Predictive Machine Research Articles

Predictive machine learning-based error correction in GPS/IMU localization to improve navigation of autonomous vehicles

Precise localization is crucial for the safety-critical factor and effective navigation of autonomous vehicles. This applied research examines machine learning models’ use to estimate, predict and correct errors in Global Positioning System (GPS)/ Inertial Measurement Unit (IMU) localization for autonomous vehicles indoors and outdoors applications. This ongoing development aims to improve localization accuracy by utilizing exploratory data analysis (EDA) and implementing models such as linear regression, random forest regressor, and decision tree regressor. The assessment is performed with the mean squared error (MSE) metric, yielding values of 1.7069427028104143e−05 for the decision tree, linear regression, and random forest models. The results showed that the model with the highest performance is determined by evaluating the Mean Squared Error (MSE) values.

Predictive Machine Learning Models for Assessing Lebanese University Students' Depression, Anxiety, and Stress During COVID-19.

University students are experiencing a mental health crisis. COVID-19 has exacerbated this situation. We have surveyed students in 2 universities in Lebanon to gauge their mental health challenges. We have constructed a machine learning (ML) approach to predict symptoms of depression, anxiety, and stress based on demographics and self-rated health measures. Our approach involved developing 8 ML predictive models, including Logistic Regression (LR), multi-layer perceptron (MLP) neural network, support vector machine (SVM), random forest (RF) and XGBoost, AdaBoost, Naïve Bayes (NB), and K-Nearest neighbors (KNN). Following their construction, we compared their respective performances. Our evaluation shows that RF (AUC = 78.27%), NB (AUC = 76.37%), and AdaBoost (AUC = 72.96%) have provided the highest-performing AUC scores for depression, anxiety, and stress, respectively. Self-rated health is found to be the top feature in predicting depression, while age was the top feature in predicting anxiety and stress, followed by self-rated health. Future work will focus on using data augmentation approaches and extending to multi-class anxiety predictions.

Pilot study of a real-time early agitation capture technology (REACT) for children with intellectual and developmental disabilities.

Children and adolescents with intellectual and developmental disabilities (IDD), particularly those with autism spectrum disorder, are at increased risk of challenging behaviors such as self-injury, aggression, elopement, and property destruction. To mitigate these challenges, it is crucial to focus on early signs of distress that may lead to these behaviors. These early signs might not be visible to the human eye but could be detected by predictive machine learning (ML) models that utilizes real-time sensing. Current behavioral assessment practices lack such proactive predictive models. This study developed and pilot-tested real-time early agitation capture technology (REACT), a real-time multimodal ML model to detect early signs of distress, termed "agitations." Integrating multimodal sensing, ML, and human expertise could make behavioral assessments for people with IDD safer and more efficient. We leveraged wearable technology to collect behavioral and physiological data from three children with IDD aged 6 to 9 years. The effectiveness of the REACT system was measured using F1 score, assessing its usefulness at the time of agitation to 20s prior. The REACT system was able to detect agitations with an average F1 score of 78.69% at the time of agitation and 68.20% 20s prior. The findings support the use of the REACT model for real-time, proactive detection of agitations in children with IDD. This approach not only improves the accuracy of detecting distress signals that are imperceptible to the human eye but also increases the window for timely intervention before behavioral escalation, thereby enhancing safety, well-being, and inclusion for this vulnerable population. We believe that such technological support system will enhance user autonomy, self-advocacy, and self-determination.

Prediction of Alzheimer's Disease Based on 3D Genome Selected circRNA.

