ML-based binary classifiers
Dotaz
Zobrazit nápovědu
In this paper, we present the results of the research concerning extraction of informative gene expression profiles from high-dimensional array of gene expressions considering the state of patients' health using clustering method, ML-based binary classifiers and fuzzy inference system. Applying of the proposed stepwise procedure can allow us to extract the most informative genes taking into account both the subtypes of disease or state of the patient's health for further reconstruction of gene regulatory networks based on the allocated genes and following simulation of the reconstructed models. We used the publicly available gene expressions data as the experimental ones which were obtained using DNA microarray experiments and contained two types of patients' gene expression profiles-the patients with lung cancer tumor and healthy patients. The stepwise procedure of the data processing assumes the following steps-in the beginning, we reduce the number of genes by removing non-informative genes in terms of statistical criteria and Shannon entropy; then, we perform the stepwise hierarchical clustering of gene expression profiles at hierarchical levels from 1 to 10 using the SOTA (Self-Organizing Tree Algorithm) clustering algorithm with correlation distance metric. The quality of the obtained clustering was evaluated using the complex clustering quality criterion which is considered both the gene expression profiles distribution relative to center of the clusters where these gene expression profiles are allocated and the centers of the clusters distribution. The result of this stage execution was a selection of the optimal cluster at each of the hierarchical levels which corresponded to the minimum value of the quality criterion. At the next step, we have implemented a classification procedure of the examined objects using four well known binary classifiers-logistic regression, support-vector machine, decision trees and random forest classifier. The effectiveness of the appropriate technique was evaluated based on the use of ROC (Receiver Operating Characteristic) analysis using criteria, included as the components, the errors of both the first and the second kinds. The final decision concerning the extraction of the most informative subset of gene expression profiles was taken based on the use of the fuzzy inference system, the inputs of which are the results of the appropriate single classifiers operation and the output is the final solution concerning state of the patient's health. To our mind, the implementation of the proposed stepwise procedure of the informative gene expression profiles extraction create the conditions for the increasing effectiveness of the further procedure of gene regulatory networks reconstruction and the following simulation of the reconstructed models considering the subtypes of the disease and/or state of the patient's health.
- Publikační typ
- časopisecké články MeSH
BACKGROUND AND OBJECTIVES: Frontotemporal lobar degeneration (FTLD) as the second most common dementia encompasses a range of syndromes and often shows overlapping symptoms with other subtypes or neurodegenerative diseases, which poses a significant clinical diagnostic challenge. Recent advancements in artificial intelligence (AI), specifically the application of machine learning (ML) algorithms to neuroimaging, have significantly progressed in addressing this challenge. This study aims to assess the diagnostic and predictive efficacy of neuroimaging feature-based AI algorithms for FTLD. METHODS: We conducted a systematic review and meta-analysis following PRISMA guidelines. We searched Pubmed, Scopus, and Web of Science for English-language, peer-reviewed studies using the following three umbrella terms: artificial intelligence, frontotemporal lobar degeneration, and neuroimaging modality. Our survey focused on computer-aided diagnosis for FTLD, employing machine/deep learning with neuroimaging radiomic features. RESULTS: The meta-analysis includes 75 articles with 20,601 subjects, including 8,051 FTLD patients. The results reveal that FTLD can be automatically classified against healthy controls (HC) with pooled sensitivity and specificity of 86% and 89%, respectively. Likewise, FTLD versus Alzheimer's disease (AD) classification exhibits pooled sensitivity and specificity of 84% and 81%, while FTLD versus Parkinson's disease (PD) demonstrates pooled sensitivity and specificity of 84% and 75%, respectively. Classification performance distinguishing FTLD from atypical Parkinsonian syndromes (APS) showed pooled sensitivity and specificity of 84% and 79%, respectively. Multiclass classification sensitivity ranges from 42% to 100%, with lower sensitivity occurring in higher class distinctions (e.g., 5-class and 11-class). DISCUSSION: Our study demonstrates the effectiveness of utilizing neuroimaging features to distinguish FTLD from HC, AD, APS, and PD in binary classification. Utilizing deep learning with multimodal neuroimaging data to differentiate FTLD subtypes and perform multiclassification among FTLD and other neurodegenerative disease holds promise for expediting diagnosis. In sum, the meta-analysis supports translation of machine learning tools in combination with imaging to clinical routine paving the way to precision medicine.
