Confusion Matrix Research Articles

Evaluating the Performance and Challenges of Machine Learning Models in Network Anomaly Detection

The application of machine learning algorithms for anomaly detection in network traffic data is examined in this study. Using a collection of network flow records that includes attributes such as IP addresses, ports, protocols, and timestamps, the study makes use of correlation heatmaps, box plots, and data visualization to identify trends in numerical characteristics. After preprocessing, which includes timestamp conversion to Unix format, three machine learning models Support Vector Machine (SVM), Gaussian Naive Bayes, and Random Forest are used for anomaly identification. The Random Forest Classifier outperforms SVM and Naive Bayes classifiers with better precision and recall for anomaly diagnosis, achieving an accuracy of 87%. Confusion matrices and classification reports are used to evaluate the models, and they show that the Random Forest Classifier performs better than the other models in identifying abnormalities in network traffic. These results provide significant value to the field of cybersecurity by highlighting the effectiveness of machine learning models specifically, the Random Forest Classifier in boosting anomaly detection capacities for network environment security.

Personalized venlafaxine dose prediction using artificial intelligence technology: a retrospective analysis based on real-world data.

Venlafaxine dose regimens vary considerably between individuals, requiring personalized dosing. This study aimed to identify dose-related influencing factors of venlafaxine through real-world data analysis and to construct a personalized dose model using advanced artificial intelligence techniques. We conducted a retrospective study on patients with depression treated with venlafaxine. Significant variables were selected through a univariate analysis. Subsequently, the predictive performance of seven models (XGBoost, LightGBM, CatBoost, GBDT, ANN, TabNet, and DT) was compared. The algorithm that demonstrated optimal performance was chosen to establish the dose prediction model. Model validation used confusion matrices and ROC analysis. Additionally, a dose subgroup analysis was conducted. A total of 298 patients were included. TabNet was selected to establish the venlafaxine dose prediction model, which exhibited the highest performance with an accuracy of 0.80. The analysis identified seven crucial variables correlated with venlafaxine daily dose, including blood venlafaxine concentration, total protein, lymphocytes, age, globulin, cholinesterase, and blood platelet count. The area under the curve (AUC) for predicting venlafaxine doses of 75mg, 150mg, and 225mg were 0.90, 0.85, and 0.90, respectively. We successfully developed a TabNet model to predict venlafaxine doses using real-world data. This model demonstrated substantial predictive accuracy, offering a personalized dosing regimen for venlafaxine. These findings provide valuable guidance for the clinical use of the drug.

Using satellite image data to identify rice varieties through linear spectral unmixing method (case study: Karangjati Sub District, Ngawi Regency)

<p>Remote sensing technology has increasingly emerged as a potent tool for precision agriculture, particularly in facilitating the mapping and monitoring of crops on a large scale. An application of this technology is the identification of different types of rice by analyzing the pixels acquired in satellite images. Regrettably, the pixels in the image have been mixed from different recorded items. Therefore, they have the potential to influence the outcome of the identification. An effective approach to address this problem is to employ the linear spectral unmixing (LSU) technique. The LSU approach quantifies the ratio of pure objects in every pixel of an image by utilizing the spectral value associated with the endmember of the rice variety. The investigation was carried out in the Karangjati District during the generative stage (70 ± DAP) of the rice planting season. The data indicates that the dominant variety is Inpari 32 HDB. The data validation tests, which involved the use of a confusion matrix and Kappa analysis, resulted in an overall accuracy rate of 85.48% and a Kappa analysis score of 70.6%.<strong></strong></p>

Comparative evaluation of the diagnostic and prognostic performance of CNSide™ versus standard cytology for leptomeningeal disease