Alzheimer's disease (AD) is a neurodegenerative disease and there is by far no effective treatment for it, especially in its late stage. Circular RNAs (circRNAs), known as a class of non-coding RNAs are widely observed in eukaryotic transcriptomes, and are reported to play an important role in neurodegenerative diseases including AD. circRNAs usually act as microRNA (miRNA) inhibitors or «sponges» to regulate the function of miRNAs, leading to subsequent changes in protein activities and functions. Accumulating evidence indicates that circRNAs can serve as potential biomarker in AD early prediction. The functional roles of circRNAs are very versatile including miRNAs binding - thus affecting downstream gene expression, generating abnormally translated protein peptides, and affecting epigenetic modifications which subsequently affect AD related gene expressions. Therefore, identifying AD-related circRNAs can contribute to AD early diagnosis and intervention. In this work, we collected and curated an AD-related circRNA dataset; by exploring the circRNAs' corresponding DNA loci distribution in chromatin 3D conformation (3D genome) and utilize the such 3D genome information, we were able to selected a concise yet predictively effective circRNA panel, based on which, significantly better AD prediction machine learning models were achieved.

Predictive machine learning models based on VASARI features for WHO grading, isocitrate dehydrogenase mutation, and 1p19q co-deletion status: a multicenter study.

The objective of our study was to develop predictive models using Visually Accessible Rembrandt Images (VASARI) magnetic resonance imaging (MRI) features combined with machine learning techniques to predict the World Health Organization (WHO) grade, isocitrate dehydrogenase (IDH) mutation status, and 1p19q co-deletion status of high-grade gliomas. To achieve this, we retrospectively included 485 patients with high-grade glioma from the First Affiliated Hospital of Xinjiang Medical University, of which 312 patients were randomly divided into a training set (n=218) and a test set (n=94) in a 7:3 ratio. Twenty-five VASARI MRI features were selected from an initial set of 30, and three machine learning models - Multilayer Perceptron (MP), Bernoulli Naive Bayes (BNB), and Logistic Regression (LR) - were trained using the training set. The most informative features were identified using recursive feature elimination. Model performance was assessed using the test set and an independent validation set of 173 patients from Beijing Tiantan Hospital. The results indicated that the MP model exhibited the highest predictive accuracy on the training set, achieving an area under the curve (AUC) close to 1, indicating perfect discrimination. However, its performance decreased in the test and validation sets; particularly for predicting the 1p19q co-deletion status, the AUC was only 0.703, suggesting potential overfitting. On the other hand, the BNB model demonstrated robust generalization on the test and validation sets, with AUC values of 0.8292 and 0.8106, respectively, for predicting IDH mutation status and 1p19q co-deletion status, indicating high accuracy, sensitivity, and specificity. The LR model also showed good performance with AUCs of 0.7845 and 0.8674 on the test and validation sets, respectively, for predicting IDH mutation status, although it was slightly inferior to the BNB model for the 1p19q co-deletion status. In conclusion, integrating VASARI MRI features with machine learning techniques shows promise for the non-invasive prediction of glioma molecular markers, which could guide treatment strategies and improve prognosis in glioma patients. Nonetheless, further model optimization and validation are necessary to enhance its clinical utility.

ACCURATE PREDICTION OF DRILL BIT PENETRATION RATE IN ROCK USING SUPERVISED MACHINE LEARNING TECHNIQUES BASE ON LABORATORY TEST DATA

Knowing the rate of penetration of a drill bit in rocks is among the most important parameters in their behaviour measurement. However, the direct measurement of ROP in rocks is a high-cost and time-intensive process. Therefore, obtaining the ROP parameter through a method other than direct measurement can be very useful and effective. Predictive machine learning methods are among the strong and precise techniques for the indirect measurement of ROP. To this end, 492 samples were tested under different UCS, µ, WOB, and ω conditions to obtain the corresponding ROP. To achieve an accurate model, three methods of linear regression analysis, lasso regression, and ridge regression were compared in terms of prediction accuracy. These models were compared through performance criteria of the prediction process and error-based charts. The performance criteria were measured using three measures: mean absolute percentage error, D-squared pinball score, and mean Poisson deviance error. For the MAPE index, the Lasso and Ridge models performed the best with values of 0.2557. Concerning the D2PS index, the linear regression model and Ridge performed better with values of 0.4083 and 0.4025, respectively. Finally, for the MPDE index, the Ridge model provided a more accurate performance with a value of 0.0105. For a better comparison, an objective function was created and calculated by combining these three indicators. The results showed the best rank for the Ridge model with an estimated value of 659.475. Finally, it was concluded that the Ridge model is a reliable and accurate model for predicting the ROP.

