Can large language models pass official high-grade exams of the European Society of Neuroradiology courses? A direct comparison between OpenAI chatGPT 3.5, OpenAI GPT4 and Google Bard.

We compared different LLMs, notably chatGPT, GPT4, and Google Bard and we tested whether their performance differs in subspeciality domains, in executing examinations from four different courses of the European Society of Neuroradiology (ESNR) notably anatomy/embryology, neuro-oncology, head and neck and pediatrics. Written exams of ESNR were used as input data, related to anatomy/embryology (30 questions), neuro-oncology (50 questions), head and neck (50 questions), and pediatrics (50 questions). All exams together, and each exam separately were introduced to the three LLMs: chatGPT 3.5, GPT4, and Google Bard. Statistical analyses included a group-wise Friedman test followed by a pair-wise Wilcoxon test with multiple comparison corrections. Overall, there was a significant difference between the 3 LLMs (p < 0.0001), with GPT4 having the highest accuracy (70%), followed by chatGPT 3.5 (54%) and Google Bard (36%). The pair-wise comparison showed significant differences between chatGPT vs GPT 4 (p < 0.0001), chatGPT vs Bard (p < 0. 0023), and GPT4 vs Bard (p < 0.0001). Analyses per subspecialty showed the highest difference between the best LLM (GPT4, 70%) versus the worst LLM (Google Bard, 24%) in the head and neck exam, while the difference was least pronounced in neuro-oncology (GPT4, 62% vs Google Bard, 48%). We observed significant differences in the performance of the three different LLMs in the running of official exams organized by ESNR. Overall GPT 4 performed best, and Google Bard performed worst. This difference varied depending on subspeciality and was most pronounced in head and neck subspeciality.

Multiple Ring-Enhancing Brain Lesions: Fulminant Diffuse Cerebral Toxoplasmosis or Cerebral Metastases?

ABSTRACT: In this case report, we describe a 76-year-old woman, presenting with dizziness for the past 2 months, without other focal neurological signs. A magnetic resonance imaging of the brain was ordered by her GP. The MRI demonstrated multiple ring-enhancing lesions, both supratentorial and infratentorial. Lumbar puncture showed normal findings, in particular a normal cell count and culture. Because of the radiologic appearance, initially thought to be suggestive of cerebral abscesses, antibiotics were started. However, further workup revealed a new diagnosis of a stage IV (metastatic) small cell lung carcinoma, making diffuse brain metastases more likely. The patient was transferred to oncology/pneumology, where she was started on whole-brain radiotherapy, after which systemic therapy would start. However, because of further clinical deterioration, she was admitted at the palliative ward, where she died only 3 months after the initial presentation. In this case report, we emphasize the importance of keeping a broad differential diagnosis and briefly review the various possible pathologies causing ring-enhancing lesions.

From Mild Gait Difficulties to a Sudden Coma: A Rare Case of Marchiafava-Bignami Disease.

ABSTRACT: In this case report we describe the case of a 66-year old man with subacute gait difficulties, with a progression to confusion coma with multiple generalised epileptic seizures during the following days. Biochemical analysis showed hyperglycaemia, cerebrospinal fluid (CSF) testing showed a mild lymphocytic pleocytosis and an elevated protein and lactate. Broad-spectrum antibiotics and antiviral therapy where initiated. However, all other CSF testing remained negative. Magnetic resonance imaging of the brain showed remarkably symmetric hyperintense T2 white matter lesions most noticable in the corpus callosum. The lesion pattern was suggestive of a metabolic or toxic encephalopathy, the preponderance for the corpus callosum was furthermore suggestive for Marchiafava-Bignami disease (MDB), as was the clinical course since admission of the patient. A high dose IV substitution of vitamin B1, B6 and B12 was started and antibiotic and antiviral therapy was discontinued. After one day the patient showed progressive regaining of consciousness and he returned to premorbid functioning in a matter of 1-2 weeks. MRI of the brain after 1 week showed notable improvement of the white matter lesions. At routine follow-up two weeks later he presented with icterus and a diagnosis of Epstein-Barr virus (EBV) hepatitis was made, lymph node biopsies showed an EBV positive diffuse large cell B-cell lymphoma (DLCBL). MDB is mostly associated with severe alcoholism, with malnourishment being the second leading cause, however there are case reports describing MDB in patients with chronically poorly controlled diabetes mellitus. We hypothesize that his condition may have been precipitated by his poorly controlled diabetes mellitus. However it is also possible that weight loss (probably related to the DLCBL diagnosis) might have contributed to a state of malnourishment and therefore played a role in the aetiology as well.

