Interpretation Ability Research Articles

Entropy-Regularized Iterative Weighted Shrinkage-Thresholding Algorithm (ERIWSTA) for inverse problems in imaging.

The iterative shrinkage-thresholding algorithm (ISTA) is a classic optimization algorithm for solving ill-posed linear inverse problems. Recently, this algorithm has continued to improve, and the iterative weighted shrinkage-thresholding algorithm (IWSTA) is one of the improved versions with a more evident advantage over the ISTA. It processes features with different weights, making different features have different contributions. However, the weights of the existing IWSTA do not conform to the usual definition of weights: their sum is not 1, and they are distributed over an extensive range. These problems may make it challenging to interpret and analyze the weights, leading to inaccurate solution results. Therefore, this paper proposes a new IWSTA, namely, the entropy-regularized IWSTA (ERIWSTA), with weights that are easy to calculate and interpret. The weights automatically fall within the range of [0, 1] and are guaranteed to sum to 1. At this point, considering the weights as the probabilities of the contributions of different attributes to the model can enhance the interpretation ability of the algorithm. Specifically, we add an entropy regularization term to the objective function of the problem model and then use the Lagrange multiplier method to solve the weights. Experimental results of a computed tomography (CT) image reconstruction task show that the ERIWSTA outperforms the existing methods in terms of convergence speed and recovery accuracy.

Subspace learning using low-rank latent representation learning and perturbation theorem: Unsupervised gene selection.

In recent years, gene expression data analysis has gained growing significance in the fields of machine learning and computational biology. Typically, microarray gene datasets exhibit a scenario where the number of features exceeds the number of samples, resulting in an ill-posed and underdetermined equation system. The presence of redundant features in high-dimensional data leads to suboptimal performance and increased computational time for learning algorithms. Although feature extraction and feature selection are two approaches that can be employed to deal with this challenge, feature selection has greater interpretability ability which causes it to receive more attention. In this study, we propose an unsupervised feature selection which is based on pseudo label latent representation learning and perturbation theory. In the first step, pseudo labels are extracted and constructed using latent representation learning. In the second step, the least square problem is solved for original data matrix and perturbed data matrix. Features are clustered based on the similarity between the original data matrix and the perturbed data matrix using k-means. In the last step, features in each subcluster are ranked based on information gain criterion. To showcase the efficacy of the proposed approach, numerical experiments were carried out on six benchmark microarray datasets and two RNA-Sequencing benchmark datasets. The outcomes indicate that the proposed technique surpasses eight state-of-the-art unsupervised feature selection methods in both clustering accuracy and normalized mutual information.

Integration of AI-Assisted in Digital Cervical Cytology Training: A Comparative Study.

This study aimed to investigate the supporting role of artificial intelligence (AI) in digital cervical cytology training. A total of 104 trainees completed both manual reading and AI-assisted reading tests following the AI-assisted digital training regimen. The interpretation scores and the testing time in different groups were compared. Also, the consistency, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of diagnoses were further analysed through the confusion matrix and inconsistency evaluation. The mean interpretation scores were significantly higher in the AI-assisted group compared with the manual reading group (81.97 ± 16.670 vs. 67.98 ± 21.469, p < 0.001), accompanied by a reduction in mean interpretation time (32.13 ± 11.740 min vs. 11.36 ± 4.782 min, p < 0.001). The proportion of trainees' results with complete consistence (Category O) were improved from 0.645 to 0.803 and the averaged pairwise κ scores were improved from 0.535 (moderate) to 0.731 (good) with AI assistance. The number of correct answers, accuracies, sensitivities, specificities, PPV, NPV and κ scores of most class-specific diagnoses (NILM, Fungi, HSV, LSIL, HSIL, AIS, AC) also improved with AI assistance. Moreover, 97.8% (89/91) of trainees reported substantial improvement in cervical cytology interpretation ability, and all participants (100%, 91/91) expressed a strong willingness to integrate AI-assisted diagnosis into their future practice. The utilisation of an AI-assisted digital cervical cytology training platform positively impacted trainee performance and received high satisfaction and acceptance among clinicians, suggesting its potential as a valuable adjunct to medical education.

