Classification Model Research Articles

Unsupervised Domain Adaptation of Object Detectors: A Survey.

Recent advances in deep learning have led to the development of accurate and efficient models for various computer vision applications such as classification, segmentation, and detection. However, learning highly accurate models relies on the availability of large-scale annotated datasets. Due to this, model performance drops drastically when evaluated on label-scarce datasets having visually distinct images, termed as domain adaptation problem. There are a plethora of works to adapt classification and segmentation models to label-scarce target dataset through unsupervised domain adaptation. Considering that detection is a fundamental task in computer vision, many recent works have focused on developing novel domain adaptive detection techniques. Here, we describe in detail the domain adaptation problem for detection and present an extensive survey of the various methods. Furthermore, we highlight strategies proposed and the associated shortcomings. Subsequently, we identify multiple aspects of the problem that are most promising for future research. We believe that this survey shall be valuable to the pattern recognition experts working in the fields of computer vision, biometrics, medical imaging, and autonomous navigation by introducing them to the problem, and familiarizing them with the current status of the progress while providing promising directions for future research.

SAR target recognition through adaptive kernel sparse representation model based on local contrast perception

Sparse representation-based methods have shown promising prospects in synthetic aperture radar (SAR) target recognition due to their robustness under complex application conditions. However, these methods are still susceptible to the feature representation of SAR images. In this paper, we propose a new SAR target recognition method using an adaptive kernel sparse representation model based on local contrast perception (LCP). The LCP method combines a specially designed random projection matrix with two-dimensional filtering to capture extensive intensity contrasts between local regions in SAR images. A response map reconstruction technique shows that LCP can obtain discriminative properties between target and background regions and improve feature representation. The LCP features are then transformed into the kernel representation to improve linear separability by defining a reproducing kernel Hilbert space (RKHS) and using a nonlinear kernel function. The kernel representation is utilized to construct a new kernel sparse representation model for classification. Furthermore, neighborhood analysis (NA) is proposed to establish the relationship between the parameter and feature distribution, with an adaptive regularization parameter selection strategy to mitigate the parameter sensitivity. Experiments conducted on the moving and stationary target acquisition and recognition dataset demonstrate the superiority and robustness of the proposed method under different application conditions.

Breast cancer detection through attention based feature integration model

<span lang="EN-US">Breast cancer is detected by screening mammography wherein X-rays are used to produce images of the breast. Mammograms for screening can detect breast cancer early. This research focuses on the challenges of using multi-view mammography to diagnose breast cancer. By examining numerous perspectives of an image, an attention-based feature-integration mechanism (AFIM) model that concentrates on local abnormal areas associated with cancer and displays the essential features considered for evaluation, analyzing cross-view data. This is segmented into two views the bi-lateral attention module (BAM) module integrates the left and right activation maps for a similar projection is used to create a spatial attention map that highlights the impact of asymmetries. Here the module's focus is on data gathering through medio-lateral oblique (MLO) and bilateral craniocaudal (CC) for each breast to develop an attention module. The proposed AFIM model generates using spatial attention maps obtained from the identical image through other breasts to identify bilaterally uneven areas and</span><span lang="EN-US">class activation map (CAM) generated from two similar breast images to emphasize the feature channels connected to a single lesion in a breast. AFIM model may easily be included in ResNet-style architectures to develop multi-view classification models.</span>

TSSG-CNN: A Tuberculosis Semantic Segmentation-Guided Model for Detecting and Diagnosis Using the Adaptive Convolutional Neural Network.