Abstract Background This retrospective study compares real-world performance of cerebrospinal fluid (CSF) CNSide™ versus cytology in leptomeningeal disease (LMD). Methods: Consecutive patients with suspected LMD who underwent lumbar punctures for CSF cytology and CNSide™ from January 2020 to December 2022 were reviewed. LMD was classified by EANO criteria. Descriptive statistics, confusion matrix, Kaplan Meier curves, and Cox proportional regression were used. Results Median age for 87 evaluable patients was 63 years (range:23-93); 82 (94%) met EANO criteria for possible/probable/confirmed LMD (EANO/LMD). The commonest primary cancers were breast (36,44.0%) and lung (34,41.5%). Primary lung harbored actionable mutations in 18 (53.0%); primary breast expressed hormone receptors in 27 (75%) and HER2 amplification in 8 (22%). Uncontrolled systemic disease was detected in 35 (40%), while 25 (46%) received systemic therapy with medium/high CNS penetrance at LMD diagnosis. Median time from initial cancer to LMD diagnosis was 31 months (range:13-73). LMD was confirmed by CSF cytology in 23/82 (28%), all identified by CNSide™. CNSide™ identified 13 additional cases (36/82,43.9%), increasing diagnostic yield by 56.5%. Median overall survival (mOS) was 31 weeks (95%CI:21-43), significantly worse for CNSide™ positive versus negative: 4.0 versus 16.0 weeks, respectively (HR=0.50,p=0.010). While survival since LMD diagnosis did not differ by histology, time to LMD diagnosis from initial cancer diagnosis was longer for breast (48.5 months,IQR:30.0-87.5) versus lung (8 months,IQR:0.5-16.0) cohorts. mOS was longer for patients eligible for intrathecal chemotherapy (HR:0.189, 95%CI:0.053-0.672, p=0.010). Conclusions This retrospective, real-world analysis of CNSide™ showed increased sensitivity versus cytology and provided clinically relevant molecular CSF analyses.

Hot rolled steel surface defect detection and classification using an automatic ensemble approach

Abstract This study introduces an ensemble-based Deep Neural Network (DNN) model for detecting defects on steel surfaces. The method suggested in this study classifies steel surface conditions into six possible fault categories, namely, crazing, inclusion, rolled in, pitted surface, scratches, and patches. The images undergo preprocessing and extraction of features in spatial and frequency domains using image segmentation techniques such as grey level difference method (GLDM), fast Fourier Transform (FFT), grey level co-occurrence matrix (GLCM), texture analysis and discrete wavelet transform (DWT). The ensembling of image features into a fused feature pool is carried out after the preprocessing of input images that are provided as input to a light-weight neural network model for training and testing. The performance of the model is comprehensively evaluated via an ablation study both before and after ensembling. In addition, the model capability is effectively analyzed using receiver operating characteristics (ROC) curve, confusion matrix from which classification accuracy of the model could be obtained and other parameters including precision and f1-score. It was observed that the proposed deep learning network presents phenomenally high accuracy of 99.72% for detection and classification of steel surface faults. This result was found to be superior when compared with the performance of the same neural network over each feature type individually. This study also compares the classification results of the model built based on the ensembled feature set with the results of various other classification approaches available in literature. The ensemble-based model could potentially be integrated into existing inspection systems for real-time, efficient and robust condition monitoring of steel surfaces.

Artificial intelligence-based tools with automated segmentation and measurement on CT images to assist accurate and fast diagnosis in acute pancreatitis.