Machine learning for predicting cognitive deficits using auditory and demographic factors.

Predicting neurocognitive deficits using complex auditory assessments could change how cognitive dysfunction is identified, and monitored over time. Detecting cognitive impairment in people living with HIV (PLWH) is important for early intervention, especially in low- to middle-income countries where most cases exist. Auditory tests relate to neurocognitive test results, but the incremental predictive capability beyond demographic factors is unknown. Use machine learning to predict neurocognitive deficits, using auditory tests and demographic factors. The Infectious Disease Center in Dar es Salaam, Tanzania. Participants were 939 Tanzanian individuals from Dar es Salaam living with and without HIV who were part of a longitudinal study. Patients who had only one visit, a positive history of ear drainage, concussion, significant noise or chemical exposure, neurological disease, mental illness, or exposure to ototoxic antibiotics (e.g., gentamycin), or chemotherapy were excluded. This provided 478 participants (349 PLWH, 129 HIV-negative). Participant data were randomized to training and test sets for machine learning. The main outcome was whether auditory variables combined with relevant demographic variables could predict neurocognitive dysfunction (defined as a score of <26 on the Kiswahili Montreal Cognitive Assessment) better than demographic factors alone. The performance of predictive machine learning algorithms was primarily evaluated using the area under the receiver operational characteristic curve. Secondary metrics for evaluation included F1 scores, accuracies, and the Youden's indices for the algorithms. The percentage of individuals with cognitive deficits was 36.2% (139 PLWH and 34 HIV-negative). The Gaussian and kernel naïve Bayes classifiers were the most predictive algorithms for neurocognitive impairment. Algorithms trained with auditory variables had average area under the curve values of 0.91 and 0.87, F1 scores (metric for precision and recall) of 0.81 and 0.76, and average accuracies of 86.3% and 81.9% respectively. Algorithms trained without auditory variables as features were statistically worse (p < .001) in both the primary measure of area under the curve (0.82/0.78) and the secondary measure of accuracy (72.3%/74.5%) for the Gaussian and kernel algorithms respectively. Auditory variables improved the prediction of cognitive function. Since auditory tests are easy-to-administer and often naturalistic tasks, they may offer objective measures or predictors of neurocognitive performance suitable for many global settings. Further research and development into using machine learning algorithms for predicting cognitive outcomes should be pursued.

Rapid screening B-site doping ferroelectric perovskite with high curie temperature for electronic applications by a novel IDBO-RF approach

Ferroelectric perovskites with high Curie temperatures (Tc) have emerged as a significant focus in the field of electronic materials in recent years. However, accurately predicting Tc remains a longstanding challenge in the discovery of new functional materials. Therefore, it is crucial to explore the relationship between potential materials descriptors and Tc, which can greatly expedite the identification of ferroelectric perovskite materials with high Tc. In this study, a novel intelligence approach was proposed as a promising solution for identifying Pb[Bx′B1−x′′]O3-type ferroelectric perovskite materials with B-site doping. To achieve this, a comprehensive set of features was utilized to generate synthetic dataset based on available experimental observations from literature and databases. By applying predictive machine learning (ML) models to analyze the synthetic dataset, four key descriptors were identified that exhibited strong correlations with Tc in these materials. First, considering the application of compressed sensing method to reduce high dimensional features, a novel descriptor cos(BCCvd)/(Ec*BCCe) was created and identified, which combined with the first ionization energy (IE1), the lattice energy difference between atoms and their neighboring coordinated atoms (BCCed), and the spin magnetic moment (GSm). These descriptors were found to be closely correlated with Tc of ferroelectric perovskites. Additionally, the Shapley Additive Explanation (SHAP) method was employed to analyze the relationship between Tc and the selected descriptors. Furthermore, based on the selected Random Forest (RF) model with superior performance among several machine learning models tested, an optimization algorithm was proposed combining Improved Dung Beetle Optimizer (IDBO) and RF regression for predicting Tc. Then the designed hybrid IDBO-RF model achieved a mean absolute error (MAE) of 13.95 K in Tc prediction, surpassing traditional methods based on tolerance factor and ionic displacement, which had a MAE of 30.2 K. From a pool of 2745 candidate materials, two potential ferroelectric perovskites with high Tc were successfully screened and identified. This proposed framework will facilitate Tc prediction technology for the discovery of ferroelectric perovskite materials with favorable performance.