基于改进轻量化卷积神经网络模型的玉露香梨缺陷识别研究

The skin of Yu-Lu-Xiang pears is brittle and easily damaged during picking and sorting. In order to reduce the secondary damage caused by mechanical automatic sorting of Yu-Lu-Xiang pear after harvest, optimize the sorting process and improve the sorting accuracy. Based on the MobileV2Net model, a lightweight convolutional neural network model EC-MobileV2Net-Fast, which integrated transfer learning and attention mechanism, was proposed to identify skin damage defects of Yu-Lu-Xiang pears. According to the defects of Yu-Lu-Xiang pears with different damage degrees, a dataset containing four characteristics was created. The model accuracy rate, single defect identification accuracy rate, recall, specificity, parameter and so on were taken as evaluation indexes, and the interpretation ability of the model decision was analyzed by Grad-CAM thermal map. Preliminary evaluation results showed that the model produced the highest level of accuracy, underscoring the potential of deep learning algorithms to significantly enhance defect recognition and classification. It can improve sorting efficiency, reduce labor costs and strictly control after-sales quality.

Comparative Analysis of Large Language Models and Spine Surgeons in Surgical Decision-Making and Radiological Assessment for Spine Pathologies

AimThis study aimed to investigate the accuracy of large language models (LLMs), specifically ChatGPT and Claude, in surgical decision-making and radiological assessment for spine pathologies compared to experienced spine surgeons. MethodsThe study employed a comparative analysis between the LLMs and a panel of attending spine surgeons. Five written clinical scenarios encompassing various spine pathologies were presented to the LLMs and surgeons, who provided recommended surgical treatment plans. Additionally, MRI images depicting spine pathologies were analyzed by the LLMs and surgeons to assess their radiological interpretation abilities. Spino-pelvic parameters were estimated from a scoliosis radiograph by the LLMs. ResultsQualitative content analysis revealed limitations in the LLMs' consideration of patient-specific factors and the breadth of treatment options. Both ChatGPT and Claude provided detailed descriptions of MRI findings but differed from the surgeons in terms of specific levels and severity of pathologies. The LLMs acknowledged the limitations of accurately measuring spino-pelvic parameters without specialized tools. The accuracy of surgical decision-making for the LLMs (20%) was lower than that of the attending surgeons (100%). Statistical analysis showed no significant differences in accuracy between the groups. ConclusionThe study highlights the potential of LLMs in assisting with radiological interpretation and surgical decision-making in spine surgery. However, the current limitations, such as the lack of consideration for patient-specific factors and inaccuracies in treatment recommendations, emphasize the need for further refinement and validation of these AI models. Continued collaboration between AI researchers and clinical experts is crucial to address these challenges and realize the full potential of AI in spine surgery.

Accurate Spatial Heterogeneity Dissection and Gene Regulation Interpretation for Spatial Transcriptomics using Dual Graph Contrastive Learning.

Recent advances in spatial transcriptomics have enabled simultaneous preservation of high-throughput gene expression profiles and the spatial context, enabling high-resolution exploration of distinct regional characterization in tissue. To effectively understand the underlying biological mechanisms within tissue microenvironments, there is a requisite for methods that can accurately capture external spatial heterogeneity and interpret internal gene regulation from spatial transcriptomics data. However, current methods for region identification often lack the simultaneous characterizing of spatial structure and gene regulation, thereby limiting the ability of spatial dissection and gene interpretation. Here, stDCL is developed, a dual graph contrastive learning method to identify spatial domains and interpret gene regulation in spatial transcriptomics data. stDCL adaptively incorporates gene expression data and spatial information via a graph embedding autoencoder, thereby preserving critical information within the latent embedding representations. In addition, dual graph contrastive learning is proposed to train the model, ensuring that the latent embedding representation closely resembles the actual spatial distribution and exhibits cluster similarity. Benchmarking stDCL against other state-of-the-art clustering methods using complex cortex datasets demonstrates its superior accuracy and effectiveness in identifying spatial domains. Ouranalysis of the imputation matrices generated by stDCL reveals its capability to reconstruct spatial hierarchical structures and refine differential expression assessment. Furthermore, it is demonstrated that the versatility of stDCL in interpretability of gene regulation, spatial heterogeneity at high resolution, and embryonic developmental patterns. In addition, it is also showed that stDCL can successfully annotate disease-associated astrocyte subtypes in Alzheimer's disease and unravel multiple relevant pathways and regulatorymechanisms.