To develop an artificial intelligence (AI) tool with automated pancreas segmentation and measurement of pancreatic morphological information on CT images to assist improved and faster diagnosis in acute pancreatitis. This study retrospectively contained 1124 patients suspected for AP and received non-contrast and enhanced abdominal CT examination between September 2013 to September 2022. Patients were divided into training (N = 688), validation (N = 145), testing dataset (N = 291, N = 104 for normal pancreas, N = 98 for AP, N = 89 for AP complicated with PDAC (AP&PDAC)). A model based on convolutional neural network (MSAnet) was developed. The pancreas segmentation and measurement were performed via eight open-source models and MSAnet based tools, and the efficacy was evaluated using Dice similarity coefficient (DSC) and Intersection over union (IoU). The DSC and IoU for patients with different ages were also compared. The outline of tumor and edema in the AP and were segmented by clustering. The diagnostic efficacy for radiologists with or without the assistance of MSAnet tool in AP and AP&PDAC was evaluated using receiver operation curve and confusion matrix. Among all models, MSAnet based tool showed best performance on the training and validation dataset, and had high efficacy on testing dataset. The performance was age-affected. With assistance of the AI tool, the diagnosis time was significantly shortened by 26.8% and 32.7% for junior and senior radiologists, respectively. The area under curve in diagnosis of AP was improved from 0.91 to 0.96 for junior radiologist and 0.98 to 0.99 for senior radiologist. In AP&PDAC diagnosis, AUC was increased from 0.85 to 0.92 for junior and 0.97 to 0.99 for senior. MSAnet based tools showed good pancreas segmentation and measurement performance, which help radiologists improve diagnosis efficacy and workflow in both AP and AP with PDAC conditions. This study developed an AI tool with automated pancreas segmentation and measurement and provided evidence for AI tool assistance in improving the workflow and accuracy of AP diagnosis.

Interactive machine learning for segmenting pores of sandstone in computed tomography images

In the realm of gas field development and CO2/H2 geological storage, the precise characterization of reservoir parameters stands as a critical role. In recent years, digital rock technology has emerged as a cutting-edge tool for examining micro-pore structures, permeability parameter characteristics, and reservoir flow mechanisms. However, there are still certain defects in image segmentation, where low-density pore bodies in low-resolution images make it difficult to clearly distinguish between pores and solid minerals, affecting the accuracy of pore segmentation. This article aims to improve the accuracy of sandstone pore segmentation. We utilized the interactive machine learning segmentation method to segment pore structures and compare the segmented images obtained with the traditional ImageJ grayscale threshold segmentation methods. Furthermore, we evaluated the effectiveness of different segmentation methods quantitatively using the confusion matrix. The research results indicate that the segmentation quality (F1 score) of the machine learning method is higher than the grayscale methods, and the segmentation effectiveness of machine learning method is more stable. To further validate the accuracy of the machine learning method, we simulated the permeability and the results were compared with the experimental results, and the results show that the deviation in permeability segmented through machine learning methods is the smallest. Therefore, our results demonstrate the effective application of the machine learning method in segmenting sandstone pores in the digital rock images.

Bayesian network analysis of factors influencing type 2 diabetes, coronary heart disease, and their comorbidities

ObjectiveBayesian network (BN) models were developed to explore the specific relationships between influencing factors and type 2 diabetes mellitus (T2DM), coronary heart disease (CAD), and their comorbidities. The aim was to predict disease occurrence and diagnose etiology using these models, thereby informing the development of effective prevention and control strategies for T2DM, CAD, and their comorbidities.MethodEmploying a case-control design, the study compared individuals with T2DM, CAD, and their comorbidities (case group) with healthy counterparts (control group). Univariate and multivariate Logistic regression analyses were conducted to identify disease-influencing factors. The BN structure was learned using the Tabu search algorithm, with parameter estimation achieved through maximum likelihood estimation. The predictive performance of the BN model was assessed using the confusion matrix, and Netica software was utilized for visual prediction and diagnosis.ResultThe study involved 3,824 participants, including 1,175 controls, 1,163 T2DM cases, 982 CAD cases, and 504 comorbidity cases. The BN model unveiled factors directly and indirectly impacting T2DM, such as age, region, education level, and family history (FH). Variables like exercise, LDL-C, TC, fruit, and sweet food intake exhibited direct effects, while smoking, alcohol consumption, occupation, heart rate, HDL-C, meat, and staple food intake had indirect effects. Similarly, for CAD, factors with direct and indirect effects included age, smoking, SBP, exercise, meat, and fruit intake, while sleeping time and heart rate showed direct effects. Regarding T2DM and CAD comorbidities, age, FBG, SBP, fruit, and sweet intake demonstrated both direct and indirect effects, whereas exercise and HDL-C exhibited direct effects, and region, education level, DBP, and TC showed indirect effects.ConclusionThe BN model constructed using the Tabu search algorithm showcased robust predictive performance, reliability, and applicability in forecasting disease probabilities for T2DM, CAD, and their comorbidities. These findings offer valuable insights for enhancing prevention and control strategies and exploring the application of BN in predicting and diagnosing chronic diseases.