A novel machine learning based lumping approach for the reduction of large kinetic mechanisms for plasma-assisted combustion applications

The development of skeletal mechanisms has become essential for multi-dimensional simulations of plasma-assisted combustion (PAC). However, reduction tools developed for traditional combustion applications are not always applicable to PAC, due to the complex interplay between non-equilibrium plasma and combustion kinetics. Plasma direct relation graph with error propagation (P-DRGEP) is a recent plasma-specific reduction method developed in order to incorporate plasma energy branching in the reduction. In the first part of this work, the applicability of P-DRGEP to large kinetic mechanisms is investigated. A detailed isooctane/air plasma mechanism containing 2805 species and 18457 reactions is reduced to 415 species and 4716 reactions, keeping errors on ignition time within 3% for a wide range of initial conditions: from 750 K to 1200 K, 10 atm and equivalence ratios from 0.75 to 1.50. The second part focuses on isomer lumping, which is another reduction technique widely used in combustion. When applied to PAC, it is shown that the resulting lumped mechanism produces poor results. A novel plasma-specific isomer lumping strategy using machine learning is proposed instead. With the supervised algorithm of gradient boosting, predictive regression models are generated, which describe rate coefficients of lumped reactions adequately. These models are trained with simulation data. Leveraging this newly proposed lumping approach on the reduced mechanism, allows for an additional 28% reduction in the number of species and 19% reduction in the number of reactions. Two different versions are presented: in the first one the models are trained using one input feature (1D), while in the second one, two input features are selected (2D). The resulting lumped mechanism is shown to produce accurate predictions of PAC over the entire parameter space of interest, while significantly decreasing the computational time. Indicatively, with the 1D version the maximum error on ignition time in this range of conditions is 6%. The 2D approach produces even lower errors, which do not exceed 3%.Novelty and significance statementIn this work, a novel approach for isomer lumping, in plasma-assisted combustion mechanisms, is demonstrated. This plasma-specific approach, uses predictive machine learning regression models to describe the complex evolution of lumped reaction rate coefficients. Combining it with the plasma direct relation graph with error propagation, a powerful reduction framework is created, which is successfully demonstrated on a detailed isooctane/air plasma kinetic mechanism, via zero-dimensional ignition simulations. This framework constitutes a useful tool towards the creation of highly accurate skeletal mechanisms, which significantly reduce the computational costs of simulations.

Spectral clustering algorithm based web mining and quadratic support vector machine for learning style prediction in E-learning platform

A learning system, which is composed of a computer and the internet as the major elements, is termed an e-learning platform. It also promotes the education standard with the utilization of modern technology and equipment. Meanwhile, to enhance the standard of education significantly, it is important to predict the learning style of the users with the assistants of feedback and supervision. Nevertheless, it will avert the inherent correlation among e-learning behaviors. Hence, to predict the learning style automatically we propose a novel Spectral Clustering algorithm based Quadratic Support Vector Machine (E-SVM) approach. Our proposed approach employs two phases: (i) Utilizing the Web usage mining approach the learning secrets are extracted from the log files of learners. (ii) The classification of learning styles of learners is effectuated with the proposed approach. Experiments are demonstrated with Python package and analyzed the performance. For simulation, we have utilized real-time dataset and compared the results with the state-of-art approaches. Our approach surpasses all the other approaches.

