Advances in colorectal cancer diagnosis using optimal deep feature fusion approach on biomedical images

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Feb. 4, 2025

Colorectal cancer (CRC) is the second popular in females and third males, with an increased number of cases. Pathology diagnoses complemented predictive prognostic biomarker information first step for personalized treatment. Histopathological image (HI) analysis benchmark pathologists to rank colorectal various kinds. However, pathologists' are highly subjective susceptible inaccurate diagnoses. The improved diagnosis load pathology laboratory, incorporated reported intra- inter-variability assessment, has prompted quest consistent machine-based techniques be integrated into routine practice. In healthcare field, artificial intelligence (AI) achieved extraordinary achievements applications. Lately, computer-aided (CAD) based on HI progressed rapidly increase machine learning (ML) deep (DL) models. This study introduces a novel Cancer Diagnosis using Optimal Deep Feature Fusion Approach Biomedical Images (CCD-ODFFBI) method. primary objective CCD-ODFFBI technique examine biomedical images identify (CRC). technique, median filtering (MF) approach initially utilized noise elimination. utilizes fusion three DL models, MobileNet, SqueezeNet, SE-ResNet, feature extraction. Moreover, models' hyperparameter selection performed Osprey optimization algorithm (OOA). Finally, belief network (DBN) model employed classify CRC. A series simulations accomplished highlight significant results method under Warwick-QU dataset. comparison showed superior accuracy value 99.39% over existing techniques.

Language: Английский

---

Stacking and Chaining of Normalization Methods in Deep Learning-Based Classification of Colorectal Cancer Using Gut Microbiome Data

IEEE Access, Journal Year: 2021, Volume and Issue: 9, P. 97296 - 97319

Published: Jan. 1, 2021

Machine learning (ML)-based detection of diseases using sequence-based gut microbiome data has been great interest within the artificial intelligence in medicine (AIM) community. The approach offers a non-invasive alternative for colorectal cancer detection, which is based on stool samples. Considering limitations existing methods CRC medical research shown use high throughput to identify disease. Owing several conventional ML algorithms, deep (DL) are becoming more popular due their outstanding performance related fields. However, DL affected by such as dimensionality, sparsity, and feature dominance inherent data. This proposes stacking chaining normalization address limitations. While technique robust, easy use, interpretable augmenting other tabular data, an that dynamically adjusts underlying properties towards normal distribution. proposed techniques combined with rank transformation selection further improve model, area under curve (AUC) values between 0.857 0.987 publicly available datasets.

Language: Английский

---

Classification of Microbiome Data from Type 2 Diabetes Mellitus Individuals with Deep Learning Image Recognition

Big Data and Cognitive Computing, Journal Year: 2023, Volume and Issue: 7(1), P. 51 - 51

Published: March 17, 2023

Microbiomic analysis of human gut samples is a beneficial tool to examine the general well-being and various health conditions. The balance intestinal flora important prevent chronic infections adiposity, as well pathological alterations connected diseases. evaluation microbiome data based on next-generation sequencing (NGS) complex their interpretation often challenging can be ambiguous. Therefore, we developed an innovative approach for examination classification microbiomic into healthy diseased by visualizing radial heatmap in order apply deep learning (DL) image classification. differentiation between 674 272 type 2 diabetes mellitus (T2D) was chosen proof concept. residual network with 50 layers (ResNet-50) model trained optimized, providing discrimination 96% accuracy. Samples from persons were detected specificity 97% those T2D individuals sensitivity 92%. Image using DL NGS enables precise diabetic individuals. In future, this could enable different diseases imbalances causative genera.