Improving Automated Hemorrhage Detection in Sparse-view CT via U-Net-based Artifact Reduction.

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. Purpose To explore the potential benefits of deep learning-based artifact reduction in sparse-view cranial CT scans and its impact on automated hemorrhage detection. Materials and Methods In this retrospective study, a U-Net was trained for artifact reduction on simulated sparseview cranial CT scans from 3000 patients obtained from a public dataset and reconstructed with varying sparse-view levels. Additionally, the EfficientNetB2 was trained on full-view CT data from 17,545 patients for automated hemorrhage detection. Detection performance was evaluated using the area under the receiver operator characteristic curve (AUC), with differences assessed using the DeLong test, along with confusion matrices. A total variation (TV) postprocessing approach, commonly applied to sparse-view, served as the basis for comparison. A Bonferronicorrected significance level of 0.001/6 = 0.00017 was used to accommodate for multiple hypotheses testing. Results Images with U-Net postprocessing were better than unprocessed and TV-processed images with respect to image quality and automated hemorrhage detection. With U-Net postprocessing, the number of views could be reduced from 4096 (AUC: 0.97; 95% CI: 0.97-0.98) to 512 (0.97; 0.97-0.98; P < .00017) and to 256 views (0.97; 0.96-0.97; P < .00017) with minimal decrease in hemorrhage detection performance. This was accompanied by mean structural similarity index measure increases of 0.0210 (95% CI: 0.0210-0.0211) and 0.0560 (95% CI: 0.0559-0.0560) relative to unprocessed images. Conclusion U-Net based artifact reduction substantially enhances automated hemorrhage detection in sparse-view cranial CTs. ©RSNA, 2024.

Considerations for implanting speech brain computer interfaces based on functional magnetic resonance imaging

Objective. Brain-computer interfaces (BCIs) have the potential to reinstate lost communication faculties. Results from speech decoding studies indicate that a usable speech BCI based on activity in the sensorimotor cortex (SMC) can be achieved using subdurally implanted electrodes. However, the optimal characteristics for a successful speech implant are largely unknown. We address this topic in a high field blood oxygenation level dependent functional magnetic resonance imaging (fMRI) study, by assessing the decodability of spoken words as a function of hemisphere, gyrus, sulcal depth, and position along the ventral/dorsal-axis. Approach. Twelve subjects conducted a 7T fMRI experiment in which they pronounced 6 different pseudo-words over 6 runs. We divided the SMC by hemisphere, gyrus, sulcal depth, and position along the ventral/dorsal axis. Classification was performed on in these SMC areas using multiclass support vector machine (SVM). Main results. Significant classification was possible from the SMC, but no preference for the left or right hemisphere, nor for the precentral or postcentral gyrus for optimal word classification was detected. Classification while using information from the cortical surface was slightly better than when using information from deep in the central sulcus and was highest within the ventral 50% of SMC. Confusion matrices where highly similar across the entire SMC. An SVM-searchlight analysis revealed significant classification in the superior temporal gyrus and left planum temporale in addition to the SMC. Significance. The current results support a unilateral implant using surface electrodes, covering the ventral 50% of the SMC. The added value of depth electrodes is unclear. We did not observe evidence for variations in the qualitative nature of information across SMC. The current results need to be confirmed in paralyzed patients performing attempted speech.

Eigenvalue Distributions in Random Confusion Matrices: Applications to Machine Learning Evaluation

This paper examines the distribution of eigenvalues for a 2×2 random confusion matrix used in machine learning evaluation. We also analyze the distributions of the matrix’s trace and the difference between the traces of random confusion matrices. Furthermore, we demonstrate how these distributions can be applied to calculate the superiority probability of machine learning models. By way of example, we use the superiority probability to compare the accuracy of four disease outcomes machine learning prediction tasks.