Guidelines for magnetic resonance imaging in pediatric head and neck pathologies: a multicentre international consensus paper.

The use of standardized imaging protocols is paramount in order to facilitate comparable, reproducible images and, consequently, to optimize patient care. Standardized MR protocols are lacking when studying head and neck pathologies in the pediatric population. We propose an international, multicenter consensus paper focused on providing the best combination of acquisition time/technical requirements and image quality. Distinct protocols for different regions of the head and neck and, in some cases, for specific pathologies or clinical indications are recommended. This white paper is endorsed by several international scientific societies and it is the result of discussion, in consensus, among experts in pediatric head and neck imaging.

DNA methylation based glioblastoma subclassification is related to tumoral T-cell infiltration and patient survival.

BACKGROUND: Histologically classified glioblastomas (GBM) can have different clinical behavior and response to therapy, for which molecular subclassifications have been proposed. We evaluated the relationship of epigenetic GBM subgroups with immune cell infiltrations, systemic immune changes during radiochemotherapy, and clinical outcome. METHODS: 450K genome-wide DNA methylation was assessed on tumor tissue from 93 patients with newly diagnosed GBM, treated with standard radiochemotherapy and experimental immunotherapy. Tumor infiltration of T cells, myeloid cells, and Programmed cell death protein 1 (PD-1) expression were evaluated. Circulating immune cell populations and selected cytokines were assessed on blood samples taken before and after radiochemotherapy. RESULTS: Forty-two tumors had a mesenchymal, 27 a receptor tyrosine kinase (RTK) II, 17 RTK I, and 7 an isocitrate dehydrogenase (IDH) DNA methylation pattern. Mesenchymal tumors had the highest amount of tumor-infiltrating CD3+ and CD8+ T cells and IDH tumors the lowest. There were no significant differences for CD68+ cells, FoxP3+ cells, and PD-1 expression between groups. Systemically, there was a relative increase of CD8+ T cells and CD8+ PD-1 expression and a relative decrease of CD4+ T cells after radiochemotherapy in all subgroups except IDH tumors. Overall survival was the longest in the IDH group (median 36 mo), intermediate in RTK II tumors (27 mo), and significantly lower in mesenchymal and RTK I groups (15.5 and 16 mo, respectively). CONCLUSIONS: Methylation based stratification of GBM is related to T-cell infiltration and survival, with IDH and mesenchymal tumors representing both ends of a spectrum. DNA methylation profiles could be useful in stratifying patients for immunotherapy trials.

Bilateral lesions of the basal ganglia and thalami (central grey matter)-pictorial review.

The basal ganglia and thalami are paired deep grey matter structures with extensive metabolic activity that renders them susceptible to injury by various diseases. Most pathological processes lead to bilateral lesions, which may be symmetric or asymmetric, frequently showing characteristic patterns on imaging studies. In this comprehensive pictorial review, the most common and/or typical genetic, acquired metabolic/toxic, infectious, inflammatory, vascular and neoplastic pathologies affecting the central grey matter are subdivided according to the preferential location of the lesions: in the basal ganglia, in the thalami or both. The characteristic imaging findings are described with emphasis on the differential diagnosis and clinical context.

The diagnostic role of diffusional kurtosis imaging in glioma grading and differentiation of gliomas from other intra-axial brain tumours: a systematic review with critical appraisal and meta-analysis.

PURPOSE: We aim to illustrate the diagnostic performance of diffusional kurtosis imaging (DKI) in the diagnosis of gliomas. METHODS: A review protocol was developed according to the (PRISMA-P) checklist, registered in the international prospective register of systematic reviews (PROSPERO) and published. A literature search in 4 databases was performed using the keywords 'glioma' and 'diffusional kurtosis'. After applying a robust inclusion/exclusion criteria, included articles were independently evaluated according to the QUADAS-2 tool and data extraction was done. Reported sensitivities and specificities were used to construct 2 × 2 tables and paired forest plots using the Review Manager (RevMan®) software. A random-effect model was pursued using the hierarchical summary receiver operator characteristics. RESULTS: A total of 216 hits were retrieved. Considering duplicates and inclusion criteria, 23 articles were eligible for full-text reading. Ultimately, 19 studies were eligible for final inclusion. The quality assessment revealed 9 studies with low risk of bias in the 4 domains. Using a bivariate random-effect model for data synthesis, summary ROC curve showed a pooled area under the curve (AUC) of 0.92 and estimated sensitivity of 0.87 (95% CI 0.78-0.92) in high-/low-grade gliomas' differentiation. A mean difference in mean kurtosis (MK) value between HGG and LGG of 0.22 (95% CI 0.25-0.19) was illustrated (p value = 0.0014) with moderate heterogeneity (I2 = 73.8%). CONCLUSION: DKI shows good diagnostic accuracy in the differentiation of high- and low-grade gliomas further supporting its potential role in clinical practice. Further exploration of DKI in differentiating IDH status and in characterising non-glioma CNS tumours is however needed.