Double Decomposition and Fuzzy Cognitive Graph-Based Prediction of Non-Stationary Time Series.

Deep learning models, such as recurrent neural network (RNN) models, are suitable for modeling and forecasting non-stationary time series but are not interpretable. A prediction model with interpretability and high accuracy can improve decision makers' trust in the model and provide a basis for decision making. This paper proposes a double decomposition strategy based on wavelet decomposition (WD) and empirical mode decomposition (EMD). We construct a prediction model of high-order fuzzy cognitive maps (HFCM), called the WE-HFCM model, which considers interpretability and strong reasoning ability. Specifically, we use the WD and EDM algorithms to decompose the time sequence signal and realize the depth extraction of the signal's high-frequency, low-frequency, time-domain, and frequency domain features. Then, the ridge regression algorithm is used to learn the HFCM weight vector to achieve modeling prediction. Finally, we apply the proposed WE-HFCM model to stationary and non-stationary datasets in simulation experiments. We compare the predicted results with the autoregressive integrated moving average (ARIMA) and long short-term memory (LSTM) models.For stationary time series, the prediction accuracy of the WE-HFCM model is about 45% higher than that of the ARIMA, about 35% higher than that of the SARIMA model, and about 16% higher than that of the LSTM model. For non-stationary time series, the prediction accuracy of the WE-HFCM model is 69% higher than that of the ARIMA and SARIMA models.

Research on temperature prediction model of molten steel of tundish in continuous casting

To achieve the desired superheat of molten steel during the continuous casting process, optimization of process parameters such as molten steel temperature in ladle furnace, casting speed, and baking temperature is necessary. Therefore, obtaining the superheat corresponding to these process parameters in advance is particularly important. To address this issue, a model for predicting the temperature of molten steel in the tundish during continuous casting is designed. The model adopts a combined modeling approach of mechanistic model and data model. To address the issue of the mechanism model’s inability to capture the variation of the lining’s thermal parameters, this article improves the traditional physics-informed neural network (PINN) algorithm. It combines the constraints from both the forward and inverse problems, allowing for obtaining solutions to the equations while capturing the variation of equation parameters. Actual data from multiple casting sequences at a steel plant are collected to validate the accuracy and interpretability of the model. The results show that the error of the model is about 2.1k which has better accuracy compared to pure mechanistic model and pure data model. Additionally, it can capture the variation patterns of tundish lining thermal parameters under different operating conditions. Therefore, the model designed in this article can provide both profound physical interpretation ability and more practical predictions of molten steel temperature.

A SOH estimation method of lithium-ion batteries based on partial charging data

Accurate and reliable assessment of lithium-ion battery SOH is critical to improving the safety performance of electric vehicles. In the existing data-driven methods, the models are often complex, the calculation is large and the interpretation ability is not strong, so it is difficult to realize the commercial application. In this paper, a novel SOH estimation method is proposed. Firstly, the incremental capacity (IC) curve is used to extract the aging features from part of the charging data, and a linear battery aging state space equation is established based on the features. The aging state of the battery can be directly observed under the condition of small calculation amount. Second, the aging state of the battery can be tracked and corrected by double Kalman filter to improve estimation accuracy. The performance of the proposed method is verified on CACLE dataset, Oxford dataset and DUT dataset respectively. The estimated errors are all <2.5 %, and the mean RMSE is 0.0066. The results show that the method can be effectively applied to different lithium-ion batteries. In order to further demonstrate the advantages of this estimator, we also compare it with literature methods using the same dataset. In the CACLE data set and Oxford data, the average time of this method is 0.848 s and the average RMSE is 0.0050, which is in a dominant position. The proposed method provides a simple and feasible way to estimate the SOH of electric vehicles.

