scholarly journals Machine Learning-based Texture Analysis of Contrast-enhanced MR Imaging to Differentiate between Glioblastoma and Primary Central Nervous System Lymphoma

2019 ◽  
Vol 18 (1) ◽  
pp. 44-52 ◽  
Author(s):  
Akira Kunimatsu ◽  
Natsuko Kunimatsu ◽  
Koichiro Yasaka ◽  
Hiroyuki Akai ◽  
Kouhei Kamiya ◽  
...  
2021 ◽  
pp. 197140092199897
Author(s):  
Sarv Priya ◽  
Caitlin Ward ◽  
Thomas Locke ◽  
Neetu Soni ◽  
Ravishankar Pillenahalli Maheshwarappa ◽  
...  

Objectives To evaluate the diagnostic performance of multiple machine learning classifier models derived from first-order histogram texture parameters extracted from T1-weighted contrast-enhanced images in differentiating glioblastoma and primary central nervous system lymphoma. Methods Retrospective study with 97 glioblastoma and 46 primary central nervous system lymphoma patients. Thirty-six different combinations of classifier models and feature selection techniques were evaluated. Five-fold nested cross-validation was performed. Model performance was assessed for whole tumour and largest single slice using receiver operating characteristic curve. Results The cross-validated model performance was relatively similar for the top performing models for both whole tumour and largest single slice (area under the curve 0.909–0.924). However, there was a considerable difference between the worst performing model (logistic regression with full feature set, area under the curve 0.737) and the highest performing model for whole tumour (least absolute shrinkage and selection operator model with correlation filter, area under the curve 0.924). For single slice, the multilayer perceptron model with correlation filter had the highest performance (area under the curve 0.914). No significant difference was seen between the diagnostic performance of the top performing model for both whole tumour and largest single slice. Conclusions T1 contrast-enhanced derived first-order texture analysis can differentiate between glioblastoma and primary central nervous system lymphoma with good diagnostic performance. The machine learning performance can vary significantly depending on the model and feature selection methods. Largest single slice and whole tumour analysis show comparable diagnostic performance.


2006 ◽  
Vol 21 (5) ◽  
pp. 1-9 ◽  
Author(s):  
April F. Eichler ◽  
Tracy T. Batchelor

✓ Primary central nervous system lymphoma (PCNSL) is a rare form of non-Hodgkin lymphoma that affects the brain, spinal cord, leptomeninges, and eyes. The clinical presentation and neuroimaging appearance of PCNSL differ in immunocompetent patients and in those with acquired immunodeficiency syndrome (AIDS). A magnetic resonance (MR) image of the brain in immunocompetent patients with PCNSL typically demonstrates one or more homogeneously enhancing lesions located in the periventricular white matter, characteristically spanning the corpus callosum. In patients with AIDS, multiple ring-enhancing lesions are more common. After neuroimages raising the suspicion of PCNSL are obtained, a definitive diagnosis should be established in both immunocompetent and AIDS patients by performing pathological analysis of cerebrospinal fluid (CSF), vitreous fluid, or a biopsy specimen. Brain biopsy sampling remains the gold standard for PCNSL diagnosis in all patients, although the possibility of establishing routine, minimally invasive diagnostic procedures in which Epstein–Barr virus polymerase chain reaction (PCR) analysis of the CSF and nuclear imaging are used is currently under investigation in the population of patients with AIDS. At the time of diagnosis, the patient should undergo further evaluation, which should include a physical examination, ophthalmic evaluation with a slit-lamp examination, serum lactate dehydrogenase levels, human immunodeficiency virus testing, computed tomography scans of the chest/abdomen/pelvis, bone marrow biopsy sampling, contrast-enhanced brain MR imaging, and lumbar puncture (LP). Testicular ultrasonography studies should be considered in men. In patients who cannot undergo LP or in those with evidence of spinal cord dysfunction, contrast-enhanced MR imaging of the entire spine should be considered.


2013 ◽  
Vol 2013 ◽  
pp. 1-5 ◽  
Author(s):  
G. Manenti ◽  
F. Di Giuliano ◽  
A. Bindi ◽  
V. Liberto ◽  
V. Funel ◽  
...  

Primary central nervous system lymphomas (PCNSLs) are mainly B-cells lymphomas. A risk factor for the development of PCNSL is immunodeficiency, which includes congenital disorders, iatrogenic immunosuppression, and HIV. The clinical course is rapidly fatal; these patients usually present signs of increased intracranial pressure, nausea, papilledema, vomiting, and neurological and neuropsychiatric symptoms. PCNSL may have a characteristic appearance on CT and MR imaging. DWI sequences and MR spectroscopy may help to differentiate CNS lymphomas from other brain lesions. In this paper, we report a case of a 23-year-old man with T-primary central nervous system lymphoma presenting with a mass in the right frontotemporal lobe. We describe clinical, CT, and MRI findings. Diagnosis was confirmed by stereotactic biopsy of the lesion.


Radiology ◽  
2014 ◽  
Vol 272 (3) ◽  
pp. 843-850 ◽  
Author(s):  
Philipp Kickingereder ◽  
Benedikt Wiestler ◽  
Felix Sahm ◽  
Sabine Heiland ◽  
Matthias Roethke ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document