Tuberculosis (TB) is an infectious disease caused by Mycobacterium. It primarily impacts the lungs but can also endanger other organs, such as the renal system, spine, and brain. When an infected individual sneezes, coughs, or speaks, the virus can spread through the air, which contributes to its high contagiousness. The goal is to enhance detection recognition with an X-ray image dataset. This paper proposed a novel approach, named the Tuberculosis Segmentation-Guided Diagnosis Model (TSSG-CNN) for Detecting Tuberculosis, using a combined semantic segmentation and adaptive convolutional neural network (CNN) architecture. The proposed approach is distinguished from most of the previously proposed approaches in that it uses the combination of a deep learning segmentation model with a follow-up classification model based on CNN layers to segment chest X-ray images more precisely as well as to improve the diagnosis of TB. It contrasts with other approaches like ILCM, which is optimized for sequential learning, and explainable AI approaches, which focus on explanations. Moreover, our model is beneficial for the simplified procedure of feature optimization from the perspectives of approach using the Mayfly Algorithm (MA). Other models, including simple CNN, Batch Normalized CNN (BN-CNN), and Dense CNN (DCNN), are also evaluated on this dataset to evaluate the effectiveness of the proposed approach. The performance of the TSSG-CNN model outperformed all the models with an impressive accuracy of 98.75% and an F1 score of 98.70%. The evaluation findings demonstrate how well the deep learning segmentation model works and the potential for further research. The results suggest that this is the most accurate strategy and highlight the potential of the TSSG-CNN Model as a useful technique for precise and early diagnosis of TB.

Predicting the Risk of Breast Cancer Recurrence and Metastasis based on miRNA Expression

Background: Even after surgery, breast cancer patients still suffer from recurrence and metastasis. Thus, it is critical to predict accurately the risk of recurrence and metastasis for individual patients, which can help determine the appropriate adjuvant therapy. Methods: The purpose of this study is to investigate and compare the performance of several categories of molecular biomarkers, i.e., microRNA (miRNA), long non-coding RNA (lncRNA), messenger RNA (mRNA), and copy number variation (CNV), in predicting the risk of breast cancer recurrence and metastasis. First, the molecular data (miRNA, lncRNA, mRNA, and CNV) of 483 breast cancer patients were downloaded from the Cancer Genome Atlas, which were then randomly divided into the training and test sets with a ratio of 7:3. Second, the feature selection process was applied by univariate Cox and multivariate Cox variance analysis on the training set (e.g., 15 miRNAs). According to the selected features (e.g., 15 miRNAs), a random forest classifier and several other classification methods were established according to the label of recurrence and metastasis. Finally, the performances of the classification models were compared and evaluated on the test set. Results: The area under the ROC curve was 0.70 for miRNA, better than those using other biomarkers. Conclusion: These results indicated that miRNA has important guiding significance in predicting recurrence and metastasis of breast cancer.

Prediction of Interactomic HUB Genes in Periodontitis With Acute Myocardial Infarction.

Acute myocardial infarction (AMI) risk correlates with C-reactive protein (CRP) levels, suggesting systemic inflammation is present well before AMI. Studying different types of periodontal disease (PD), extremely common in individuals at risk for AMI, has been one important research topic. According to recent research, AMI and PD interact via the systemic production of certain proinflammatory and anti-inflammatory cytokines, small signal molecules, and enzymes that control the onset and development of both disorders' chronic inflammatory reactions. This study uses machine learning to identify the interactome hub biomarker genes in acute myocardial infarction and periodontitis. GSE208194 and GSE222883 were chosen for our research after a thorough search using keywords related to the study's goal from the gene expression omnibus (GEO) datasets. DEGs were identified from the GEOR tool, and the hub gene was identified using Cytoscape-cytohubba. Using expression values, Random Forest, Adaptive Boosting, and Naive Bayes, widgets-generated transcriptomics data, were labelled, and divided into 80/20 training and testing data with cross-validation. ROC curve, confusion matrix, and AUC were determined. In addition, Functional Enrichment Analysis of Differentially Expressed Gene analysis was performed. Random Forest, AdaBoost, and Naive Bayes models with 99%, 100%, and 75% AUC, respectively. Compared to RF, AdaBoost, and NB classification models, AdaBoost had the highest AUC. Categorization algorithms may be better predictors than important biomarkers. Machine learning model predicts hub and non-hub genes from genomic datasets with periodontitis and acute myocardial infarction.

