Classification performance assessment for imbalanced multiclass data

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: May 10, 2024

Abstract The evaluation of diagnostic systems is pivotal for ensuring the deployment high-quality solutions, especially given pronounced context-sensitivity certain systems, particularly in fields such as biomedicine. Of notable importance are predictive models where target variable can encompass multiple values (multiclass), when these classes exhibit substantial frequency disparities (imbalance). In this study, we introduce Imbalanced Multiclass Classification Performance (IMCP) curve, specifically designed multiclass datasets (unlike ROC curve), and characterized by its resilience to class distribution variations (in contrast accuracy or F $$_\beta$$ β -score). Moreover, IMCP curve facilitates individual performance assessment each within system, shedding light on confidence associated with prediction—an aspect particular significance medical diagnosis. Empirical experiments conducted real-world data a context (involving 35 types tumors) featuring high level imbalance demonstrate that both area under serve excellent indicators classification quality.

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

---

Discovery of robust and highly specific microbiome signatures of non-alcoholic fatty liver disease

Microbiome, Journal Year: 2025, Volume and Issue: 13(1)

Published: Jan. 14, 2025

The pathogenesis of non-alcoholic fatty liver disease (NAFLD) with a global prevalence 30% is multifactorial and the involvement gut bacteria has been recently proposed. However, finding robust bacterial signatures NAFLD great challenge, mainly due to its co-occurrence other metabolic diseases. Here, we collected public metagenomic data integrated taxonomy profiles in silico generated community outputs, detailed clinical data, 1206 Chinese subjects w/wo diseases, including (obese lean), obesity, T2D, hypertension, atherosclerosis. We identified highly specific microbiome through building accurate machine learning models (accuracy = 0.845–0.917) for high portability (generalizable) low prediction rate (specific) when applied as well approach involving differential co-abundance ecological networks. Moreover, using these coupled further mediation analysis dependency modeling, propose synergistic defined microbial consortia associated phenotype overweight lean individuals, respectively. Our study reveals offers more realistic microbiome-therapeutics over individual species this complex disease.

Language: Английский

---

Synergizing Artificial Intelligence and Probiotics: A Comprehensive Review of Emerging Applications in Health Promotion and Industrial Innovation

Trends in Food Science & Technology, Journal Year: 2025, Volume and Issue: unknown, P. 104938 - 104938

Published: Feb. 1, 2025

Language: Английский

---

Performance of Gut Microbiome as an Independent Diagnostic Tool for 20 Diseases: Cross-Cohort Validation of Machine-Learning Classifiers

Gut Microbes, Journal Year: 2023, Volume and Issue: 15(1)

Published: May 4, 2023

Cross-cohort validation is essential for gut-microbiome-based disease stratification but was only performed limited diseases. Here, we systematically evaluated the cross-cohort performance of gut microbiome-based machine-learning classifiers 20 Using single-cohort classifiers, obtained high predictive accuracies in intra-cohort (~0.77 AUC), low validation, except intestinal diseases (~0.73 AUC). We then built combined-cohort trained on samples combined from multiple cohorts to improve non-intestinal diseases, and estimated required sample size achieve >0.7. In addition, observed higher using metagenomic data than 16S amplicon further quantified marker consistency a Marker Similarity Index similar trends. Together, our results supported microbiome as an independent diagnostic tool revealed strategies based identified determinants consistent alterations.

Language: Английский

---

Emerging roles of the gut microbiota in cancer immunotherapy

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

Published: Feb. 22, 2023

Gut microbiota represents a hidden treasure vault encompassing trillions of microorganisms that inhabit the intestinal epithelial barrier host. In past decade, numerous in-vitro , animal and clinical studies have revealed profound roles gut in maintaining homeostasis various physiological functions, especially immune modulation, remarkable differences configuration microbial communities between cancers healthy individuals. addition, although considerable efforts been devoted to cancer treatments, there remain many patients succumb their disease with incremental burden worldwide. Nevertheless, compared stability human genome, plasticity renders it promising opportunity for individualized treatment. Meanwhile, burgeoning findings indicate is involved close interactions outcomes diverse immunotherapy protocols, including checkpoint blockade therapy, allogeneic hematopoietic stem cell transplantation, chimeric antigen receptor T therapy. Here, we reviewed evidence capacity microflora modulate immunotherapies, highlighted opportunities microbiota-based prognostic prediction, as well microbiotherapy by targeting potentiate anticancer efficacy while attenuating toxicity, which will be pivotal development personalized treatment strategies.

