Lung cancer is the most significant cancer that heavily contributes to cancer-related mortality rate, due to its violent nature and late diagnosis at advanced stages. Early identification of lung cancer is essential for improving the survival rate. Various imaging modalities, including X-rays and computed tomography (CT) scans, are employed to diagnose lung cancer. Computer-aided diagnosis (CAD) models are necessary for minimizing the burden upon radiologists and enhancing detection efficiency. Currently, computer vision (CV) and deep learning (DL) models are employed to detect and classify the lung cancer in a precise manner. In this background, the current study presents a cat swarm optimization-based computer-aided diagnosis model for lung cancer classification (CSO-CADLCC) model. The proposed CHO-CADLCC technique initially pre-process the data using the Gabor filtering-based noise removal technique. Furthermore, feature extraction of the pre-processed images is performed with the help of NASNetLarge model. This model is followed by the CSO algorithm with weighted extreme learning machine (WELM) model, which is exploited for lung nodule classification. Finally, the CSO algorithm is utilized for optimal parameter tuning of the WELM model, resulting in an improved classification performance. The experimental validation of the proposed CSO-CADLCC technique was conducted against a benchmark dataset, and the results were assessed under several aspects. The experimental outcomes established the promising performance of the CSO-CADLCC approach over recent approaches under different measures.

This paper proposes a hybrid deep learning model integrating DenseNet201 and InceptionV3 to address the challenges in achieving accurate and reliable cervical cancer classification. Current models often exhibit limitations in balancing precision and recall, which are critical for dependable clinical applications. The hybrid model leverages DenseNet201’s efficient feature reuse and InceptionV3’s capacity for handling multi-scale and hierarchical features through fine-tuning and feature fusion techniques. The methodology involves rigorous data preprocessing, including normalization, augmentation, and dataset splitting, to ensure robust training and validation. Feature extraction and dimensionality optimization are employed to identify the most critical and discriminative features for classification. The experimental setup utilizes Python, TensorFlow, and Keras within a GPU-enabled environment to handle computational demands effectively. Comprehensive evaluation metrics, including accuracy, precision, recall, and F1-score, indicate that the proposed model achieves an accuracy of 96.54%, 95.91% Presicion, 96.44% Recall and 96.17% F1 Score surpassing state-of-the-art models such as ResNet-50, DenseNet-201, InceptionV3, and Xception. Visualization tools, including high-resolution confusion matrices and ROC curves, further demonstrate the hybrid model’s capability to differentiate between cervical cancer cell classes accurately. Comparative analyses validate the model’s superior performance and its potential as a dependable tool for clinical implementation. This study presents a robust and efficient classification system that addresses the limitations of existing models. Future research will focus on further improving the system’s performance and investigating its applicability to other medical imaging tasks. The proposed model is expected to contribute significantly to early and accurate cervical cancer diagnosis, enhancing patient outcomes and supporting healthcare professionals in clinical decision-making.

Introduction: Brain tumors are among the most aggressive forms of cancer, requiring precise diagnosis and treatment planning to improve patient outcomes. This study aims to develop an efficient deep learning-based framework for the classification of brain tumors using MRI data. Methods: The methodology employs Convolutional Neural Networks (CNNs) to accurately classify tumors into four categories: normal, glioma, pituitary, and meningioma. Key preprocessing techniques, including noise reduction,resizing, and data augmentation, were applied to enhance the robustness of the model. Advanced architectures such as DenseNet50, VGG19, and other transfer learning models, along with CNN variants, were trained and evaluated for their performance. Explainable AI (XAI) techniques, including Grad-CAM, LIME, and feature map visualizations, played a crucial role in providing better visualizations of the model’s decision-making process and identifying areas of improvement during model training and to establish a better model. Results: The best-performing model, a 4-conv-1-dense-1-dropout CNN, achieved a classification accuracy of 95.86%, outperforming deeper architectures and transfer learning approaches. The findings underscore the potential of deep learning models for reliable and efficient brain tumor classification. This work concludes with recommendations for real-time deployment in clinical settings and explores future integration with Large Language Models (LLMs) to generate detailed diagnostic reports.

The absence of reliable early treatment serves as one of the main causes of cervical cancer. Hence, it is crucial to detect cervical cancer early. The biggest challenge in diagnosing cervical cancer early is that it is asymptomatic until it develops into invasive carcinoma. In medical applications, the use of machine learning and deep learning is successful as a classifier in the preliminary identification of cancerous cells in the cervical region. In our study, we present a CNN-based method for the classification of cervical cancer cells. We present a method for accurately classifying Pap smear images into abnormal or healthy cells by extracting essential information using a variety of deep-learning approaches. Experiments are performed using the SIPaKMeD and Herlev datasets. Several pre-trained convolutional neural network (CNN) models are used via transfer learning methods, hence predicting and evaluating the accurate classifier with the best optimal solution. Classification of cervical cell clusters in whole slide images (WSI) has usually comprised two stages: segmentation to extract individual cell patches, and subsequently single-cell categorization. As a result, segmentation accuracy determines the classification pipeline’s performance. We propose a direct classification of WSI cervical cell groups without segmentation and demonstrate that segmentation is not strictly necessary for good classification results. Our solution outperformed prior methods and benchmarks, with an accuracy of 96.74% for WSI patches and 97.55% for full-cell images for the SIPaKMeD dataset, and an accuracy of 90.42% for the Herlev dataset. The results show that the suggested approach may accurately distinguish cervical cancerous and non-cancerous cells.

Artificial Intelligence has changed our outlook towards the whole world, and it is regularly used to better understand all the data and information that surround us in our everyday lives. One such application of Artificial Intelligence in a real-world scenario is the extraction of data from various images and interpreting them in different ways. This includes applications like object detection, image segmentation, image restoration, etc. While every technique has its own area of application, image segmentation has a variety of applications extending from the complex medical field to regular pattern identification. The aim of this paper is to research about several FCNN-based semantic segmentation techniques to develop a deep learning model that is able to segment tumours in brain MRI images to a high degree of precision and accuracy. The aim is to try several different architectures and experiments with several loss functions to improve the accuracy of our model and obtain the best model for our classification including newer loss functions like dice loss function, hierarchical dice loss function cross entropy, etc.