Domestication constrains the ability of dogs to convey emotions via facial expressions in comparison to their wolf ancestors

Dogs (Canis lupus familiaris) are the domestically bred descendant of wolves (Canis lupus). However, selective breeding has profoundly altered facial morphologies of dogs compared to their wolf ancestors. We demonstrate that these morphological differences limit the abilities of dogs to successfully produce the same affective facial expressions as wolves. We decoded facial movements of captive wolves during social interactions involving nine separate affective states. We used linear discriminant analyses to predict affective states based on combinations of facial movements. The resulting confusion matrix demonstrates that specific combinations of facial movements predict nine distinct affective states in wolves; the first assessment of this many affective facial expressions in wolves. However, comparative analyses with kennelled rescue dogs revealed reduced ability to predict affective states. Critically, there was a very low predictive power for specific affective states, with confusion occurring between negative and positive states, such as Friendly and Fear. We show that the varying facial morphologies of dogs (specifically non-wolf-like morphologies) limit their ability to produce the same range of affective facial expressions as wolves. Confusion among positive and negative states could be detrimental to human–dog interactions, although our analyses also suggest dogs likely use vocalisations to compensate for limitations in facial communication.

Speech Emotion Recognition Using ANN

The speech is the most effective means of communication, to recognize the emotions in speech is the most crucial task. In this paper we are using the Artificial Neural Network to recognize the emotions in speech. Hence, providing an efficient and accurate technique for speech based emotion recognition is also an important task. This study is focused on seven basic human emotions (angry, disgust, fear, happy, neutral, surprise, sad). The training and validating accuracy and also lose can be seen in a graph while training the dataset. According to it confusion matrix for model is created. The seven features Frequency, Pitch, Amplitude and formant of speech is used to recognize seven basic emotions from speech. Keywords: Emotion detection, Emotion classification, verbal emotion classification.

Hybrid Transfer Learning and Support Vector Machine Models for Asphalt Pavement Distress Classification

Pavement condition evaluation plays a crucial role in assisting with the management of the highway infrastructure. However, the current methods used for assessing pavement conditions are costly, time-consuming, and subjective. There is a growing need to automate these assessment tactics and leverage low-cost technologies to enable widespread deployment. This study aims to develop robust and highly accurate models for classifying asphalt pavement distresses using transfer learning (TL) techniques based on pretrained deep learning (DL) networks. This topic has gained considerable attention in the field since 2015 when DL became the mainstream choice for various computer vision tasks. While progress has been made in TL model development, challenges persist in areas of accuracy, repeatability, and training cost. To tackle these challenges, this study proposes hybrid models that combine DL networks with support vector machines (SVMs). Three strategies were evaluated: single DL models using transfer learning (TLDL), hybrid models combining DL and SVM (DL+SVM), and hybrid models combining TLDL and SVM (TLDL+SVM). The performance of each strategy was assessed using statistical metrics based on the confusion matrix. Results consistently showed that the TLDL+SVM strategy outperformed the other approaches in accuracy and F1 scores, regardless of the DL network type. On average, the hybrid models achieved an accuracy of 95%, surpassing the 80% accuracy of the best single model and the 55% accuracy for DL+SVM without TL. The results clearly indicate that employing transfer-learned models as feature extractors, in combination with SVM as the classifier, consistently achieves exceptional performance.

Need to revise classification of physical activity intensity in older adults? The use of estimated METs, measured METs and V̇O2Reserve.