Performance Evaluation of a Supervised Machine Learning Pain Classification Model Developed by Neonatal Nurses.

Early-life pain is associated with adverse neurodevelopmental consequences; and current pain assessment practices are discontinuous, inconsistent, and highly dependent on nurses' availability. Furthermore, facial expressions in commonly used pain assessment tools are not associated with brain-based evidence of pain. To develop and validate a machine learning (ML) model to classify pain. In this retrospective validation study, using a human-centered design for Embedded Machine Learning Solutions approach and the Neonatal Facial Coding System (NFCS), 6 experienced neonatal intensive care unit (NICU) nurses labeled data from randomly assigned iCOPEvid (infant Classification Of Pain Expression video) sequences of 49 neonates undergoing heel lance. NFCS is the only observational pain assessment tool associated with brain-based evidence of pain. A standard 70% training and 30% testing split of the data was used to train and test several ML models. NICU nurses' interrater reliability was evaluated, and NICU nurses' area under the receiver operating characteristic curve (AUC) was compared with the ML models' AUC. Nurses weighted mean interrater reliability was 68% (63%-79%) for NFCS tasks, 77.7% (74%-83%) for pain intensity, and 48.6% (15%-59%) for frame and 78.4% (64%-100%) for video pain classification, with AUC of 0.68. The best performing ML model had 97.7% precision, 98% accuracy, 98.5% recall, and AUC of 0.98. The pain classification ML model AUC far exceeded that of NICU nurses for identifying neonatal pain. These findings will inform the development of a continuous, unbiased, brain-based, nurse-in-the-loop Pain Recognition Automated Monitoring System (PRAMS) for neonates and infants.

Classification of tea leaf disease using convolutional neural network approach

Leaf diseases on tea plants affect the quality of tea. This issue must be overcome since preparing tea drinks requires high-quality tea leaves. Various automatic models for identifying disease in tea leaves have been developed; however, their performance is typically low since the extracted features are not selective enough. This work presents a classification model for tea leaf disease that distinguishes six leaf classes: algal spot, brown, blight, grey blight, helopeltis, red spot, and healthy. Deep learning using a convolutional neural network (CNN) builds an effective model for detecting tea leaf illness. The Kaggle public dataset contains 5,980 tea leaf images on a white background. Pre-processing was performed to reduce computing time, which involved shrinking and normalizing the image prior to augmentation. Augmentation techniques included rotation, shear, flip horizontal, and flip vertical. The CNN model was used to classify tea leaf disease using the MobileNetV2 backbone, Adam optimizer, and rectified linear unit (ReLU) activation function with 224×224 input data. The proposed model attained the highest performance, as evidenced by the accuracy value 0.9455.

Hybrid deep learning model for YouTube spam comment detection

Social media platforms, including YouTube and Facebook, allow users to interact through comments and videos. However, the openness of these platforms also makes them susceptible to spammers engaging in phishing, malware distribution, and advertisement dissemination. In response, our study introduces an innovative technique for detecting features indicative of spam within comments associated with shared videos. The initial phase involves data collection from the University of California, Irvine (UCI) machine learning repository and preprocessing using tokenization and lemmatization. Subsequently, a rigorous feature selection process is executed, and experiments are conducted with various proposed classification models. The performance evaluation demonstrates outstanding accuracy in identifying spam comments on YouTube: convolutional neural network with gated recurrent unit (CNN-GRU) at 95.92%, convolutional neural network with long short-term memory (CNN-LSTM) at 95.41%, convolutional neural network with bidirectional long short-term memory (CNN-biLSTM) at 96.43%, gated recurrent unit (GRU) at 95.41%, long short-term memory (LSTM) at 94.13%, and bidirectional long short-term memory (biLSTM) at 96.94% and convolutional neural network (CNN) at 94.64%. These results highlight the substantial contribution of our approach to spam detection and the fortification of online security.