Chordoid Glioma as a Differential Diagnosis of Anterior Third Ventricle Tumours: A Rare Case Report and Five-Year Follow-Up.

Chordoid glioma is a rare and relatively recently defined tumour entity. Worldwide there have only been around 90 cases described until now. A chordoid glioma comprises a low-grade suprasellar neuroepithelial neoplasm originating in the anterior part of the third ventricle, with consistent radiological features on MRI. This lesion should be considered as a differential of third ventricle tumours. The patient described in this paper is quite unique in the sense that despite only partial tumour resection was obtained, the residual tumour was not progressive during several years of follow-up. Preoperative recognition of this disease entity is crucial to modify the treatment approach and improve patient outcome.

Glioblastoma with extracranial parotid, lymph node, and pulmonary metastases: a case report.

We present a rare case of an isocitrate dehydrogenase-wildtype glioblastoma with histologically proven parotid, cervical lymph node, and lung metastases. While recent therapy advances are likely to increase glioblastoma mid- and long-term survival, this will also increase the time window for extraneural glioblastoma spread. Radiologists should be aware of this risk, so they can accurately detect and interpret metastatic glioblastoma disease.

Retro-Odontoid Pseudotumor in a Patient with Atlanto-Occipital Assimilation.

A retro-odontoid pseudotumor is an uncommon non-neoplastic mass. They are mostly associated with rheumatoid arthritis and atlanto-axial subluxation. The pathogenesis is degeneration of the transverse ligament due to chronic mechanical stress. In this case report, an atlanto-occipital assimilation altered the biomechanics of the cervical spine, causing chronic mechanical stress on the transverse ligament and subsequently the development of a retro-odontoid pseudotumor. This is in accordance with previous studies that have attributed the development of retro-odontoid pseudotumor to a loss of mobility of the cervical spine, in cases without associated rheumatoid arthritis or atlanto-axial subluxation.

An advanced MRI and MRSI data fusion scheme for enhancing unsupervised brain tumor differentiation.

Proton Magnetic Resonance Spectroscopic Imaging (1H MRSI) has shown great potential in tumor diagnosis since it provides localized biochemical information discriminating different tissue types, though it typically has low spatial resolution. Magnetic Resonance Imaging (MRI) is widely used in tumor diagnosis as an in vivo tool due to its high resolution and excellent soft tissue discrimination. This paper presents an advanced data fusion scheme for brain tumor diagnosis using both MRSI and MRI data to improve the tumor differentiation accuracy of MRSI alone. Non-negative Matrix Factorization (NMF) of the spectral feature vectors from MRSI data and the image fusion with MRI based on wavelet analysis are implemented jointly. Hence, it takes advantage of the biochemical tissue discrimination of MRSI as well as the high resolution of MRI. The feasibility of the proposed frame work is validated by comparing with the expert delineations, giving mean correlation coefficients for the tumor source of 0.97 and the Dice score of tumor region overlap of 0.90. These results compare favorably against those obtained with a previously proposed NMF method where MRSI and MRI are integrated by stacking the MRSI and MRI features.

Classifying Glioblastoma Multiforme Follow-Up Progressive vs. Responsive Forms Using Multi-Parametric MRI Features.

