Arijit Bhattacharya

Polycystic Ovary Syndrome (PCOS) is a leading cause of female infertility and is associated with various health complications, including preterm abortions, anovulation, and ovarian cancer. It affects approximately 5% to 10% of women in their reproductive years. PCOS diagnosis often relies on ultrasound imaging to assess ovarian follicle size, count and arrangement. Accurately diagnosing PCOS in clinical practice poses significant challenges for radiologists due to the variability in follicle sizes and their complex relationships with surrounding blood vessels and tissues. This process is labor-intensive, prone to errors, and time-consuming. To address these challenges, numerous research efforts have focused on automating the detection of PCOS-affected ovaries. While advancements have been made, further improvements are needed to enhance diagnostic accuracy. Convolutional Neural Networks (CNNs) have shown promise in PCOS classification, but models relying solely on global features may achieve suboptimal results, as regional features are often overlooked. This paper introduces a feature fusion model named PCOSFusionNet designed to improve the accuracy of PCOS classification. The proposed system combines handcrafted features extracted using the Histogram of Oriented Gradients (HOG) descriptor with global features obtained from the VGG19 deep learning model. Additionally, Contrast Limited Adaptive Histogram Equalization (CLAHE) is applied during preprocessing to enhance image quality and improve feature extraction. The watershed method is employed for segmentation before classification. By integrating deep features with handcrafted features, the system achieves superior classification performance across multiple metrics, including accuracy, precision, recall, and F1-score, using five-fold cross-validation. The performance of the proposed PCOSFusionNet model was evaluated on two publicly available datasets. The first dataset (Dataset_1) contains 3856 ultrasound images and the second dataset (Dataset_2) comprises 12,680 ultrasound images. On these datasets, PCOSFusionNet achieved accuracies of 98.49% and 98.30%, respectively, surpassing existing state-of-the-art methods and demonstrating its effectiveness in PCOS diagnosis.

Detecting ovarian structures in ultrasound images is essential in gynecological and reproductive medicine. An automated detection system can serve as a valuable tool for physicians and assist in complex ultrasound interpretations. This study presents a CNN-based object detector designed to segment and count follicle regions in ovarian ultrasound images. Automated identification of ovarian follicles can aid in diagnosing conditions such as infertility, Polycystic Ovarian Syndrome (PCOS), ovarian cancer, and other reproductive health issues. The proposed model, Multi-Attention Residual Dilated UNet with Squeeze and Excitation (MARDSE-UNet), integrates residual UNet, dilated UNet, and squeeze-and-excitation blocks to enhance follicle detection performance. MARDSE-UNet achieved exceptional results, with 98.69% accuracy, 97.89% precision, 97.7% recall, an F1-score of 86.97%, and Intersection over Union (IoU) of 95.66% in follicle detection using 5-fold cross-validation. The USOVA3D dataset was used for experimentation. The model also incorporates a novel preprocessing stage to address noise and low contrast issues, as well as a post-processing stage to refine edges and extract features such as area, perimeter, and diameter of follicles for a more comprehensive performance comparison. The proposed model outperformed traditional CNN models and other state-of-the-art methods in comparative evaluations.

Missing value imputation is one of the major problems for any data-driven model. Credit risk prediction isn't any exception either. However, the selection of appropriate missing value imputation techniques for credit default prediction remains an underexplored area in existing literature. The paper introduces ‘CRES-SMVI,' a ‘Credit Risk Expert System based on Selective Missing Value Imputation,' developed using ensemble learning techniques. The model integrates advanced preprocessing techniques, including SMOTE for handling imbalanced data and seven missing value imputation methods for effective comparison. By employing multiple classifiers such as Decision Tree (DT), Catboost (CB), and Extra Trees (ET), the system aimed to provide accurate credit default predictions. The study evaluated twenty-one models in total and finally found the best among them. As per the study, Multivariate imputation by chained equations (MICE)-based ET model performed better than all the other models, irrespective of the datasets.

This paper explores the application of stacking models for breast cancer detection, integrating key techniques such as data balancing, hyperparameter tuning, and feature selection. We implemented five different stacking configurations. Initially, Logistic Regression (LR) was used as the meta-classifier, while the base estimators included Decision Tree (DT), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Random Forest (RF) classifiers. In the second configuration, we reversed the roles: DT acted as the meta-classifier, with SVM, KNN, RF, and LR serving as the base estimators. In a third setup, SVM was used as the meta-classifier, with DT, LR, KNN, and RF as the base learners. Fourth, we implemented KNN as the stacking classifier, with LR, DT, SVM, and RF as the base estimators. Finally, in the fifth configuration, RF was the meta-classifier, supported by LR, DT, KNN, and SVM as base learners. The evaluation of stacking models was conducted in five phases, starting with a baseline with no adjustments, followed by applying data balancing alone, then adding hyperparameter tuning, applying Chi-square feature selection with data balancing, and finally using correlation-based feature selection with data balancing, all systematically excluding certain elements to analyze their individual impact. Among all cases, the stacking model with LR delivers the best performance, achieving an accuracy of 97.63%, precision of 97.68%, recall of 97.63%, and an F-measure of 97.63%, showcasing its exceptional reliability and balanced effectiveness. All models were evaluated using 10-fold cross-validation.