Enhancing Academic Outcomes through an Adaptive Learning Framework Utilizing a Novel Machine Learning-Based Performance Prediction Method

Introduction:E landscapes have been transformed by technological advancements, enabling adaptive and flexible learning through AI-based and decision-oriented adaptive learning systems. The increasing importance of this solutions is underscored by the pivotal role of the learner model, representing the core of the teaching-learning dynamic. This model, encompassing qualities, knowledge, abilities, behaviors, preferences, and unique distinctions, plays a crucial role in customizing the learning experience. It influences decisions related to learning materials, teaching strategies, and presentation styles. Objective: This study meets the need for applying AI-driven adaptive learning in education, implementing a novel method that uses self-esteem (ES), emotional intelligence (EQ), and demographic data to predict student performance and adjust the learning process. Methods: Our study involved collecting and processing data, constructing a predictive machine learning model, implementing it as an online solution, and conducting an experimental study with 146 high school students in computer science and French as foreign language. The aim was to tailor the teaching-learning process to the learners' performance. Results: significant correlations were observed between self-esteem, emotional intelligence, demographic data, and final grades. The predictive model demonstrated a 90 % accuracy rate. In the experimental group, the results indicated higher scores, with an average of 15,78/20 compared to the control group's 12,53/20 in computer science. Similarly, in French as a foreign language, the experimental group achieved an average of 13,78/20, surpassing the control group's 10,47/20. Conclusion: the achieved results motivate the creation of a multifactorial AI-driven adaptive learning platform. Recognizing the necessity for improvement, we aim to refine the predicted performance score through the incorporation of a diagnostic evaluation, ensuring an optimal grouping of learners

Speaker landscapes: machine learning opens a window on the everyday language of opinion

ABSTRACT We propose a new method that embeds speakers into a spatial representation according to the linguistic similarity of their contributions to a debate. Such “speaker landscapes” can be constructed quantitatively using word embeddings by annotating text corpora of speech samples by tokens representing the speakers. The way embeddings are constructed from predictive machine learning models means that speaker-tokens are placed closer together if they are easier to confuse given their speech samples. The result is a nuanced measure of similarity in speech which takes into account the wealth of linguistic signification and structure that the text prediction model can learn. We validate this tool using two South African case studies: the Twitter debate around the arrest and imprisonment of former president Jacob Zuma, and quotes from the news media about land reform. The results show that speaker landscapes make social qualities such as group membership and opinions evident, and we discuss how speaker landscapes open up new methods for studying opinion discourse with text data.

Multiomics Blood Test for Increasing Asymptomatic AD Patients Proportion in Preventive Clinical Trials