Improved performance of fake account classifiers with percentage overlap features selection

Feature selection plays a crucial role in the development of high-performance classification models. We propose an innovative method for detecting fake accounts. This method leverages the percentage overlap technique to refine feature selection. We introduce our technique upon earlier work that showcased the enhanced efficacy of the Naïve Bayesian classifier through dataset normalization. Our study employs a dataset of account profiles sourced from Twitter, which we normalize using the Min-Max method. We analyze the results through a series of comprehensive experiments involving diverse classification algorithms—such as Naïve Bayes, decision tree, k-nearest neighbors (KNN), deep learning, and support vector machines (SVM). Our experimental results demonstrate a 100% accuracy achieved by the SVM and deep learning classifiers. The results are attributed to the percentage overlap technique, which facilitates the identification of four highly informative features. These findings outperform models with more extensive feature sets, underscoring the efficacy of our approach.

Classification of Soybean Genotypes as to Calcium, Magnesium, and Sulfur Content Using Machine Learning Models and UAV–Multispectral Sensor

Making plant breeding programs less expensive, fast, practical, and accurate, especially for soybeans, promotes the selection of new soybean genotypes and contributes to the emergence of new varieties that are more efficient in absorbing and metabolizing nutrients. Using spectral information from soybean genotypes combined with nutritional information on secondary macronutrients can help genetic improvement programs select populations that are efficient in absorbing and metabolizing these nutrients. In addition, using machine learning algorithms to process this information makes the acquisition of superior genotypes more accurate. Therefore, the objective of the work was to verify the classification performance of soybean genotypes regarding secondary macronutrients by ML algorithms and different inputs. The experiment was conducted in the experimental area of the Federal University of Mato Grosso do Sul, municipality of Chapadão do Sul, Brazil. Soybean was sown in the 2019/20 crop season, with the planting of 103 F2 soybean populations. The experimental design used was randomized blocks, with two replications. At 60 days after crop emergence (DAE), spectral images were collected with a Sensifly eBee RTK fixed-wing remotely piloted aircraft (RPA), with autonomous takeoff control, flight plan, and landing. At the reproductive stage (R1), three leaves were collected per plant to determine the macronutrients calcium (Ca), magnesium (Mg), and sulfur (S) levels. The data obtained from the spectral information and the nutritional values of the genotypes in relation to Ca, Mg, and S were subjected to a Pearson correlation analysis; a PC analysis was carried out with a k-means algorithm to divide the genotypes into clusters. The clusters were taken as output variables, while the spectral data were used as input variables for the classification models in the machine learning analyses. The configurations tested in the models were spectral bands (SBs), vegetation indices (VIs), and a combination of both. The combination of machine learning algorithms with spectral data can provide important biological information about soybean plants. The classification of soybean genotypes according to calcium, magnesium, and sulfur content can maximize time, effort, and labor in field evaluations in genetic improvement programs. Therefore, the use of spectral bands as input data in random forest algorithms makes the process of classifying soybean genotypes in terms of secondary macronutrients efficient and important for researchers in the field.

Machine learning approach with a posteriori-based feature to predict service life of a thermal cracking furnace with coking deposition

A thermal cracking furnace is an important equipment in the petrochemical industry that is typically used for breaking long hydrocarbons into short chains and producing coke as a byproduct. Deposition of the generated coke increases the temperature at the outside coil wall, necessitating regular furnace maintenance to prevent coil failure. Therefore, this study proposed a machine learning approach with a posteriori-based feature to predict the service life of the furnace to runtime failure. The proposed approach consists of a two-level machine learning model, which aims to improve prediction accuracy and reduce feature sensitivity. The label is classified as a week-range label, which can be categorized by classification criteria into three classes: weekly, bi-weekly, and quarter-weekly. The first-level model is utilized to extract sensor features into the posterior probability class label score. These scores are then processed and sorted into moving windows to generate features for the second-level model. The results showed that the proposed model could extract process variation and identify service needs, which improved classification accuracy by 23.94 % and 17.67 % for the clean and coke-contaminated datasets compared to the conventional classification model, respectively. Additionally, the most general class range (quarter-weekly) provided the best performance compared to the bi-weekly and weekly classes. Therefore, the model has the potential for the prediction of the furnace service life under pseudo-steady state conditions, where coking evolves gradually over time.