Multiples of resting metabolic rate (RMR) are often used to classify physical activity intensity, a concept known as the Metabolic Equivalent of Task (MET). However, the METs metrics may misclassify physical activity intensity in older adults because of age related changes in RMR and maximal aerobic capacity (V̇O2max). This study aimed to 1) compare classifications of activity intensity by estimated (METsestimated) and measured (METsmeasured) METs and 2) compare physical activity classified by absolute (METsmeasured) versus relative intensity (%V̇O2Reserve) in older adults. Ninety-eight adults aged 75-90 years participated in the study. RMR and V̇O2 during sitting, standing, daily activities and 6-minute-walking-test were measured. V̇O2Reserve was defined as the difference between V̇O2max and RMR. Moderate and vigorous intensity was classified as 3 and 6 METs and 40% and 60% of V̇O2Reserve, respectively. Paired t-tests and a confusion matrix were used to investigate aim 1 and 2, respectively. METsmeasured was 24% lower than the standard 1 MET of 3.5ml O2·min-1·kg-1. METsestimated underestimated the intensity during daily and walking activities when compared to METsmeasured. Nevertheless, when comparing METsmeasured to percentages of V̇O2Reserve, a mismatch was shown for moderate intensity in 47-67% of the participants during daily activities, and 21% of the participants during self-selected gait speed. Applying METsestimated for older adults leads to potential underestimation of physical activity intensity, suggesting that current classification metrics should be revised for older adults. V̇O2Reserve is a candidate metric for establishing precise physical activity intensity cut-points for older adults.

Development of Attenuated Total Reflectance Fourier Transform Infrared Spectroscopy Coupled with Multivariate Classification Chemometric Model for Routine Screening of Paracetamol, Ibuprofen, and Aspirin Adulteration in Herbal Products

Objective: The objective of this study is to develop and validate a routine screening test for the determination of three common antipyretic-analgesic synthetic drugs (paracetamol, ibuprofen, and aspirin) adulteration in herbal products using Attenuated Total Reflectance Fourier Transform Infrared (ATR-FTIR) coupled with chemometric method. Method: ATR-FTIR spectra of sixteen testing sets of herbal product samples for pain and fever indications were used for multivariate chemometrics model construction. Linear Discriminant Analysis (LDA) was selected as a method for model construction with IBM SPSS for statistical analysis. Model development employed feature selection, such as the stepwise method for variable selection. The model with a high %correct classification and cross-validation was selected and was then validated with an independent testing data set with an auto-prediction test, confusion matrix, and Receiver Operating Characteristic (ROC) curve. To validate the developed test for routine use, the result from ATR-FTIR method was compared with the standard HPLC and TLC analyses used for adulteration screening. objective: Creating validated screening tools for herbal products adulterated with three common antipyretic-analgesic synthetic drugs (Paracetamol, Ibuprofen, and Aspirin) using ATR-FTIR couple with chemometric method Results: The selected model's overall %correct classification result was 97.7%, with a cross-validation of 93.8% rate in training set samples. External validation with an independent testing dataset gave an overall correct classification of 93.8%, with an area under the curve of ROC at 0.979. Comparative testing revealed that model performance was comparable with the HPLC and TLC methods, which routinely detect the presence of paracetamol, aspirin, and ibuprofen. The results of testing set samples classification were consistent with training set samples. Conclusion: Against the standard chromatographic methods, the multivariate chemometric model based on ATR-FTIR demonstrates comparable detection capability to determine adulteration of paracetamol, ibuprofen, and aspirin in herbal products.

Prediction of student exam performance using data mining classification algorithms

