Skip to content
Open Imaging Index
MRI Brain

Ring-enhancing brain lesion MRI (UNESP/Unicamp)

Inflammatory lesions and brain tumors: is it possible to differentiate them based on texture features in magnetic resonance imaging?

Brain MRI of patients with a ring-enhancing lesion from one Brazilian medical school hospital, used to tell inflammatory lesions from brain tumors with texture features and machine learning. 67 patients on one 1.5T scanner. Not shared.

Go to dataset Edit Report

Overview

This retrospective cohort was assembled by researchers at São Paulo State University (UNESP, Botucatu) and the University of Campinas (Unicamp) to test whether texture features on routine brain MRI can separate inflammatory lesions from brain tumors when both show a ring-enhancing lesion with a non-enhancing core. All exams come from one 1.5T scanner in a medical school hospital in Brazil; the paper thanks the radiology staff of the Unicamp School of Medical Sciences. The images are not released.

Composition

67 patients scanned between 2010 and 2020 were included: 30 with inflammatory lesions and 37 with tumors. Diagnoses were confirmed by histopathology, cerebrospinal fluid analysis or follow-up imaging. The inflammatory group covers fungal, parasitic, bacterial and tuberculous infections (toxoplasmosis is the largest subgroup with 8 patients), 3 multiple sclerosis cases and single cases of neuromyelitis, progressive multifocal leukoencephalopathy and vasculitis. The tumor group consists of gliomas: 15 glioblastomas, 11 anaplastic astrocytomas, 8 low-grade astrocytomas, 2 gliosarcomas and 1 anaplastic ependymoma. Patients with prior biopsy or surgery, lesions under 10 mm, a visible scolex in neurocysticercosis or intracranial malformations were excluded. Ages and sex per patient are given in the article's supplementary files.

Acquisition

A Philips 1.5T scanner with an 8-channel head coil was used. The protocol included spin-echo T1-weighted images before and after gadolinium, fast spin-echo T2-weighted and proton density images, axial FLAIR and diffusion-weighted images with b = 1000. Axial T1, contrast-enhanced T1, T2, FLAIR and DWI were used for the analysis.

Annotations

Two radiologists with over 15 years of experience picked the axial slice with the largest lesion diameter and placed 10 x 10 pixel regions of interest inside the solid and necrotic parts of the lesion, excluding edema. 63 GLCM, run length and wavelet features were extracted per sequence and classified with SVM, kNN and random forest.

Known limitations

  • Not shared. The ethics approval limits the images to this study, and the paper states they are not available on request.
  • Single centre, single scanner and a small, heterogeneous set of diagnoses.
  • Annotations are small ROIs on one slice, not lesion masks.

Cohort

Aggregate numbers from the sources below. Bars are relative to the 67 subjects.

Condition

Groups can overlap

  • Brain tumor 37 55%
  • Glioma 37 55%
  • Central nervous system infection 24 36%
  • Glioblastoma 15 22%
  • Multiple sclerosis 3 4%

Table 1 in Alves et al. 2020, J Venom Anim Toxins incl Trop Dis 26, e20200011

Country

  • Brazil 67 100%

Affiliations; Methods (Image bank) in Alves et al. 2020, J Venom Anim Toxins incl Trop Dis 26, e20200011

Scanner vendor

  • Philips 67 100%

Methods (Image bank) in Alves et al. 2020, J Venom Anim Toxins incl Trop Dis 26, e20200011

Field strength

  • 1.5 T 67 100%

Methods (Image bank) in Alves et al. 2020, J Venom Anim Toxins incl Trop Dis 26, e20200011

License and access

Our reading of the license, not legal advice. Before you use the data, read the original license and confirm that your use is allowed. We take no responsibility for how you use a dataset. Full disclaimer

Access
Not shared

The data are not available outside the institution or study that holds them

The paper's availability statement says no patient data (DICOM images) are available on request, because the ethics committee approved their use for this study only. The linked figshare record holds only the article's figures and result tables.

Access page

No published license or data use terms

The data holder has published no license and no data use terms. Any use, sharing or commercial right has to be agreed with the data holder, and nothing can be assumed to be allowed. Default copyright and data protection law still apply.

The images are not released and no data use terms are published.

Original license text Checked 2026-10-08

What you can do

  • Commercial use Not stated
  • Not stated
  • Train ML models Not stated
  • Create derived data Not stated
  • Publish results Not stated

What you can share

  • Share the data Not stated
  • Share derived data Not stated
  • Share trained models Not stated

What you must do

  • Cite or credit Not stated
  • Share alike Not stated
  • Sign an agreement Not stated
  • Ethics approval Not stated
  • Manuscript review Not stated
  • Release code Not stated
  • Return results Not stated
  • Delete after use Not stated

Limits

  • No re-identification Not stated
  • Location limits Not stated

Citation

Alves AFF, Miranda JRA, Reis F, de Souza SAS, Alves LLR, Feitoza LM, de Castro JTS, de Pina DR. Inflammatory lesions and brain tumors: is it possible to differentiate them based on texture features in magnetic resonance imaging? J Venom Anim Toxins incl Trop Dis 26, e20200011 (2020). doi:10.1590/1678-9199-JVATITD-2020-0011

Sources

Every number on this page comes from one of these documents. Each chart names the table or page it is taken from. The raw numbers are in stats.csv.