Optimal deep learning based fusion model for biomedical image classification

Expert Systems, Journal Year: 2021, Volume and Issue: 39(3)

Published: July 11, 2021

Abstract Automated examination of biomedical signals plays a vital role to diagnose diseases and offers useful data several applications in the areas physiology, sports medicine, human–computer interface. The latest advancements Artificial Intelligence (AI) have ability manage analyse enormous datasets resulting clinical decision making real time applications. At same time, Colorectal cancer (CRC) is third most deadly disease affecting people over globe. utilization AI techniques for earlier identification CRC has gained significant interest among research communities. Therefore, this paper presents novel based fusion model diagnosis classification, named AIFM‐CRC. presented AIFM‐CRC primarily undergoes Gaussian filtering noise removal contrast enhancement as preprocessing stage. In addition, feature extraction process takes place where SIFT handcrafted features Inception v4 deep are fused together. Besides, whale optimization algorithm tuned support vector machine employed classification technique determine existence CRC. order highlight proficient results analysis model, comprehensive simulation place. resultant experimental values pointed out betterment by accomplishing maximum accuracy 96.18%.

Language: Английский

---

Statistical normalization methods in microbiome data with application to microbiome cancer research

Gut Microbes, Journal Year: 2023, Volume and Issue: 15(2)

Published: Aug. 25, 2023

Mounting evidence has shown that gut microbiome is associated with various cancers, including gastrointestinal (GI) tract and non-GI cancers. But data have unique characteristics pose major challenges when using standard statistical methods causing results to be invalid or misleading. Thus, analyze data, it not only needs appropriate methods, but also requires normalized prior analysis. Here, we first describe the of in analyzing them (Section 2). Then, provide an overall review on available normalization 16S rRNA shotgun metagenomic along examples their applications cancer research 3). In Section 4, comprehensively investigate how are evaluated. Finally, summarize conclude remarks 5). Altogether, this aims a broad comprehensive view promises examples.

Language: Английский

---

A review of machine learning methods for cancer characterization from microbiome data

npj Precision Oncology, Journal Year: 2024, Volume and Issue: 8(1)

Published: May 30, 2024

Abstract Recent studies have shown that the microbiome can impact cancer development, progression, and response to therapies suggesting microbiome-based approaches for characterization. As cancer-related signatures are complex implicate many taxa, their discovery often requires Machine Learning approaches. This review discusses methods characterization from data. It focuses on implications of choices undertaken during sample collection, feature selection pre-processing. also ML model selection, guiding how choose an model, validation. Finally, it enumerates current limitations these may be surpassed. Proposed methods, based Random Forests, show promising results, however insufficient widespread clinical usage. Studies report conflicting results mainly due models with poor generalizability. We expect evaluating expanded, hold-out datasets, removing technical artifacts, exploring representations other than taxonomical profiles, leveraging advances in deep learning, developing better adapted characteristics data will improve performance generalizability enable usage clinic.

Language: Английский

---

Revolutionizing colorectal cancer detection: A breakthrough in microbiome data analysis

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(1), P. e0316493 - e0316493

Published: Jan. 29, 2025

The emergence of Next Generation Sequencing (NGS) technology has catalyzed a paradigm shift in clinical diagnostics and personalized medicine, enabling unprecedented access to high-throughput microbiome data. However, the inherent high dimensionality, noise, variability data present substantial obstacles conventional statistical methods machine learning techniques. Even promising deep (DL) are not immune these challenges. This paper introduces novel feature engineering method that circumvents limitations by amalgamating two sets derived from input generate new dataset, which is then subjected selection. innovative approach markedly enhances Area Under Curve (AUC) performance Deep Neural Network (DNN) algorithm colorectal cancer (CRC) detection using gut data, elevating it 0.800 0.923. proposed constitutes significant advancement field, providing robust solution intricacies analysis amplifying potential DL disease detection.

Language: Английский

---

DeepFeature: feature selection in nonimage data using convolutional neural network

Briefings in Bioinformatics, Journal Year: 2021, Volume and Issue: 22(6)

Published: July 15, 2021

Artificial intelligence methods offer exciting new capabilities for the discovery of biological mechanisms from raw data because they are able to detect vastly more complex patterns association that cannot be captured by classical statistical tests. Among these methods, deep neural networks currently among most advanced approaches and, in particular, convolutional (CNNs) have been shown perform excellently a variety difficult tasks. Despite applications this type high-dimensional omics importantly, meaningful interpretation results returned such models biomedical context remains an open problem. Here we present, approach applying CNN nonimage feature selection. Our pipeline, DeepFeature, can both successfully transform into form is optimal fitting model and also return sets important genes used internally computing predictions. Within framework, Snowfall compression algorithm introduced enable elements fixed pixel region accumulation element decoder developed find or class activation maps. In comparative tests cancer prediction task, DeepFeature simultaneously achieved superior predictive performance better ability discover key pathways processes context. Capabilities offered proposed framework effective use powerful learning facilitate causal data.

