Evaluating the Efficacy of On-Device AI in Radiological Image Classification: Using CoreML for Brain Tumor Detection

Authors

Keywords:

Artificial Intelligence, Radiology, Magnetic Resonance Imaging (MRI), Brain Neoplasms, Image Interpretation, Computer-Assisted, Machine Learning, Pattern Recognition Automated

Abstract

Background: The rapidly expanding prevalence of AI in patient care and treatment planning reflects its transformative potential in improving healthcare outcomes and efficiency. AI applications in radiology can aid in accuracy of diagnosis and alleviating physician workload. However, one barrier to integration of AI is its financial burden due to subscription-based software costs, ongoing updates, and hardware requirements. A potential way to offset these costs is the adoption of local, on-device machine learning workflows, implemented through accessible development environments such as Apple’s Xcode in combination with Create ML and the Core ML framework. 

 

Methods: This project evaluated the accuracy of a Core ML–based image classification model for detecting brain tumors on MRI images and assessed the feasibility of leveraging the Xcode development environment to train machine learning (ML) image classification models with Create ML. Utilizing de-identified, publicly available brain MRI datasets, a supervised image classification model developed using Apple’s Create ML Image Classifier was trained to differentiate between glioma, meningioma, pituitary tumors, and healthy brain images. Standardization of images was done using Create ML Image Feature Print v2, which automatically scaled each image to 360 x 360 pixels. A total of 7,023 images were in the dataset with 5,970 (85%) images utilized for training and 1,053 (15%) images used for testing. 

 

Results: After 21 training iterations, the classifier achieved 93% accuracy on the test images, with precisions of 96% for healthy, 95% for pituitary tumor, 93% for glioma, and 87% for meningioma. 

 

Conclusion: The model demonstrates consistently higher performance across all categories, with slightly lower scores for glioma and meningioma. As shown by our CoreML model achieving 93% accuracy in identifying brain MRI tumor subtypes, utilizing the Xcode development environment and CoreML is a viable and promising way for future development of effective and cost-efficient diagnostic radiology tools for brain tumor detection.

Downloads

Published

2026-08-19

How to Cite

Ninan, D., Khandavilli, V., Castro, R., Kansal , P., Memon, A., Lin, J., … Havo, D. (2026). Evaluating the Efficacy of On-Device AI in Radiological Image Classification: Using CoreML for Brain Tumor Detection. International Journal of Medical Students. Retrieved from https://ijms.info/IJMS/article/view/4904

Issue

Forthcoming

Section

Abstracts of the WCMSR

Categories

License

Copyright (c) 2026 Daniel Ninan, Vaishnavi Khandavilli, Ralph Castro, Paarth Kansal , Ammarah Memon, Jessica Lin, Bridget Le, Daisy Havo

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

  1. The Author retains copyright in the Work, where the term “Work” shall include all digital objects that may result in subsequent electronic publication or distribution.
  2. Upon acceptance of the Work, the author shall grant to the Publisher the right of first publication of the Work.
  3. The Author shall grant to the Publisher and its agents the nonexclusive perpetual right and license to publish, archive, and make accessible the Work in whole or in part in all forms of media now or hereafter known under a Creative Commons Attribution 4.0 International License or its equivalent, which, for the avoidance of doubt, allows others to copy, distribute, and transmit the Work under the following conditions:
    1. Attribution—other users must attribute the Work in the manner specified by the author as indicated on the journal Web site; with the understanding that the above condition can be waived with permission from the Author and that where the Work or any of its elements is in the public domain under applicable law, that status is in no way affected by the license.
    2. The Author is able to enter into separate, additional contractual arrangements for the nonexclusive distribution of the journal's published version of the Work (e.g., post it to an institutional repository or publish it in a book), as long as there is provided in the document an acknowledgment of its initial publication in this journal.
    3. Authors are permitted and encouraged to post online a prepublication manuscript (but not the Publisher’s final formatted PDF version of the Work) in institutional repositories or on their Websites prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work. Any such posting made before acceptance and publication of the Work shall be updated upon publication to include a reference to the Publisher-assigned DOI (Digital Object Identifier) and a link to the online abstract for the final published Work in the Journal.
    4. Upon Publisher’s request, the Author agrees to furnish promptly to Publisher, at the Author’s own expense, written evidence of the permissions, licenses, and consents for use of third-party material included within the Work, except as determined by Publisher to be covered by the principles of Fair Use.
    5. The Author represents and warrants that:
      1. the Work is the Author’s original work;
      2. the Author has not transferred, and will not transfer, exclusive rights in the Work to any third party;
      3. the Work is not pending review or under consideration by another publisher;
      4. the Work has not previously been published;
      5. the Work contains no misrepresentation or infringement of the Work or property of other authors or third parties; and
      6. the Work contains no libel, invasion of privacy, or other unlawful matter.
    6. The Author agrees to indemnify and hold Publisher harmless from the Author’s breach of the representations and warranties contained in Paragraph 6 above, as well as any claim or proceeding relating to Publisher’s use and publication of any content contained in the Work, including third-party content.

Enforcement of copyright

The IJMS takes the protection of copyright very seriously.

If the IJMS discovers that you have used its copyright materials in contravention of the license above, the IJMS may bring legal proceedings against you seeking reparation and an injunction to stop you using those materials. You could also be ordered to pay legal costs.

If you become aware of any use of the IJMS' copyright materials that contravenes or may contravene the license above, please report this by email to contact@ijms.org

 

Infringing material

If you become aware of any material on the website that you believe infringes your or any other person's copyright, please report this by email to contact@ijms.org