Machine learning based radiomics approach for outcome prediction of meningioma – a systematic review DOI Creative Commons

S Saroh,

Saikiran Pendem,

K Prakashini

et al.

F1000Research, Journal Year: 2025, Volume and Issue: 14, P. 330 - 330

Published: March 25, 2025

Introduction Meningioma is the most common brain tumor in adults. Magnetic resonance imaging (MRI) preferred modality for assessing outcomes. Radiomics, an advanced technique, assesses heterogeneity and identifies predictive markers, offering a non-invasive alternative to biopsies. Machine learning (ML) based radiomics models enhances diagnostic prognostic accuracy of tumors. Comprehensive review on ML-based predicting meningioma recurrence survival are lacking. Hence, aim study summarize performance measures ML prediction outcomes such as progression/recurrence (P/R) overall analysis meningioma. Methods Data bases Scopus, Web Science, PubMed, Embase were used conduct literature search order find pertinent original articles that concentrated outcome prediction. PRISMA (Preferred reporting items systematic reviews meta-analysis) recommendations extract data from selected studies. Results Eight included study. MRI Radiomics-based combined with clinical pathological showed strong recurrence. A decision tree model achieved 90% accuracy, outperforming apparent diffusion coefficient (ADC) (83%). support vector machine (SVM) reached area under curve (AUC) 0.80 radiomic features, improving 0.88 ADC integration. clinico-pathological (CPRM) AUC testing. Key predictors include values, scores, ki-67 index, Simpson grading. For meningioma, clinicopathological features 0.78. Conclusion Integrating through greatly improved These surpass conventional aggressiveness, providing crucial insights personalized treatment surgical planning.

Language: Английский

Machine learning based radiomics approach for outcome prediction of meningioma – a systematic review DOI Creative Commons

S Saroh,

Saikiran Pendem,

K Prakashini

et al.

F1000Research, Journal Year: 2025, Volume and Issue: 14, P. 330 - 330

Published: March 25, 2025

Introduction Meningioma is the most common brain tumor in adults. Magnetic resonance imaging (MRI) preferred modality for assessing outcomes. Radiomics, an advanced technique, assesses heterogeneity and identifies predictive markers, offering a non-invasive alternative to biopsies. Machine learning (ML) based radiomics models enhances diagnostic prognostic accuracy of tumors. Comprehensive review on ML-based predicting meningioma recurrence survival are lacking. Hence, aim study summarize performance measures ML prediction outcomes such as progression/recurrence (P/R) overall analysis meningioma. Methods Data bases Scopus, Web Science, PubMed, Embase were used conduct literature search order find pertinent original articles that concentrated outcome prediction. PRISMA (Preferred reporting items systematic reviews meta-analysis) recommendations extract data from selected studies. Results Eight included study. MRI Radiomics-based combined with clinical pathological showed strong recurrence. A decision tree model achieved 90% accuracy, outperforming apparent diffusion coefficient (ADC) (83%). support vector machine (SVM) reached area under curve (AUC) 0.80 radiomic features, improving 0.88 ADC integration. clinico-pathological (CPRM) AUC testing. Key predictors include values, scores, ki-67 index, Simpson grading. For meningioma, clinicopathological features 0.78. Conclusion Integrating through greatly improved These surpass conventional aggressiveness, providing crucial insights personalized treatment surgical planning.

Language: Английский

Citations

0