Evaluating vision-capable chatbots in interpreting kinematics graphs: a comparative study of free and subscription-based models

This study investigates the performance of eight large multimodal model (LMM)-based chatbots on the Test of Understanding Graphs in Kinematics (TUG-K), a research-based concept inventory. Graphs are a widely used representation in STEM and medical fields, making them a relevant topic for exploring LMM-based chatbots’ visual interpretation abilities. We evaluated both freely available chatbots (Gemini 1.0 Pro, Claude 3 Sonnet, Microsoft Copilot, and ChatGPT-4o) and subscription-based ones (Gemini 1.0 Ultra, Gemini 1.5 Pro API, Claude 3 Opus, and ChatGPT-4). We found that OpenAI’s chatbots outperform all the others, with ChatGPT-4o showing the overall best performance. Contrary to expectations, we found no notable differences in the overall performance between freely available and subscription-based versions of Gemini and Claude 3 chatbots, with the exception of Gemini 1.5 Pro, available via API. In addition, we found that tasks relying more heavily on linguistic input were generally easier for chatbots than those requiring visual interpretation. The study provides a basis for considerations of LMM-based chatbot applications in STEM and medical education, and suggests directions for future research.

Estimation on compressive strength of recycled aggregate self-compacting concrete using interpretable machine learning-based models

PurposeRecycled aggregate self-compacting concrete (RASCC) has the potential for sustainable resource utilization and has been widely applied. Predicting the compressive strength (CS) of RASCC is challenging due to its complex composite nature and nonlinear behavior.Design/methodology/approachThis study comprehensively evaluated commonly used machine learning (ML) techniques, including artificial neural networks (ANN), random trees (RT), bagging and random forests (RF) for predicting the CS of RASCC. The results indicate that RF and ANN models typically have advantages with higher R2 values, lower root mean square error (RMSE), mean square error (MSE) and mean absolute error (MAE) values.FindingsThe combination of ML and Shapley additive explanation (SHAP) interpretable algorithms provides physical rationality, allowing engineers to adjust the proportion based on parameter analysis to predict and design RASCC. The sensitivity analysis of the ML model indicates that ANN’s interpretation ability is weaker than tree-based algorithms (RT, BG and RF). ML regression technology has high accuracy, good interpretability and great potential for predicting the CS of RASCC.Originality/valueML regression technology has high accuracy, good interpretability and great potential for predicting the CS of RASCC.

Comparison of Problem-based Learning and Didactic Lecture as a Teaching–Learning Method among Undergraduate Medical Students: An Interventional Study

Background Didactic lecture (DL) is a teaching method that has been there for a long time in our educational system, but student involvement is minimal. Problem-based learning (PBL) differs in that it involves students in the learning process. Various skills are learned, such as self-learning, group discussion, communication skills, teamwork and knowledge retention. Critical thinking, analysis and the ability to solve real problems are some benefits of PBL, which can help medical students understand clinical scenarios and analyse, interpret and come up with possible solutions based on sound knowledge. Thus, there is a need to assess the effectiveness of PBL in recall, interpretation and problem-solving skills as compared to the traditional lecture method, as there are very few Indian studies related to PBL in Indian medical students. Materials and Methods A non-randomised interventional study was carried out in the Department of Pharmacology at East Point Medical College, Bengaluru. Groups A and B, each with 60 medical students, participated in PBL and DL for either of the two competencies (myocardial infarction and iron deficiency anaemia). After the completion of the DL and PBL sessions, 20 multiple-choice questions (MCQs) were given to the students. The effectiveness of the teaching method was determined by comparing the mean scores of MCQs of both groups. Results The group scores were compared using an independent sample t-test and expressed as mean ± standard deviation. The total mean score of PBL for the topic myocardial infarction (11.00 ± 2.64, p value = .001) was statistically significant compared with the total mean score in the DL group (9.38 ± 2.52), and the total mean score of PBL for the topic iron deficiency anaemia (15.28 ± 3.53, p value = .001) was significant compared to the total mean score in the DL group (12.98 ± 3.61), respectively. The mean score for recall and problem-solving for the topic of myocardial infarction was statistically significant in the PBL group compared to the DL group. The mean score for recall, interpretation and problem-solving for the topic of iron deficiency anaemia was statistically significant in the PBL group compared with the DL group. Conclusion PBL showed marked improvement in recall, interpretation and problem-solving abilities compared with DL. So, including PBL as a teaching–learning method can promote active and student-centered learning.