Language: Английский

---

Advances in AI‐based cancer cytopathology

Deleted Journal, Journal Year: 2023, Volume and Issue: 1(3)

Published: July 1, 2023

Abstract Cytopathological examination plays a crucial role in cancer diagnosis as it reflects the cellular pathology of cancer. However, this process traditionally relies on visual by cytopathologists. Recent advancements computer and digital imaging technologies have enabled application artificial intelligence (AI)‐based models to identify tumor cells images, thereby assisting cytopathologists achieving enhanced performance. AI‐based can improve accuracy reproducibility image evaluation streamline clinical workflows. Moreover, analyze diverse range sample types, including peripheral blood, urine, ascites, bone marrow. cytopathological recognition help clinicians screen diagnose cancer, predict prognosis recurrence cancers, such leukemia, cervical urothelial carcinoma, gastric Additionally, types mutations leukemia. A growing number studies emphasize potential computational analysis deep learning‐based AI build novel diagnostic tools that are conducive biomedical field. This review describes recent developments offers perspective how cytopathology prediction. Future model applications further contribute improvement human health.

Language: Английский

---

Artificial intelligence-driven rechargeable batteries in multiple fields of development and application towards energy storage

Journal of Energy Storage, Journal Year: 2023, Volume and Issue: 73, P. 108926 - 108926

Published: Sept. 18, 2023

Language: Английский

---

The colon targeting efficacies of mesalazine medications and their impacts on the gut microbiome

Journal of Controlled Release, Journal Year: 2024, Volume and Issue: 369, P. 630 - 641

Published: April 11, 2024

Successful treatment of ulcerative colitis (UC) is highly dependent on several parameters, including dosing regimen and the ability to deliver drugs disease site. In this study two strategies for delivering mesalazine (5-aminosalicylic acid, 5-ASA) colon were compared in an advanced vitro model human gastrointestinal (GI) tract, SHIME® system. Herein, a prodrug strategy employing bacteria-mediated drug release (sulfasalazine, Azulfidine®) was evaluated alongside formulation that utilised pH (5-ASA, Octasa® 1600 mg). experiments performed simulating both GI physiology colonic microbiota under healthy inflammatory bowel (IBD) conditions, impact state ileal variability 5-ASA delivery. addition, effects products microbiome investigated by monitoring bacterial growth metabolites. Results demonstrated approaches resulted similar percentage recovery conditions. On contrary, during IBD patients (the target population) higher proportion delivery region as approach (P < 0.0001). Interestingly, had distinct synthesis key metabolites, such lactate short chain fatty acids, which varied according variability. Further, sulfasalazine significantly reduced faecal sourced from six humans. The findings support selected could influence effectiveness UC treatment, highlight licensed may differentially functioning microbiota.

Language: Английский

---

Past, present, and future of microbiome-based therapies

Microbiome Research Reports, Journal Year: 2024, Volume and Issue: 3(2)

Published: March 18, 2024

Technological advances in studying the human microbiome depth have enabled identification of microbial signatures associated with health and disease. This confirms crucial role microbiota maintaining homeostasis host status. Nowadays, there are several ways to modulate composition effectively improve health; therefore, development therapeutic treatments based on gut is experiencing rapid growth. In this review, we summarize influence infectious disease cancer, which two main targets microbiome-based therapies currently being developed. We analyze two-way interaction between traditional drugs order emphasize drug effectivity treatment response. explore different strategies available for modulating ecosystem our benefit, ranging from 1st generation intervention more complex 2nd their regulatory framework. Lastly, finish a quick overview what believe future these strategies, that 3rd developed use artificial intelligence (AI) algorithms.

Language: Английский

---