Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Enhancing brain tumor classification with a simplified CNN through hyperparameter optimization.

Nihal Remzan1,2, Karim Tahiry2, Abdelmajid Farchi2

  • 1Innovation in Mathematics and Intelligent Systems Laboratory, Ibn Zohr University, Agadir, Morocco.

Biomedical Physics & Engineering Express
|May 29, 2026
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same journal

Neuroinformed spectro-temporal-spatial attention with bi-hemispheric learning for EEG emotion recognition.

Biomedical physics & engineering express·2026
Same journal

Prediction of conduction velocity distribution and motor unit action potential from evoked compound muscle action potential: a novel metaheuristic-based framework.

Biomedical physics & engineering express·2026
Same journal

CRGFFNet: An adaptive inter-channel relation guided feature fusion network for enhanced EEG personality recognition.

Biomedical physics & engineering express·2026
Same journal

Exploration on early diagnosis of cerebral hemorrhage: a new method based on a portable sensor using vortex damping technology.

Biomedical physics & engineering express·2026
Same journal

A retinal vessel segmentation network with multiscale feature extraction and cross-layer attention fusion (MAF-Net).

Biomedical physics & engineering express·2026
Same journal

Customized 3D-printed breast prostheses using patient imaging at the point of care: feasibility and proof-of-concept study.

Biomedical physics & engineering express·2026

Optimizing deep learning models for brain tumor classification using Magnetic Resonance Imaging (MRI) is crucial. HyperBand optimization achieved the highest accuracy for Convolutional Neural Networks (CNNs), outperforming Genetic Algorithms and Bayesian Optimization.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computational Biology

Background:

  • Brain tumors pose a significant public health challenge, necessitating accurate and timely diagnosis.
  • Conventional Magnetic Resonance Imaging (MRI) analysis relies on subjective radiologist interpretation, leading to variability.
  • Deep learning, specifically Convolutional Neural Networks (CNNs), offers automated analysis of medical images for improved accuracy.

Purpose of the Study:

  • To investigate the impact of hyperparameter optimization strategies on CNN performance for brain tumor classification from MRI.
  • To compare Genetic Algorithms, Bayesian Optimization, and HyperBand for optimizing a lightweight CNN model.
  • To assess the practical applicability and efficiency of different optimization methods.

Main Methods:

Keywords:
CNNMRIbrain tumordeep learninghyperband algorithm.hyperparameter optimization

Related Experiment Videos

  • A simplified, lightweight CNN architecture was developed for brain tumor classification.
  • Three hyperparameter optimization strategies were employed: Genetic Algorithms, Bayesian Optimization, and HyperBand.
  • Experiments were conducted on a public MRI dataset including glioma, meningioma, pituitary tumors, and healthy controls.

Main Results:

  • Genetic Algorithm achieved 96.41% ± 0.36% mean accuracy.
  • Bayesian Optimization improved performance to 97.37% ± 0.30% mean accuracy.
  • HyperBand demonstrated the highest average accuracy at 97.73% ± 0.48%, indicating superior hyperparameter configuration identification.

Conclusions:

  • Hyperparameter optimization significantly impacts CNN performance in brain tumor classification.
  • HyperBand is an effective strategy for optimizing lightweight CNNs for MRI-based brain tumor detection.
  • Automated deep learning approaches show promise in enhancing the accuracy and reliability of brain tumor diagnosis.