A Comprehensive Study on Deep Learning Models for the Detection of Ovarian Cancer and Glomerular Kidney Disease using Histopathological Images

Archives of Computational Methods in Engineering, Год журнала: 2024, Номер unknown

Опубликована: Июнь 1, 2024

Язык: Английский

---

Multiple perception contrastive learning for automated ovarian tumor classification in CT images

Abdominal Radiology, Год журнала: 2025, Номер unknown

Опубликована: Март 12, 2025

Язык: Английский

---

Artificial Intelligence for Ovarian Cancer Detection with Medical Images: A Review of the Last Decade (2013–2023)

Archives of Computational Methods in Engineering, Год журнала: 2025, Номер unknown

Опубликована: Март 17, 2025

Язык: Английский

---

Deformable Dual Graph Aggregation Transformer Convolutional Networks with Spider Wasp Optimizer for Ovarian Tumor Classification Using Magnetic Resonance Imaging

Deleted Journal, Год журнала: 2025, Номер unknown

Опубликована: Апрель 4, 2025

Язык: Английский

---

X-ray image classification with dual-model information fusion and improved PSO algorithm

Signal Image and Video Processing, Год журнала: 2025, Номер 19(7)

Опубликована: Май 9, 2025

Язык: Английский

---

Deep fine-KNN classification of ovarian cancer subtypes using efficientNet-B0 extracted features: a comprehensive analysis

Journal of Cancer Research and Clinical Oncology, Год журнала: 2024, Номер 150(7)

Опубликована: Июль 25, 2024

This study presents a robust approach for the classification of ovarian cancer subtypes through integration deep learning and k-nearest neighbor (KNN) methods. The proposed model leverages powerful feature extraction capabilities EfficientNet-B0, utilizing its features subsequent fine-grained using fine-KNN approach. UBC-OCEAN dataset, encompassing histopathological images five distinct subtypes, namely, high-grade serous carcinoma (HGSC), clear-cell (CC), endometrioid (EC), low-grade (LGSC), mucinous (MC), served as foundation our investigation. With dataset comprising 725 images, divided into 80% training 20% testing, exhibits exceptional performance. Both validation testing phases achieved 100% accuracy, underscoring efficacy methodology. In addition, area under curve (AUC), key metric evaluating model's discriminative ability, demonstrated high performance across various with AUC values 0.94, 0.78, 0.69, 0.92, 0.94 MC. Furthermore, positive likelihood ratios (LR

Язык: Английский

---

Accident Detection using Images and Videos with CNN, LSTM, and Interpreting the Results using LIME & GradCAM

2022 IEEE 7th International conference for Convergence in Technology (I2CT), Год журнала: 2024, Номер unknown, С. 1 - 8

Опубликована: Апрель 5, 2024

Accident detection is a crucial issue in modern society, as accidents can have devastating consequences for those involved. One way to improve accident through the use of images and videos, which provide valuable information about scene an accident. In this paper, we explore convolutional neural networks (CNNs), recurrent (RNNs), explainable AI concepts detecting using videos. Each frame video processed deep learning convolution network model that has been trained identify frames or non-accident. We also incorporate RNNs into our model, they are ideal processing sequential data such frames. addition CNNs RNNs, model. This allows us explanations model's predictions, help build trust system its overall effectiveness. Our initial results show approach effective at With impressive accuracy, categorize movie Additionally, clear predictions.

Язык: Английский

---

Neural Networks and Emotions: A Deep Learning Perspective

2022 IEEE 7th International conference for Convergence in Technology (I2CT), Год журнала: 2024, Номер unknown, С. 1 - 7

Опубликована: Апрель 5, 2024

Emotion recognition represents a critical facet of human-centric artificial intelligence systems. This paper delves into the forefront emotion detection by leveraging cutting-edge deep learning models across three distinct modalities: textual, visual, and auditory. Our text-based model harnesses potency Bidirectional Long Short-Term Memory (BiLSTM) networks, adept at capturing intricate semantic relationships contextual nuances within textual data. Simultaneously, we employ Convolutional Neural Networks (CNNs) in domain image-based detection, effectively extracting discriminative spatial features to discern emotional states visual content. For speech-based recognition, (LSTM) capitalizing on their ability capture temporal dependencies acoustic signals. These modalities converge offer comprehensive insights multimodal where fusion auditory information enhances classification accuracy. research not only underscores importance analysis but also holds great potential for applications human-computer interaction, sentiment analysis, mental health diagnostics, multimedia content understanding. By elucidating strengths synergies these modalities, this contributes significantly burgeoning field emotion-aware AI systems, promising more nuanced understanding human emotions an increasingly digital world.

Язык: Английский

---

Intelligent system based on multiple networks for accurate ovarian tumor semantic segmentation

Heliyon, Год журнала: 2024, Номер 10(17), С. e37386 - e37386

Опубликована: Сен. 1, 2024

Ovarian tumors, especially malignant ones, represent a global concern, with increased prevalence in recent years. More accurate medical support systems are urgently needed to staff obtaining an efficient ovarian tumors diagnosis since detection early stages could lead immediately applying appropriate treatment, and implicitly improving the survival rate. The current paper aims demonstrate that more be designed by combining different convolutional neural networks using custom combination approaches selecting involved ensemble model achieve best performance metrics. It is essential understand if all experimented or only best-performing ones always most effective results not. structured three main phases. first step propose individual experiments. Five DeepLab-V3+ encoders (ResNet-18, ResNet-50, MobileNet-V2, InceptionResNet-V2, Xception) were used. In second step, proposes algorithm combine multiple semantic segmentation networks, while last describes iterative selection approach for combined so obtained. system performing types of covering both benign achieved 91.18 % Intersection over union (IoU), thus overperforming networks. proposed method extended powerful deep learning models

Язык: Английский

---

Emotion Detection from Textual Data Using Supervised Machine Learning Models

Опубликована: Май 26, 2023

Emotion detection and recognition from text is a recent field of research that closely related to Sentiment analysis. Many people express themselves using text, photographs, music, video. Text communication web-based networking platforms, however, could be little overwhelming. Every second, substantial amount unstructured data produced on the Internet as result social media sites. This where sentiment analysis, which recognises polarity in texts, can useful. It assesses author's attitude towards specific object, administration, person, or location concludes if it positive, negative, neutral. In some cases, analysis inadequate, necessitating emotion detection, precisely ascertains person's mental/emotional state. The development text-based prediction model primary goal this work. confronting several market hurdles, with accuracy being key one. As result, Decision Trees, Naive Bayes, Support Vector Machine, Logistic Regression, k-Nearest Neighbors Random Forest, supervised machine learning classification algorithms were examined. six main emotions recognized by Ekman are joy, fear, anger, love, surprise, sadness, these foundation through was constructed. strategies for preprocessing containing stemming, stop-words, numerals, punctuation marks removal, tokenization, spelling correction implemented. review paper delves into degrees models well technique text.

Язык: Английский

---