conclusion The study showed that deep learning breast cancer risk prediction can be further improved by using sequential past mammograms instead of random mammograms for AI model training. This can also potentially enhance other AI risk prediction models that employ combined mammogram and traditional breast cancer clinical risk factors. A larger validation study will be useful. Contact information: email Lester Leong at lester.leong.c.h@singhealth.com.sg ROC curve for breast cancer risk prediction using AI deep learning training with sequential past mammograms. Citation Format: Lester Leong, Mingjie Xu, Weimin Huang, Erli Zhang, Engracia Loh, Sze Yiun Teo, Geok Hoon Lim, Veronique Kiak Mien Tan, Yirong Sim, Fredrik Strand, Ryan Tan. Machine learning breast cancer risk prediction using sequential past mammograms - a pilot study [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-07-03.

Lightweight JPEG image steganalysis using dilated blind-spot network

The primary goal of image steganalysis is to detect subtle steganographic signals hidden within digital images. An essential aspect of digital image steganalysis involves the extraction of meaningful steganographic signal features. However, existing JPEG steganalysis networks have not delved deep enough into the research on extracting steganographic noise signals, especially in the preprocessing stage of the model. In this paper, we propose a novel approach called the Dilated Blind-Spot Steganalysis Network (DBS2Net), specifically designed for JPEG steganalysis. DBS2Net incorporates the dilated blind-spot convolutional network into the preprocessing stage to extract noise residuals, thereby enhancing detection accuracy. The blind-spot network estimates central outputs based on adjacent regions to effectively suppress semantic information and uses dilated convolutions to expand the receptive field. Additionally, we design the transition block to further enhance noise signals by mapping noise residuals extracted by the blind-spot network. In feature extraction, depth-wise separable convolutions are used to minimize the parameter count, while dual-view pooling layers compress features, resulting in improved detection performance. Experimental results validate the efficacy of this approach, and comprehensive analyses of JPEG steganographic noise support its success. In particular, DBS2Net outperforms state-of-the-art steganalysis methods, achieving superior detection accuracy with minimal model complexity.