AbstractBackgroundWith the first positive phase III trials for anti‐amyloid approaches in MCI participants, the drug evaluation during the asymptomatic Alzheimer’s disease (AD) stage is now more relevant than ever. Currently, cognitively unimpaired participants are included in preventive trials based on risk factors such as amyloid positivity. The lifetime risks for an individual without cognitive impairment does not exceed 30% for a 65‐year‐old with amyloidosis alone (Brookmeyer et al., 2018). Based on this risk factor, most participants included will never develop AD and therefore will not respond to anti‐AD therapies. There is a need to develop new biomarkers with high specificity predicting which individuals in the cognitively unimpaired population will exhibit AD symptoms and which will not to conduct effective preventive clinical trials.MethodTargeted mass spectrometry assays were developed for 81 blood biomarkers (45 proteins and 36 metabolites) pre‐identified in AAV‐AD rats (Audrain et al. 2018). 287 participants were collected in plasma at baseline and followed clinically from 1 to 18 years. Participants’ diagnosis (AD, or Healthy Controls) was established at the last clinical visit based on reference diagnosis standard. 73.9% of participants had available amyloid status, 42.9% were determined as amyloid positive. Predictive machine learning models for asymptomatic AD participants (n = 48, followed for an average of 6.1 years) among individuals without cognitive decline (HC, n = 239, followed for an average of 4.6 years) was developed. The training dataset (70%) aimed to select the biomarkers subset, train the algorithm, and define the cut‐off. External test dataset (30%) aimed to validate in blind the locked predictive model.ResultWith 20 blood biomarkers and 2 covariates (age and gender), the model predicted asymptomatic AD (n = 16) from HC participants (n = 72) with 81.9% specificity and 56.3% sensitivity during external validation (AUROC = 69.1%, p = 0.02). When current amyloid tests (CSF or PET) and this predictive ML model were applied in series, and participants positive for both were predicted as asymptomatic AD, 90.9% specificity and 54.5% sensitivity were achieved.ConclusionMultiomics blood peripheral biomarkers predict asymptomatic AD patients among cognitively unimpaired individuals with high specificity. This could constitute an additional tool for the inclusion of participants in preventive clinical trials.

OP63 Incorporating Machine Learning Methods In Health Economic Evaluations: A Case Study On Depression Prevention

IntroductionNew methodologies such as machine learning are becoming widely available and are increasingly used. However, more guidance on their role in the context of economic evaluations would be beneficial.MethodsWe developed a machine learning model to predict recurrent depressive episodes and incorporated the model outcomes in a health economic model to assess the cost effectiveness of offering targeted prevention of recurrent depression. We considered the impact on cost effectiveness (defined as cost per quality-adjusted life-year) for machine learning models with different thresholds for classifying a patient as being at risk, resulting in different precision-recall pairs.ResultsTargeted prevention of recurrent depression could enhance the cost effectiveness of depression treatment by preventing a small number of recurrent depressive episodes in patients where the estimated risk of recurrence is relatively high. More depressive episodes could be prevented with the trade-off of less cost effectiveness for the healthcare system.ConclusionsHealth economic modeling approaches can be augmented with machine learning methods, which broadens the areas in which evidence can be generated for policy makers to base their budget allocation. The precision of such predictive machine learning models must be high enough to be able to improve a care-as-usual healthcare system. Machine learning models generally let you set the level of precision acquired, at the cost of a possibly low recall, thereby limiting the impact on the healthcare system as a whole. More and better data for training these machine learning models will allow developed models to better distinguish patients who will and won’t develop a recurrent depressive episode, and for higher recall given a desired precision threshold. This will translate into a more substantial improvement in the treatment of depressive disorders in the healthcare system.

Developing Topics.

The growth of social media and the continuous improvement of machine-learning algorithms suggest that social media-based screening methods for mental diseases will become increasingly feasible with high accuracy in the next few years. Additionally, artificial intelligence, particularly predictive machine learning (ML) Models, has been established as one of the more powerful approaches to building reliable models that might be useful as an early predictor for Mental disorders. Specifically, one of the current challenges in brain disorders is identifying patients with mild cognitive impairment (MCI) that might be converted to Alzheimer's (AD) or other types of dementia. Cognitive decline in older adults is associated with decreased social interaction and changes in other features. We aim to use vast quantities of historical Facebook data from participants who conducted the MoCA test as a cognitive assessment test to examine if there are changes in their language mistakes and in a range of psychological features derived from natural language processing use and social engagement over time as a sign of MCI. To develop a Machine learning model to predict MCI using (n = 225), we should first consider how we can formulate Language mistakes and social interactions. Then, we aim to perform a longitudinal and cross-sectional examination of the data to examine the changes in a range of features extracted from Facebook subjects' data over time, using NLP algorithms, LIWC 2022, Go-emotion dataset, and supervised Classification ML tools, such as support vector machine model (SVM). We are collecting identified Facebook data, de-identifying it, translating the Arabic text into English, processing text augmentation and text preprocessing, defining language mistakes, social engagement over time features, and sentiments impeded in each syntax. Our classification model SVM shows significant prediction results, differentiating between patients (MCI& dementia) and normal subjects, using a sample of 225 subjects, resulting that the SVM prediction test score is 70%. Using our suggested pipeline, Our classification model SVM shows promising primary results in building a predictive model that has the ability for early detection of MCI.

