Efficient Gastrointestinal Disease Classification Using Pretrained Deep Convolutional Neural Network

Electronics, Год журнала: 2023, Номер 12(7), С. 1557 - 1557

Опубликована: Март 26, 2023

Gastrointestinal (GI) tract diseases are on the rise in world. These can have fatal consequences if not diagnosed initial stages. WCE (wireless capsule endoscopy) is advanced technology used to inspect gastrointestinal such as ulcerative-colitis, polyps, esophagitis, and ulcers. produces thousands of frames for a single patient’s procedure which manual examination tiresome, time-consuming, prone error; therefore, an automated needed. images suffer from low contrast increases inter-class intra-class similarity reduces anticipated performance. In this paper, efficient GI disease classification technique proposed utilizes optimized brightness-controlled contrast-enhancement method improve images. The applies genetic algorithm (GA) adjusting values brightness within image by modifying fitness function, improves overall quality This improvement reported using qualitative measures, peak signal noise ratio (PSNR), mean square error (MSE), visual information fidelity (VIF), index (SI), (IQI). As second step, data augmentation performed applying multiple transformations, then, transfer learning fine-tune modified pre-trained model Finally, disease, extracted features passed through machine-learning classifiers. To show efficacy performance, results original dataset well contrast-enhanced dataset. 15.26% accuracy, 13.3% precision, 16.77% recall rate, 15.18% F-measure. comparison with existing techniques shows that framework outperforms state-of-the-art techniques.

Язык: Английский

---

A New Approach for Gastrointestinal Tract Findings Detection and Classification: Deep Learning-Based Hybrid Stacking Ensemble Models

Diagnostics, Год журнала: 2023, Номер 13(4), С. 720 - 720

Опубликована: Фев. 14, 2023

Endoscopic procedures for diagnosing gastrointestinal tract findings depend on specialist experience and inter-observer variability. This variability can cause minor lesions to be missed prevent early diagnosis. In this study, deep learning-based hybrid stacking ensemble modeling has been proposed detecting classifying system findings, aiming at diagnosis with high accuracy sensitive measurements saving workload help the objectivity in endoscopic first level of bi-level approach, predictions are obtained by applying 5-fold cross-validation three new CNN models. A machine learning classifier selected second is trained according predictions, final classification result reached. The performances models were compared models, McNemar’s statistical test was applied support results. According experimental results, performed a significant difference 98.42% ACC 98.19% MCC KvasirV2 dataset 98.53% 98.39% HyperKvasir dataset. study offer learning-oriented approach that efficiently evaluates features provides objective reliable results testing state-of-the-art studies subject. improves performance outperforms literature.

Язык: Английский

---

GastroNet: A robust attention‐based deep learning and cosine similarity feature selection framework for gastrointestinal disease classification from endoscopic images

CAAI Transactions on Intelligence Technology, Год журнала: 2023, Номер unknown

Опубликована: Июнь 11, 2023

Abstract Diseases of the Gastrointestinal (GI) tract significantly affect quality human life and have a high fatality rate. Accurate diagnosis GI diseases plays pivotal role in healthcare systems. However, processing large amounts medical image data can be challenging for radiologists other professionals, increasing risk inaccurate assessments. Computer‐aided Diagnosis systems provide help to doctors rapid accurate diagnosis, thus resulting saving lives. Recently, many techniques are found literature that uses deep Convolutional Neural Network (CNN) models disease classification. they limitations their ability detect deformation‐invariant features lack robustness. The diseased region is highlighted, using attention‐based generation superimposition with original images. A lightweight CNN model employed get significant features. These further reduced Cosine similarity‐based technique. proposed framework assessed Kvasir dataset. To verify effectiveness framework, vast experiments conducted. overall accuracy 97.68%, 99.02% precision, 96.37% recall, an F‐measure 97.68% achieved 810 This reduction resulted classification time. robustness observed not only terms considerable improvement accuracy, but also precision as well F‐measure.

Язык: Английский

---

GIT-Net: An Ensemble Deep Learning-Based GI Tract Classification of Endoscopic Images

Bioengineering, Год журнала: 2023, Номер 10(7), С. 809 - 809

Опубликована: Июль 5, 2023

This paper presents an ensemble of pre-trained models for the accurate classification endoscopic images associated with Gastrointestinal (GI) diseases and illnesses. In this paper, we propose a weighted average model called GIT-NET to classify GI-tract diseases. We evaluated on KVASIR v2 dataset eight classes. When individual are used classification, they often prone misclassification since may not be able learn characteristics all classes adequately. is due fact that each specific more efficiently than other leverages predictions three models, DenseNet201, InceptionV3, ResNet50 accuracies 94.54%, 88.38%, 90.58%, respectively. The base learners combined using two methods: averaging averaging. performances evaluated, has accuracy 92.96% whereas 95.00%. outperforms models. results from evaluation demonstrate utilizing can successfully features were incorrectly learned by learners.