AbstractStudent outcomes are of great importance in higher education institutions. Accreditation bodies focus on them as an indicator to measure the performance and effectiveness of the institution. Forecasting students’ academic performance is crucial for every educational establishment seeking to enhance performance and perseverance of its students and reduce the failure rate in the future. The main goal of this study is to predict the performance of undergraduate first-level students in the Computer Department during the years 2016 to 2021 to enhance their performance in future by discovering the best algorithm use to analyze the educational data to identify the students’ academic performance. The secondary data was collected by reviewing the Student Affairs Department at the Faculty of Specific Education at Damietta University, in addition to the Statistics Department at the university. The dataset contained 830 instances after excluding 139 instances of missing values, irrelevant rows, and outliers. The dataset was divided into train (577 instances (70%)), test (253 instances (30%)) and involved six features such year, midterm, practical exam, writing exam, final total degree, and grade. This paper use five machine learning (ML) algorithms which was selected according to the literature review and high accuracy in predicting educational data mining: For the purpose of comparison, a number of different machine learning algorithms, such as Random Forest, Decision Tree, Naive Bayes, Neural Network, and K-Nearest Neighbours, were utilized and evaluated with evaluation metrics such as confusion matrix, accuracy, precision, recall, and F-measure. The Random Forest and Decision Tree classifiers emerged as the top-performing algorithms, accurately categorizing 250 instances when predicting students' performance in the statistics course. This was determined based on the findings of the study. Out of a total of 253 instances that were included in the testing set, they only made three incorrect classifications.

Abstract PO5-21-03: Cosmetic assessment in the UNICANCER HypoG-01 trial: a deep learning approach

Abstract Introduction Cosmetic evaluation after breast cancer treatment is a clinical indicator of toxicity. User bias and inter-subject variability hamper this objective. To address this limitation, a deep learning approach was developed on the basis of the HYPOG-01 trial (NCT03127995), a phase III trial comparing hypo to normo-fractionated radiotherapy (RT) in breast cancer patients requiring nodal irradiation. Material and Methods Cosmetic outcomes using the Harris scale were assessed by a radiation oncologist using a 4-level rating system from excellent to poor. This evaluation involved photographs from 581 female patients included in the intention-to-treat population of the HYPOG-01 study analysis (mastectomy/pamectomy and non-usable cases excluded). Front images were taken with the arms along the body at baseline, 3-weeks after radiotherapy start, end of treatment, then 6 months and every year after randomization up to 5 years. Comparing manual landmark annotation with the semi-automated software BCCT.core©, the agreement rate was moderate, with an intra-class correlation coefficient (ICC) of 0.66 (95%CI 0.57-0.73). The dataset consisted of 2,348 images, which were divided into exclusive patient-based training (1,661), validation (308), and testing (377) datasets. The distribution of Harris scores in the dataset was highly imbalanced: 7% excellent, 33% good, 45% fair, and 15% poor. Nipple landmarks were used as landmarks to address picture acquisition variations by cropping and resizing images to 224 × 224 resolution. Feature extraction was performed using a Swin-TransformerV2, an advanced attention-based vision model initially trained on ImageNet. Newly integrated fully connected layers categorized the extracted features. The model was trained for 300 epochs, and the highest F1-score model was selected. Asymmetry in texture, marks, and breast geometry played a crucial role in Harris scoring. To improve the model's performance, we generated symmetrical images from the region of interest, averaging them with the original images, and we incorporated the timestamps from image captures as an additional influencing factor. We employed techniques such as contrast modulation, lighting adjustments, and geometric transformations to augment the dataset, introducing additional variations and enhancing the model's generalization and accuracy. Results The performance of our model was evaluated using balanced binary classification, multi-class accuracy, and F1-score. Comparatively, our model performed similarly to BCCT in terms of overall accuracy but demonstrated better performance in separating multiple classes, as indicated in Table 1. In Table 2, we present a confusion matrix that provides insights into the model's performance Conclusion The proposed solution simplifies and accelerates the evaluation process by utilizing only two nipple landmarks, surpassing manual and semi-automated tools. This advancement opens doors for automated, large-scale cosmetic toxicity evaluation. Continuous improvement and validation contribute to its robustness and reinforce its significant impact in assessing cosmetic outcomes after breast cancer treatment. Table. Evaluation of performance on the test set (only on cases with evaluation by BCCT (327 images)). Table. Confusion matrix between our predictions and the labels on the test set (only on cases with evaluation by BCCT (327 images)). Citation Format: Alexandre Cafaro, Amandine Ruffier, Gabriele Bielinyte, Youlia Kirova, Séverine Racadot, Mohamed Benchalal, Jean-Baptiste Clavier, Claire Charra-Brunaud, Marie-Eve Chand-Fouche, Delphine Argo-Leignel, Karine Peignaux, Ahmed Benyoucef, David Pasquier, Philippe Guilbert, Julien Blanchecotte, Agnès Tallet, Adeline Petit, Guillemette Bernadou, Xavier Zasadny, Claire Lemanski, Jacques Fourquet, Emmanuelle Malaurie, Honorine Kouto, Carole Massabeau, Alexandre Henni, Pauline Regnault, Aurélie Belliere, Yazid Belkacemi, Magali Le Blanc-Onfroy, Julien Geffrelot, Jean-Briac Prevost, Eleni Karamouza, Stefan Michiels, Marie Bergeaud, Assia Lamrani-Ghaouti, Sami Romdhani, Alexis Bombezin-Domino, Nikos Paragios, Sofia Rivera. Cosmetic assessment in the UNICANCER HypoG-01 trial: a deep learning approach [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 PO5-21-03.