Advanced CKD detection through optimized metaheuristic modeling in healthcare informatics

Data categorization is a top concern in medical data to predict and detect illnesses; thus, it is applied in modern healthcare informatics. In modern informatics, machine learning and deep learning models have enjoyed great attention for categorizing medical data and improving illness detection. However, the existing techniques, such as features with high dimensionality, computational complexity, and long-term execution duration, raise fundamental problems. This study presents a novel classification model employing metaheuristic methods to maximize efficient positives on Chronic Kidney Disease diagnosis. The medical data is initially massively pre-processed, where the data is purified with various mechanisms, including missing values resolution, data transformation, and the employment of normalization procedures. The focus of such processes is to leverage the handling of the missing values and prepare the data for deep analysis. We adopt the Binary Grey Wolf Optimization method, a reliable subset selection feature using metaheuristics. This operation is aimed at improving illness prediction accuracy. In the classification step, the model adopts the Extreme Learning Machine with hidden nodes through data optimization to predict the presence of CKD. The complete classifier evaluation employs established measures, including recall, specificity, kappa, F-score, and accuracy, in addition to the feature selection. Data related to the study show that the proposed approach records high levels of accuracy, which is better than the existing models.

Endovascular Detection of Catheter-Thrombus Contact by Vacuum Excitation.

The objective of this work is to introduce and demonstrate the effectiveness of a novel sensing modality for contact detection between an off-the-shelf aspiration catheter and a thrombus. A custom robotic actuator with a pressure sensor was used to generate an oscillatory vacuum excitation and sense the pressure inside the extracorporeal portion of the catheter. Vacuum pressure profiles and robotic motion data were used to train a support vector machine (SVM) classification model to detect contact between the aspiration catheter tip and a mock thrombus. Validation consisted of benchtop accuracy verification, as well as user study comparison to the current standard of angiographic presentation. Benchtop accuracy of the sensing modality was shown to be 99.67%. The user study demonstrated statistically significant improvement in identifying catheter-thrombus contact compared to the current standard. The odds ratio of successful detection of clot contact was 2.86 (p = 0.03) when using the proposed sensory method compared to without it. The results of this work indicate that the proposed sensing modality can offer intraoperative feedback to interventionalists that can improve their ability to detect contact between the distal tip of a catheter and a thrombus. By offering a relatively low-cost technology that affords off-the-shelf aspiration catheters as clot-detecting sensors, interventionalists can improve the first-pass effect of the mechanical thrombectomy procedure while reducing procedural times and mental burden.

A compact breath breathalyzer for identifying the non-alcoholic fatty liver disease biomarker

Non-alcoholic fatty liver disease (NAFLD) is a prevalent chronic liver disease worldwide. Currently, its diagnosis relies primarily on imaging and histological examinations, which are invasive and prone to misdiagnosis in the early stage. To address these limitations, detection and analysis of volatile organic compounds (VOCs) in human breath can be a rapid and non-invasive screening method for NAFLD. In this study, a compact breath breathalyzer was developed, utilizing a miniaturized gas chromatography chip with the STM32 microcontroller as the main control chip to manage airflow, temperature, and receive terminal signals from the photoionization detector. In the experiment, a gas mixture comprising five VOCs (pentane, acetone, toluene, octane, and decane) was selected as the simulated typical disease biomarkers in human breath to investigate the breathalyzer's performance and optimize testing conditions for multi-polar and wide-boiling-range breath samples. Results show that the breathalyzer can detect low-boiling components (< 100°C) such as the isoprene and acetone, with a detection limit less than 50 ppb which are two commonly biomarkers of NAFLD. Furthermore, breath samples were collected from 35 non-diseased individuals, and NAFLD early-stage patient samples were simulated by increasing the isoprene concentration by 10 ppb. Convolutional neural network (CNN) were used to identify the VOC signatures in gas chromatograms with predictive accuracy of 85% for the classification model. Therefore, the compact breath breathalyzer has potential application in the rapid and early screening of NAFLD.