Язык: Английский

---

Detection of various gastrointestinal tract diseases through a deep learning method with ensemble ELM and explainable AI

Expert Systems with Applications, Год журнала: 2024, Номер 256, С. 124908 - 124908

Опубликована: Июль 30, 2024

The rising prevalence of gastrointestinal (GI) tract disorders worldwide highlights the urgent need for precise diagnosis, as these diseases greatly affect human life and contribute to high mortality rates. Fast identification, accurate classification, efficient treatment approaches are essential addressing this critical health issue. Common side effects include abdominal pain, bloating, discomfort, which can be chronic debilitating. Nausea vomiting also frequent, leading difficulties in maintaining adequate nutrition hydration. current study intends develop a deep learning (DL)-based approach that automatically classifies GI diseases. For first time, GastroVision dataset with 8000 images 27 different was utilized work design computer-aided diagnosis (CAD) system. This presents novel lightweight feature extractor compact size minimum number layers named Parallel Depthwise Separable Convolutional Neural Network (PD-CNN) Pearson Correlation Coefficient (PCC) selector. Furthermore, robust classifier Ensemble Extreme Learning Machine (EELM), combined pseudo inverse ELM (ELM) L1 Regularized (RELM), has been proposed identify more precisely. A hybrid preprocessing technique, including scaling, normalization, image enhancement techniques such erosion, CLAHE, sharpening, Gaussian filtering, employed enhance representation improve classification performance. consists twenty-four only 0.815 million parameters 9.79 MB model size. PD-CNN-PCC-EELM extracts features, reduces computational overhead, achieves excellent performance on multiclass images. achieved highest precision, recall, f1, accuracy, ROC-AUC, AUC-PR values 88.12 ± 0.332 %, 87.75 0.348 87.12 0.324 98.89 92 respectively, while testing time 0.000001 s. comparative utilizes 10-fold cross-validation, ablation various state-of-the-art (SOTA) transfer (TL) models extractors. Then, PCC EELM integrated TL generate predictions, notably terms real-time processing capability; significantly outperforms other models. Moreover, explainable AI (XAI) methods, SHAP (Shapley Additive Explanations), heatmap, guided Grad-Cam (Gradient-weighted Class Activation Mapping), Grad-CAM, Saliency mapping, have explore interpretability decision-making capability model. Therefore, provides practical intelligence increasing confidence diagnosing real-world scenarios.

Язык: Английский

---

An ensemble approach of deep CNN models with Beta normalization aggregation for gastrointestinal disease detection

Biomedical Signal Processing and Control, Год журнала: 2025, Номер 105, С. 107567 - 107567

Опубликована: Фев. 4, 2025

Язык: Английский

---

Interpretable deep learning architecture for gastrointestinal disease detection: A Tri-stage approach with PCA and XAI

Computers in Biology and Medicine, Год журнала: 2024, Номер 185, С. 109503 - 109503

Опубликована: Дек. 7, 2024

Язык: Английский

---

A Convolutional Neural Network with Meta-feature Learning for Wireless Capsule Endoscopy Image Classification

Journal of Medical and Biological Engineering, Год журнала: 2023, Номер 43(4), С. 475 - 494

Опубликована: Авг. 1, 2023

Язык: Английский

---

Optimized deep learning model for comprehensive medical image analysis across multiple modalities

Neurocomputing, Год журнала: 2024, Номер 619, С. 129182 - 129182

Опубликована: Дек. 12, 2024

Язык: Английский

---

Efficient-gastro: optimized EfficientNet model for the detection of gastrointestinal disorders using transfer learning and wireless capsule endoscopy images

PeerJ Computer Science, Год журнала: 2024, Номер 10, С. e1902 - e1902

Опубликована: Март 11, 2024

Gastrointestinal diseases cause around two million deaths globally. Wireless capsule endoscopy is a recent advancement in medical imaging, but manual diagnosis challenging due to the large number of images generated. This has led research into computer-assisted methodologies for diagnosing these images. Endoscopy produces thousands frames each patient, making examination difficult, laborious, and error-prone. An automated approach essential speed up process, reduce costs, potentially save lives. study proposes transfer learning-based efficient deep learning methods detecting gastrointestinal disorders from multiple modalities, aiming detect with superior accuracy efforts costs experts. The Kvasir eight-class dataset was used experiment, where endoscopic were preprocessed enriched augmentation techniques. EfficientNet model optimized via fine tuning, compared most widely pre-trained models. model’s efficacy tested on another independent prove its robustness reliability.

Язык: Английский

---