Language: Английский

---

Deep learning methods in metagenomics: a review

Microbial Genomics, Journal Year: 2024, Volume and Issue: 10(4)

Published: April 17, 2024

The ever-decreasing cost of sequencing and the growing potential applications metagenomics have led to an unprecedented surge in data generation. One most prevalent is study microbial environments, such as human gut. gut microbiome plays a crucial role health, providing vital information for patient diagnosis prognosis. However, analysing metagenomic remains challenging due several factors, including reference catalogues, sparsity compositionality. Deep learning (DL) enables novel promising approaches that complement state-of-the-art pipelines. DL-based methods can address almost all aspects analysis, pathogen detection, sequence classification, stratification disease prediction. Beyond generating predictive models, key aspect these also their interpretability. This article reviews DL metagenomics, convolutional networks, autoencoders attention-based models. These aggregate contextualized pave way improved care better understanding microbiome's our health.

Language: Английский

---

Tunicate swarm algorithm with deep convolutional neural network-driven colorectal cancer classification from histopathological imaging data

Electronic Research Archive, Journal Year: 2023, Volume and Issue: 31(5), P. 2793 - 2812

Published: Jan. 1, 2023

<abstract> <p>Colorectal cancer (CRC) is one of the most popular cancers among both men and women, with increasing incidence. The enhanced analytical load data from pathology laboratory, integrated described intra- inter-variabilities through calculation biomarkers, has prompted quest for robust machine-based approaches in combination routine practice. In histopathology, deep learning (DL) techniques have been applied at large due to their potential supporting analysis forecasting medically appropriate molecular phenotypes microsatellite instability. Considering this background, current research work presents a metaheuristics technique convolutional neural network-based colorectal classification based on histopathological imaging (MDCNN-C3HI). presented MDCNN-C3HI majorly examines images (CRC). At initial stage, applies bilateral filtering approach get rid noise. Then, proposed uses an capsule network Adam optimizer extraction feature vectors. For CRC classification, DL modified classifier, whereas tunicate swarm algorithm used fine-tune its hyperparameters. To demonstrate performance wide range experiments was conducted. outcomes extensive experimentation procedure confirmed superior over other existing techniques, achieving maximum accuracy 99.45%, sensitivity 99.45% specificity 99.45%.</p> </abstract>

Language: Английский

---

Overview of data preprocessing for machine learning applications in human microbiome research

Frontiers in Microbiology, Journal Year: 2023, Volume and Issue: 14

Published: Oct. 5, 2023

Although metagenomic sequencing is now the preferred technique to study microbiome-host interactions, analyzing and interpreting microbiome data presents challenges primarily attributed statistical specificities of (e.g., sparse, over-dispersed, compositional, inter-variable dependency). This mini review explores preprocessing transformation methods applied in recent human studies address analysis challenges. Our results indicate a limited adoption targeting characteristics data. Instead, there prevalent usage relative normalization-based transformations that do not specifically account for specific attributes The information on before was incomplete or missing many publications, leading reproducibility concerns, comparability issues, questionable results. We hope this will provide researchers newcomers field research with an up-to-date point reference various tools assist them choosing most suitable method based their questions, objectives, characteristics.

Language: Английский

---

Novel aeroengine fault diagnosis method based on feature amplification

Engineering Applications of Artificial Intelligence, Journal Year: 2023, Volume and Issue: 122, P. 106093 - 106093

Published: March 23, 2023

Language: Английский

---

Deep learning in microbiome analysis: a comprehensive review of neural network models

Frontiers in Microbiology, Journal Year: 2025, Volume and Issue: 15

Published: Jan. 22, 2025

Microbiome research, the study of microbial communities in diverse environments, has seen significant advances due to integration deep learning (DL) methods. These computational techniques have become essential for addressing inherent complexity and high-dimensionality microbiome data, which consist different types omics datasets. Deep algorithms shown remarkable capabilities pattern recognition, feature extraction, predictive modeling, enabling researchers uncover hidden relationships within ecosystems. By automating detection functional genes, interactions, host-microbiome dynamics, DL methods offer unprecedented precision understanding composition its impact on health, disease, environment. However, despite their potential, approaches face challenges research. Additionally, biological variability datasets requires tailored ensure robust generalizable outcomes. As research continues generate vast complex datasets, these will be crucial advancing microbiological insights translating them into practical applications with DL. This review provides an overview models discussing strengths, uses, implications future studies. We examine how are being applied solve key problems highlight potential pathways overcome current limitations, emphasizing transformative could field moving forward.

Language: Английский

---