Does the Clusters of High-speed Railway Network Match the Urban Agglomerations? A Case Study in China

High-speed railways play a crucial role in intercity transportation, shaping the spatial interactions within urban agglomerations. However, a mismatch between transportation clusters and the spatial structure of urban agglomerations hampers regional economic integration. To address this issue, this paper proposes a community classification model, employing the Louvain algorithm, to analyze the aggregation of high-speed railway networks (HSRN). It compares central cities in city clusters with HSRN communities in relevant areas, examining the influence of central cities on urban agglomerations. The model is applied to a case study of China's high-speed railway network. The community classifications of HSRN are then compared with planned or future urban agglomerations at the national, provincial, and civic levels. The findings reveal that HSRN in China exhibits distinct clusters that align with the distribution of urban agglomerations. The existing central cities within these agglomerations provide good service levels in the HSRN network. However, HSRN communities within each urban agglomeration are overly scattered and should be integrated to reduce accessibility inequalities. This study offers suggestions for the development planning of HSRN and contributes to the economic integration of urban agglomerations.

Differential effects of mass-loading the eardrum and stiffening the middle ear on wideband absorbance.

The current work investigated the effects of mass-loading the eardrum on wideband absorbance in humans. A non-invasive approach to mass-loading the eardrum was utilized in which water was placed on the eardrum via ear canal access. The mass-loaded absorbance was compared to absorbance measured for two alternative middle ear states: normal and stiffened. To stiffen the ear, subjects pressurized the middle ear through either exsufflation or insufflation concurrent with Eustachian tube opening. Mass-loading the eardrum was hypothesized to reduce high-frequency absorbance, whereas pressurizing the middle ear was hypothesized to reduce low- to mid-frequency absorbance. Discriminant linear analysis classification was performed to evaluate the utility of absorbance in differentiating between conditions. Water on the eardrum reduced absorbance over the 0.7- to 6-kHz frequency range and increased absorbance at frequencies below approximately 0.5 kHz; these changes approximated the pattern of changes reported in both hearing thresholds and stapes motion upon mass-loading the eardrum. Pressurizing the middle ear reduced the absorbance over the 0.125- to 4-kHz frequency range. Several classification models based on the absorbance in two- or three-frequency bands had accuracy exceeding 88%.

Harmonization radiomics model to predict immune checkpoint inhibitor-related pneumonitis (CIP) in patients with non-small cell lung cancer (NSCLC).