AI driven interpretable deep learning based fetal health classification

In this study, a deep learning model is proposed for the classification of fetal health into 3 categories: Normal, suspect, and pathological. The primary objective is to utilize the power of deep learning to improve the efficiency and effectiveness of diagnostic processes. A deep neural network (DNN) model is proposed for fetal health analysis using data obtained from Cardiotocography (CTG). A dataset containing 21 attributes is used to carry out this work. The model incorporates multiple hidden layers, augmented with batch normalization and dropout layers for improved generalization. This study assesses the model's interpretation ability in fetal health classification using explainable deep learning. This enhances transparency in decision-making of the classifier model by leveraging feature importance and feature saliency analysis, fostering trust and facilitating the clinical adoption of fetal health assessments. Our proposed model demonstrates a remarkable performance with 0.99 accuracy, 0.93 sensitivity, 0.93 specificity, 0.96 AUC, 0.93 precision, and 0.93 F1 scores in classifying fetal health. We also performed comparative analysis with six other models including Logistic Regression, KNN, SVM, Naive Bayes, Random Forest, and Gradient Boosting to assess and compare the effectiveness of our model and the accuracies of 0.89, 0.88, 0.90, 081, 0.93, and 0.93 were achieved respectively by these baseline models. The results revealed that our proposed model outperformed all the baseline models in terms of accuracy. This indicates the potential of deep learning in improving fetal health assessment and contributing to the field of obstetrics by providing a robust tool for early risk detection.

Investigating the Applications of AI in Oncology from NIH Funded Projects: Case study of Lung Cancer and Pancreatic Cancer

ABSTRACTThis poster aims to conduct an examination of the applications of Artificial Intelligence (AI) in oncology. We collect projects from National Institutes of Health (NIH) which delve into how AI applications and techniques are used in cancer research between 2018 to 2024. Results show AI's ability to enhance lung and pancreatic cancer through novel imaging techniques and data processing. We find that refined medical imaging can be done with AI‐driven techniques such as machine learning or deep learning to improve diagnostic accuracy potentially leading to better survival rates for patients involved. Additionally, this reveals how AI transformed oncology studies underlining the movement toward complex diagnostic algorithms that allow accurate stage assignments. Consequently, these forms of teamwork are essential when determining as well as making reliable Cancer care systems by developing strong AI systems with interpretation abilities using supportive NIH funding.

Research on fault diagnosis and feature extraction mechanism visualization of rotating machinery based on improved 1D CNN

With the wide application of intelligent equipment in modern industrial production, it is particularly important to study how to intelligently sense the faults of rotating machinery, improve the diagnosis efficiency and enhance the interpretation ability of the diagnosis process. Although the traditional 1-D CNN performs well in fault diagnosis, it has limitations in capturing subtle changes and complex patterns of fault signals, and its interpretability needs to be improved. Therefore, based on the improved ELCNN model, this paper discusses its diagnosis mechanism in depth, aiming at providing a new idea for intelligent fault diagnosis of rotating machinery. The functions of convolutional layer and S-GAP layer in ELCNN are studied and analyzed. Through single-layer linear convolution, ELCNN can adaptively learn the frequency domain features of the signal and realize the lightweight of the model. At the same time, the S-GAP layer enhances the ability of ELCNN to capture the main peak frequency of fault signals through feature sparseness. The experimental results show that the accuracy of ELCNN in frequency domain feature extraction is more than 80 %. The main peak frequency extracted can effectively help engineers understand the basis of model judgment and improve the reliability of the model.

Geophysical archaeology for the study of complex historical-period burial spaces: An example from the Fort Lewis military cemetery, Colorado

Geophysical methods have long been used to locate and map unmarked graves in cemeteries and burial spaces. Despite their prevalence, geophysical cemetery surveys can be extremely complex, particularly when dealing with historical-period cemeteries. Not only are there interpretive challenges from site formation processes and noncultural features, but there are also cultural challenges when seeking to study a cemetery within its historical context. Geophysical archaeology is a specialization that can guide this type of work by underscoring the cultural context for geophysical data. Beyond simply locating unmarked graves or other “anomalies,” geophysical archaeology emphasizes increasing the interpretive ability of geophysical data by situating these surveys in cultural theory. Using the example of the Fort Lewis military cemetery, which also served the Fort Lewis Indian Boarding School, in southwest Colorado, this paper highlights how geophysical archaeology was used to identify specific features about the cemetery, including the original fence line, number of likely graves, and two different burial contexts (casketed burials and exhumed graves).

Breast cancer prognosis through the use of multi-modal classifiers: current state of the art and the way forward.