Predictive machine learning applying cross industry standard process for data mining for the diagnosis of diabetes mellitus type 2

&lt;p&gt;Currently, type 2 diabetes mellitus is one of the world's most prevalent&lt;br /&gt;diseases and has claimed millions of people's lives. The present research aims&lt;br /&gt;to know the impact of the use of machine learning in the diagnostic process&lt;br /&gt;of type 2 diabetes mellitus and to offer a tool that facilitates the diagnosis of&lt;br /&gt;the dis-ease quickly and easily. Different machine learning models were&lt;br /&gt;designed and compared, being random forest was the algorithm that generated&lt;br /&gt;the model with the best performance (90.43% accuracy), which was integrated&lt;br /&gt;into a web platform, working with the PIMA dataset, which was validated by&lt;br /&gt;specialists from the Peruvian League for the Fight against Diabetes&lt;br /&gt;organization. The result was a decrease of (A) 88.28% in the information&lt;br /&gt;collection time, (B) 99.99% in the diagnosis time, (C) 44.42% in the diagnosis&lt;br /&gt;cost, and (D) 100% in the level of difficulty, concluding that the application&lt;br /&gt;of machine learning can significantly optimize the diagnostic process of type&lt;br /&gt;2 diabetes mellitus.&lt;/p&gt;

Explainable machine learning for profiling the immunological synapse and functional characterization of therapeutic antibodies

Therapeutic antibodies are widely used to treat severe diseases. Most of them alter immune cells and act within the immunological synapse; an essential cell-to-cell interaction to direct the humoral immune response. Although many antibody designs are generated and evaluated, a high-throughput tool for systematic antibody characterization and prediction of function is lacking. Here, we introduce the first comprehensive open-source framework, scifAI (single-cell imaging flow cytometry AI), for preprocessing, feature engineering, and explainable, predictive machine learning on imaging flow cytometry (IFC) data. Additionally, we generate the largest publicly available IFC dataset of the human immunological synapse containing over 2.8 million images. Using scifAI, we analyze class frequency and morphological changes under different immune stimulation. T cell cytokine production across multiple donors and therapeutic antibodies is quantitatively predicted in vitro, linking morphological features with function and demonstrating the potential to significantly impact antibody design. scifAI is universally applicable to IFC data. Given its modular architecture, it is straightforward to incorporate into existing workflows and analysis pipelines, e.g., for rapid antibody screening and functional characterization.

Radiomics and Clinicopathological Characteristics for Predicting Lymph Node Metastasis in Testicular Cancer

Accurate prediction of lymph node metastasis (LNM) in patients with testicular cancer is highly relevant for treatment decision-making and prognostic evaluation. Our study aimed to develop and validate clinical radiomics models for individual preoperative prediction of LNM in patients with testicular cancer. We enrolled 91 patients with clinicopathologically confirmed early-stage testicular cancer, with disease confined to the testes. We included five significant clinical risk factors (age, preoperative serum tumour markers AFP and B-HCG, histotype and BMI) to build the clinical model. After segmenting 273 retroperitoneal lymph nodes, we then combined the clinical risk factors and lymph node radiomics features to establish combined predictive models using Random Forest (RF), Light Gradient Boosting Machine (LGBM), Support Vector Machine Classifier (SVC), and K-Nearest Neighbours (KNN). Model performance was assessed by the area under the receiver operating characteristic (ROC) curve (AUC). Finally, the decision curve analysis (DCA) was used to evaluate the clinical usefulness. The Random Forest combined clinical lymph node radiomics model with the highest AUC of 0.95 (±0.03 SD; 95% CI) was considered the candidate model with decision curve analysis, demonstrating its usefulness for preoperative prediction in the clinical setting. Our study has identified reliable and predictive machine learning techniques for predicting lymph node metastasis in early-stage testicular cancer. Identifying the most effective machine learning approaches for predictive analysis based on radiomics integrating clinical risk factors can expand the applicability of radiomics in precision oncology and cancer treatment.