12142 Background: The increasing use of immunotherapy in advanced non-small cell lung cancer (NSCLC) presents a significant challenge in managing adverse events, particularly checkpoint inhibitor-associated pneumonitis (CIP). This potentially life-threatening complication often necessitates discontinuation of immunotherapy, even in patients experiencing tumor response. Presently, there are no reliable models to predict the onset of CIP. This research utilizes CT radiomics to create an innovative approach for anticipating the risk of CIP in patients with NSCLC. Methods: This IRB-approved, retrospective study analyzed data from 159 stage III-IV NSCLC patients undergoing immunotherapy. We categorized patients into pneumonitis (further subdivided into immunotherapy-induced, radiation-induced, and others) and non-pneumonitis groups. Using LIFEx software, we extracted 3D-radiomic features from both tumors and surrounding 1cm thick peritumoral regions. To address scanner-associated variations, a linear mixed-effect radiomics harmonization model was applied. A Random Forest algorithm was then used to develop a classification model predicting CIP occurrence based on the pre-treatment CT radiomic and clinical data. The dataset was split into training (70%) and test (30%) sets. The accuracy of predictions was evaluated using confusion matrix statistics and bootstrapping with 1,000 iterations for median and 95% confidence interval (CI). The performance metrics included sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the ROC curve (AUC). Results: 159 patients were analyzed, of which only 19 had checkpoint inhibitor-related pneumonitis. Among which, 17 (54.8%) patients had grade 1 pneumonitis, 12 (38.7%) patients had grade 2 pneumonitis, 2 (6.5%) patients had grade 3 pneumonitis, and none had grade 4 or grade 5 pneumonitis. The model achieved a sensitivity of 0.98 (0.97, 0.99), a specificity of 0.08 (0.05, 0.14), a PPV of 0.91 (0.90, 0.91), and an NPV of 0.33 ± 0.23 for predicting CIP. The AUC of 0.59 (0.56, 0.66) indicates that the model may predict checkpoint inhibitor-associated pneumonitis with an accuracy of 59%. Conclusions: The study provides insights into the potential of radiomic analysis coupled with AI algorithms and the use of harmonization model in predicting CIP among NSCLC patients treated with immunotherapy. However, larger studies are needed to validate our findings and the utility of harmonization models in predicting CIP.

Lung Tumor Classification Optimizer with Augment Input Images

This work focuses on the accurate classification of lung tumors in Computed Tomography (CT) images to determine tumor type and stage, essential for guiding treatment decisions. In this research, pretrained model VGG19 and Generative Adversarial Networks (GANs) models are used to classify the lung tumor in order to enhance the classification accuracy by increasing the diversity of the training dataset. The dataset comprises 1097 CT image samples, with 120 classified as benign, 561 as malignant, and 416 as normal cases. The coding for this work was developed using Python, with TensorFlow as the deep learning platform, and simulations were conducted on Google Colab. Performance metrics such as accuracy, sensitivity, specificity, and F1 score are evaluated to assess the effectiveness of the classification model. The datasets were obtained from the Iraq-Oncology Teaching Hospital.

Radiomics-based machine learning approach for the prediction of grade and stage in upper urinary tract urothelial carcinoma-a step towards virtual biopsy.

Upper tract urothelial carcinoma is a rare, aggressive lesion, with early detection a key to its management. This study aimed to utilise computed tomographic urogram data to develop machine learning models for predicting tumour grading and staging in upper urothelial tract carcinoma patients and to compare these predictions with histopathological diagnosis used as reference standards. Protocol-based computed tomographic urogram data from 106 patients were obtained and visualised in 3D. Digital segmentation of the tumours was conducted by extracting textural radiomics features. They were further classified using 11 predictive models. The predicted grades and stages were compared to the histopathology of radical nephroureterectomy specimens. Classifier models worked well in mining the radiomics data and delivered satisfactory predictive machine learning models. The MultiLayer Panel showed 84% sensitivity and 93% specificity while predicting upper tract urothelial carcinoma grades. The Logistic Regression model showed a sensitivity of 83% and a specificity of 76% while staging. Similarly, other classifier algorithms (e.g., Support Vector classifier) provided a highly accurate prediction while grading upper tract urothelial carcinoma compared to clinical features alone or ureteroscopic biopsy histopathology. Data mining tools could handle medical imaging datasets from small (<2cm) tumours for upper tract urothelial carcinoma. The radiomics-based machine learning algorithms provide a potential tool to model tumour grading and staging with implications for clinical practice and the upgradation of current paradigms in cancer diagnostics. Machine learning based on radiomics features can predict upper tract urothelial cancer grading and staging with significant improvement over ureteroscopic histopathology. The study showcased the prowess of such emerging tools in the set objectives with implications towards virtual biopsy.