We present a survey of the current state-of-the-art in breast cancer detection and prognosis. We analyze the evolution of Artificial Intelligence-based approaches from using just uni-modal information to multi-modality for detection and how such paradigm shift facilitates the efficacy of detection, consistent with clinical observations. We conclude that interpretable AI-based predictions and ability to handle class imbalance should be considered priority.

Soil organic carbon sequestration can be promoted through the improvement of landscape configuration heterogeneity in typical agricultural regions of northeast China

Landscape heterogeneity is considered a promising option for building resilient and sustainable agroecosystems. Understanding the relationships between farmland soil organic carbon (SOC) and landscape heterogeneity can support soil carbon sequestration and better serve food security and climate change. However, the influence extent and optimal scale of farmland landscape heterogeneity (i.e. landscape composition and landscape configuration heterogeneity) on SOC remains unclear. In this study, we established the relationships between multi-scale landscape heterogeneity and SOC in a typical grain-production county of northeast China. Stepwise regression results showed that when the buffer radius was 2000 m, the interpretation of SOC by landscape heterogeneity was the largest. The effects of landscape composition and landscape configuration on SOC were further decomposed by Variance Partitioning Analysis, and we found that independent interpretation ability of landscape configuration (8%) exceeded landscape composition (7%). The result of soil mapping combined with landscape indexes also showed that landscape configuration contributed more to the increased accuracy. Moreover, we found that correlation between configuration indexes and SOC at the class level was less related than that at the landscape level, among which the two most important indexes were Mean Fractal Dimension Index (FRAC_MN) and Number of Patches (NP). FRAC_MN was even more important than natural factors, indicating the validity of landscape as an indicator of human activities should not be ignored when considering farmland SOC. Overall, the results of this study revealed that the negative effects of agricultural intensification on SOC can be buffered to a certain extent by increasing the complexity of patch shape and reducing the degree of landscape fragmentation at the landscape level, providing hope for the sustainable development in intensive agricultural areas. In addition, due to the scale effect of landscape heterogeneity on farmland SOC, we suggest that decision makers should consider the spatial scale in landscape allocation and planning. This study provides a scientific reference for realizing the balance between grain production and ecological function in intensive agricultural areas.

Narrative Techniques in English Literature: Enhancing Critical Thinking Skills in EFL (English as a Foreign Language) Classrooms

This article explores the role of narrative techniques in English literature as an effective tool for enhancing critical thinking skills in English as a Foreign Language (EFL) classrooms. With the growing emphasis on developing higher-order thinking skills, literature provides a rich context for fostering students' analytical, interpretative, and evaluative abilities. The study highlights key narrative techniques such as foreshadowing, irony, metaphor, and point of view, demonstrating how these elements encourage deeper engagement with texts. By analyzing narrative structures, students can challenge assumptions, explore different perspectives, and develop the ability to critically assess various situations. The article examines the integration of these literary techniques into EFL curricula, focusing on how they support language development alongside critical thinking. It also addresses potential challenges faced by both teachers and students, such as language proficiency barriers and the need for culturally relevant texts. Through case studies and classroom examples, the research offers practical strategies for EFL instructors to incorporate narrative techniques into their teaching, enhancing both language acquisition and cognitive skills. The findings suggest that teaching English literature through a focus on narrative elements not only improves linguistic competence but also cultivates critical thinking, preparing students to navigate complex real-world issues.

Developmet of Android E-book Forms of Energy to Improve Students' Critical Thinking and Mathematical Representation Skills

Critical thinking and mathematical representation skills play an important role for students, especially in physics. This research’ aims are to study the feasibility and effectiveness of the android e-book forms of energy to improve students' critical thinking and mathematical representation skills. This research was developed using the ADDIE model. The experimental design was one group pretest posttest. This research found that the E-book were feasible to be applied in physics learning. It content qualify the writing and content requirements for a physics E-book. The android e-book is effective for improving students' ability in interpretation and explanation, effective enough in analyze, and less effectively in the inference and argument and evaluation aspects. E-book is less effective in the aspect use pictures, graphs, tables, or diagrams to solve problems and effective enough in training students use symbols or mathematical equations in solving problems and explain data or other representations in sentence form. The global N-Gain value shows that the Android e-book forms of energy is effective enough to improve critical thinking and mathematical representation skills.