Predictive machine learning model for mental health issues in higher education students due to COVID-19 using HADS assessment

PurposeStudents pursuing different professional courses at the higher education level during 2021–2022 saw the first-time occurrence of a pandemic in the form of coronavirus disease 2019 (COVID-19), and their mental health was affected. Many works are available in the literature to assess mental health severity. However, it is necessary to identify the affected students early for effective treatment.Design/methodology/approachPredictive analytics, a part of machine learning (ML), helps with early identification based on mental health severity levels to aid clinical psychologists. As a case study, engineering and medical course students were comparatively analysed in this work as they have rich course content and a stricter evaluation process than other streams. The methodology includes an online survey that obtains demographic details, academic qualifications, family details, etc. and anxiety and depression questions using the Hospital Anxiety and Depression Scale (HADS). The responses acquired through social media networks are analysed using ML algorithms – support vector machines (SVMs) (robust handling of health information) and J48 decision tree (DT) (interpretability/comprehensibility). Also, random forest is used to identify the predictors for anxiety and depression.FindingsThe results show that the support vector classifier produces outperforming results with classification accuracy of 100%, 1.0 precision and 1.0 recall, followed by the J48 DT classifier with 96%. It was found that medical students are affected by anxiety and depression marginally more when compared with engineering students.Research limitations/implicationsThe entire work is dependent on the social media-displayed online questionnaire, and the participants were not met in person. This indicates that the response rate could not be evaluated appropriately. Due to the medical restrictions imposed by COVID-19, which remain in effect in 2022, this is the only method found to collect primary data from college students. Additionally, students self-selected themselves to participate in this survey, which raises the possibility of selection bias.Practical implicationsThe responses acquired through social media networks are analysed using ML algorithms. This will be a big support for understanding the mental issues of the students due to COVID-19 and can taking appropriate actions to rectify them. This will improve the quality of the learning process in higher education in Oman.Social implicationsFurthermore, this study aims to provide recommendations for mental health screening as a regular practice in educational institutions to identify undetected students.Originality/valueComparing the mental health issues of two professional course students is the novelty of this work. This is needed because both studies require practical learning, long hours of work, etc.

Quantitative patient graph analysis for transient ischemic attack risk factor distribution based on electronic medical records

A transient ischemic attack (TIA) affects millions of people worldwide. Although TIA risk factors have been identified individually, a systemic quantitative analysis of all health factors relevant to TIA using electronic medical records (EMR) remains lacking. This study employed a data-driven approach, leveraging hospital EMR data to create a TIA patient health factor graph. This graph consisted of 737 TIA and 737 control patient nodes, 740 health factor nodes, and over 33,000 relations between patients and factors. For all health factors in the graph, the connection delta ratios (CDRs) were determined and ranked, generating a quantitative distribution of TIA health factors. A literature review confirmed 56 risk factors in the distribution and unveiled a potential new risk factor “rhinosinusitis” for future validation. Moreover, the patient graph was visualized together with the TIA knowledge graph in the Unified Medical Language System. This integration enables clinicians to access and visualize patient data and international standard knowledge within a unified graph. In conclusion, graph CDR analysis can effectively quantify the distribution of TIA risk factors. The resulting TIA risk factor distribution might be instrumental in developing new risk prediction machine learning models for screening and early detection of TIA.