Language: Английский

---

Current Trends and Challenges of Microbiome Research in Prostate Cancer

Current Oncology Reports, Journal Year: 2024, Volume and Issue: 26(5), P. 477 - 487

Published: April 4, 2024

Language: Английский

---

Multi-stage biomedical feature selection extraction algorithm for cancer detection

SN Applied Sciences, Journal Year: 2023, Volume and Issue: 5(5)

Published: April 7, 2023

Abstract Cancer is a significant cause of death worldwide. Early cancer detection greatly aided by machine learning and artificial intelligence (AI) to gene microarray data sets (microarray data). Despite this, there discrepancy between the number features in set samples. Because it crucial identify markers for array data. Existing feature selection algorithms, however, generally use long-standing, are limited single-condition rarely take extraction into account. This work proposes Multi-stage algorithm Biomedical Deep Feature Selection (MBDFS) address this issue. In first, three techniques combined thorough selection, subsets obtained; second, an unsupervised neural network used create best representation subset enhance final classification accuracy. Using variety metrics, including comparison results before after performance alternative methods, we evaluate MBDFS's efficacy. The experiments demonstrate that although MBDFS uses fewer features, accuracy either unchanged or enhanced.

Language: Английский

---

Supervised machine learning for microbiomics: bridging the gap between current and best practices

Machine Learning with Applications, Journal Year: 2024, Volume and Issue: 18, P. 100607 - 100607

Published: Nov. 14, 2024

Language: Английский

---

Investigation of trends in gut microbiome associated with colorectal cancer using machine learning

Frontiers in Oncology, Journal Year: 2023, Volume and Issue: 13

Published: March 1, 2023

The rapid growth of publications on the gut microbiome and colorectal cancer (CRC) makes it feasible for text mining bibliometric analysis.Publications were retrieved from Web Science. Bioinformatics analysis was performed, a machine learning-based Latent Dirichlet Allocation (LDA) model used to identify subfield research topics.A total 5,696 related CRC Science Core Collection 2000 2022. China USA most productive countries. top 25 references, institutions, authors with strongest citation bursts identified. Abstracts all extracted that identified 50 topics in this field increasing interest. colitis animal model, expression cytokines, sequencing 16s, composition dysbiosis, cell inhibition increasingly noticed during last two years. intensively investigated further categorized into four clusters, including "microbiome tumor," compositions, interactions, treatment," molecular features mechanisms," metabolism."This explores historical tendencies identifies specific developmental trajectory, along noticeable characterized by analysis, will contribute future direction its clinical translation.

Language: Английский

---

Creating Smart House via IoT and Intelligent Computation

Intelligent Automation & Soft Computing, Journal Year: 2022, Volume and Issue: 35(1), P. 415 - 430

Published: June 6, 2022

This study mainly uses the concept of Internet Things (IoT) to establish a smart house with an indoor, comfortable, environmental, and real-time monitoring system. In house, this investigation employed temperature- humidity-sensing module lightness monitor any condition for intelligent-living house. The data temperature, humidity, are transmitted wirelessly human-machine interface. correlation weight extension theory is used analyze ideal comfortable environment so that people in indoor can feel better thermal comfort lightness. study, improved particle swarm optimization (IPSO) employed—an effective evolutionary method search function extreme. It simple has fast convergence. convergence accuracy algorithm not high at beginning, it easily fall into local extreme points. effect inertia mix PSO becomes IPSO-Extension Neural Network (ENN), which was analyzed found reliable. Motivated by idea power function, new non-linear strategy decreasing (DIW) proposed based on existing linear DIW. Then, novel hierarchical multi-sensor fusion adopting presented, factor estimated. distinctive feature its capability fusing near-optimal manner when there no available information about reliability sources, degree redundancy/complementarities structure hierarchy. obtained from data, successfully removed noise disturbance, achieved favorable results.

Language: Английский

---

DeepGum: Deep feature transfer for gut microbiome analysis using bottleneck models

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 91, P. 105984 - 105984

Published: Jan. 31, 2024

Language: Английский

---

Deep Learning-Assisted Techniques for Detection and Prediction of Colorectal Cancer From Medical Images and Microbial Modality

Microorganisms for sustainability, Journal Year: 2024, Volume and Issue: unknown, P. 151 - 169

Published: Jan. 1, 2024

Language: Английский

---