Stone decision engine accurately predicts stone removal and treatment complications for shock wave lithotripsy and laser ureterorenoscopy patients.

Kidney stones form when mineral salts crystallize in the urinary tract. While most stones exit the body in the urine stream, some can block the ureteropelvic junction or ureters, leading to severe lower back pain, blood in the urine, vomiting, and painful urination. Imaging technologies, such as X-rays or ureterorenoscopy (URS), are typically used to detect kidney stones. Subsequently, these stones are fragmented into smaller pieces using shock wave lithotripsy (SWL) or laser URS. Both treatments yield subtly different patient outcomes. To predict successful stone removal and complication outcomes, Artificial Neural Network models were trained on 15,126 SWL and 2,116 URS patient records. These records include patient metrics like Body Mass Index and age, as well as treatment outcomes obtained using various medical instruments and healthcare professionals. Due to the low number of outcome failures in the data (e.g., treatment complications), Nearest Neighbor and Synthetic Minority Oversampling Technique (SMOTE) models were implemented to improve prediction accuracies. To reduce noise in the predictions, ensemble modeling was employed. The average prediction accuracies based on Confusion Matrices for SWL stone removal and treatment complications were 84.8% and 95.0%, respectively, while those for URS were 89.0% and 92.2%, respectively. The average prediction accuracies for SWL based on Area-Under-the-Curve were 74.7% and 62.9%, respectively, while those for URS were 77.2% and 78.9%, respectively. Taken together, the approach yielded moderate to high accurate predictions, regardless of treatment or outcome. These models were incorporated into a Stone Decision Engine web application (http://peteranoble.com/webapps.html) that suggests the best interventions to healthcare providers based on individual patient metrics.

Crossflow Microfiltration of Aqueous Suspensions with Guar and Xanthan Gums: Identification of Solutions Using Artificial Neural Networks

Artificial Neural Networks (ANNs) are mathematical models used in the computational area that act in an analogous way to the central nervous system of living beings, which possess the ability of acquiring knowledge in a technique called machine learning, allowing them to recognize patterns and stop numerous applications. Therefore, the objective was to develop Neural Networks capable of identifying aqueous solutions with Guar and Xanthan gums (widely used in the food industry) during the crossflow microfiltration process. The networks were trained in the supervised learning algorithms trainscg, trainlm and traingd, all in the 70/15/15 model, for a range of five to fifteen neurons in the hidden layer, whose datasets were found in the literature, referring to temperature, flow velocity, pressure, transmembrane flow rate, time and membrane pore size. The software used to implement the ANNs was MATLAB and the evaluation criteria consisted of the analysis of the parameters confusion matrix, error histogram,performance and ROC curve. In summary, ten ANNs had satisfactory performances, presenting confusion matrices with accuracies above 98.8%, error histogram graphs being Gaussian centered at 0, decaying performance curves with stopping criterion equal to 6 errors in the validation set and ROC graphs similar to a square with vertices at (0,0), (1,0), (0,1) and (1,1), results considered satisfactory in the literature.