EUVIMEDEUVIMED entdecken
Crossref

Improving brain tumor diagnosis: A self-calibrated 1D residual network with random forest integration

A Sumithra, P M Joe Prathap, A Karthikeyan, S Dhanasekaran

Brain Research · 2025

DOI: 10.1016/j.brainres.2025.149704 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Brain Research, 2025. Verfasst von A Sumithra und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science and Engineering, Sns College of Technology, Coimbatore, Tamil Nadu, India. Electronic address: sumithragvskr@gmail.com. Hauptthemen laut MEDLINE: Brain Neoplasms, Neural Networks, Computer. Die Arbeit nennt 33 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Medical specialists need to perform precise MRI analysis for accurate diagnosis of brain tumors. Current research has developed multiple artificial intelligence (AI) techniques for the process automation of brain tumor identification. However, existing approaches often depend on singular datasets, limiting their generalization capabilities across diverse clinical scenarios. The research introduces SCR-1DResNet as a new diagnostic tool for brain tumor detection that incorporates self-calibrated Random Forest along with one-dimensional residual networks. The research starts with MRI image acquisition from multiple Kaggle datasets then proceeds through stepwise processing that eliminates noise, enhances images, and performs resizing and normalization and conducts skull stripping operations. After data collection the WaveSegNet mode l extracts important attributes from tumors at multiple scales. Components of Random Forest classifier together with One-Dimensional Residual Network form the SCR-1DResNet model via self-calibration optimization to improve prediction reliability. Tests show the proposed system produces classification precision of 98.50% accompanied by accuracy of 98.80% and recall reaching 97.80% respectively. The SCR-1DResNet model demonstrates superior diagnostic capability and enhanced performance speed which shows strong prospects towards clinical decision support systems and improved neurological and oncological patient treatments.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Brain NeoplasmsNeural Networks, ComputerHumansMagnetic Resonance ImagingArtificial IntelligenceReproducibility of ResultsRandom Forest

Stichwörter der Autor:innen

Brain TumorDatasetsDeep LearningSelf Calibrated Random forestSingle-Dimensional Residual NetworkWaveSegNet

Autor:innen und Institutionen

  • Department of Computer Science and Engineering, Sns College of Technology, Coimbatore, Tamil Nadu, India. Electronic address: sumithragvskr@gmail.com.
  • Department of Computer Science and Engineering, Saptagiri NPS University, Bangalore, India.
  • Department of Computer Science and Engineering, Panimalar Engineering College, Chennai, Tamil Nadu, India.
  • Department of Information Technology, Kalasalingam Academy of Research and Education, Krishnankoil, Srivilliputtur, Tamil Nadu, India.

Zitieren

APA

Sumithra, A., Prathap, P. M. J., Karthikeyan, A., & Dhanasekaran, S. (2025). Improving brain tumor diagnosis: A self-calibrated 1D residual network with random forest integration. Brain Research. https://doi.org/10.1016/j.brainres.2025.149704

Vancouver

Sumithra A, Prathap PMJ, Karthikeyan A, Dhanasekaran S. Improving brain tumor diagnosis: A self-calibrated 1D residual network with random forest integration. Brain Research. 2025. doi:10.1016/j.brainres.2025.149704

BibTeX
@article{Sumithra2025,
  author = {Sumithra, A and Prathap, P M Joe and Karthikeyan, A and Dhanasekaran, S},
  title = {Improving brain tumor diagnosis: A self-calibrated 1D residual network with random forest integration},
  journal = {Brain Research},
  year = {2025},
  doi = {10.1016/j.brainres.2025.149704}
}
Kostenlos starten →