Purpose: The purpose of this paper is discriminating between tumor progression and response to treatment based on follow-up multi-parametric magnetic resonance imaging (MRI) data retrieved from glioblastoma multiforme (GBM) patients. Materials and Methods: Multi-parametric MRI data consisting of conventional MRI (cMRI) and advanced MRI [i.e., perfusion weighted MRI (PWI) and diffusion kurtosis MRI (DKI)] were acquired from 29 GBM patients treated with adjuvant therapy after surgery. We propose an automatic pipeline for processing advanced MRI data and extracting intensity-based histogram features and 3-D texture features using manually and semi-manually delineated regions of interest (ROIs). Classifiers are trained using a leave-one-patient-out cross validation scheme on complete MRI data. Balanced accuracy rate (BAR)-values are computed and compared between different ROIs, MR modalities, and classifiers, using non-parametric multiple comparison tests. Results: Maximum BAR-values using manual delineations are 0.956, 0.85, 0.879, and 0.932, for cMRI, PWI, DKI, and all three MRI modalities combined, respectively. Maximum BAR-values using semi-manual delineations are 0.932, 0.894, 0.885, and 0.947, for cMRI, PWI, DKI, and all three MR modalities combined, respectively. After statistical testing using Kruskal-Wallis and post-hoc Dunn-Sidák analysis we conclude that training a RUSBoost classifier on features extracted using semi-manual delineations on cMRI or on all MRI modalities combined performs best. Conclusions: We present two main conclusions: (1) using T1 post-contrast (T1pc) features extracted from manual total delineations, AdaBoost achieves the highest BAR-value, 0.956; (2) using T1pc-average, T1pc-90th percentile, and Cerebral Blood Volume (CBV) 90th percentile extracted from semi-manually delineated contrast enhancing ROIs, SVM-rbf, and RUSBoost achieve BAR-values of 0.947 and 0.932, respectively. Our findings show that AdaBoost, SVM-rbf, and RUSBoost trained on T1pc and CBV features can differentiate progressive from responsive GBM patients with very high accuracy.

Hierarchical non-negative matrix factorization to characterize brain tumor heterogeneity using multi-parametric MRI.

Tissue characterization in brain tumors and, in particular, in high-grade gliomas is challenging as a result of the co-existence of several intra-tumoral tissue types within the same region and the high spatial heterogeneity. This study presents a method for the detection of the relevant tumor substructures (i.e. viable tumor, necrosis and edema), which could be of added value for the diagnosis, treatment planning and follow-up of individual patients. Twenty-four patients with glioma [10 low-grade gliomas (LGGs), 14 high-grade gliomas (HGGs)] underwent a multi-parametric MRI (MP-MRI) scheme, including conventional MRI (cMRI), perfusion-weighted imaging (PWI), diffusion kurtosis imaging (DKI) and short-TE (1)H MRSI. MP-MRI parameters were derived: T2, T1 + contrast, fluid-attenuated inversion recovery (FLAIR), relative cerebral blood volume (rCBV), mean diffusivity (MD), fractional anisotropy (FA), mean kurtosis (MK) and the principal metabolites lipids (Lip), lactate (Lac), N-acetyl-aspartate (NAA), total choline (Cho), etc. Hierarchical non-negative matrix factorization (hNMF) was applied to the MP-MRI parameters, providing tissue characterization on a patient-by-patient and voxel-by-voxel basis. Tissue-specific patterns were obtained and the spatial distribution of each tissue type was visualized by means of abundance maps. Dice scores were calculated by comparing tissue segmentation derived from hNMF with the manual segmentation by a radiologist. Correlation coefficients were calculated between each pathologic tissue source and the average feature vector within the corresponding tissue region. For the patients with HGG, mean Dice scores of 78%, 85% and 83% were obtained for viable tumor, the tumor core and the complete tumor region. The mean correlation coefficients were 0.91 for tumor, 0.97 for necrosis and 0.96 for edema. For the patients with LGG, a mean Dice score of 85% and mean correlation coefficient of 0.95 were found for the tumor region. hNMF was also applied to reduced MRI datasets, showing the added value of individual MRI modalities.

Tumour Relapse Prediction Using Multiparametric MR Data Recorded during Follow-Up of GBM Patients.

PURPOSE: We have focused on finding a classifier that best discriminates between tumour progression and regression based on multiparametric MR data retrieved from follow-up GBM patients. MATERIALS AND METHODS: Multiparametric MR data consisting of conventional and advanced MRI (perfusion, diffusion, and spectroscopy) were acquired from 29 GBM patients treated with adjuvant therapy after surgery over a period of several months. A 27-feature vector was built for each time point, although not all features could be obtained at all time points due to missing data or quality issues. We tested classifiers using LOPO method on complete and imputed data. We measure the performance by computing BER for each time point and wBER for all time points. RESULTS: If we train random forests, LogitBoost, or RobustBoost on data with complete features, we can differentiate between tumour progression and regression with 100% accuracy, one time point (i.e., about 1 month) earlier than the date when doctors had put a label (progressive or responsive) according to established radiological criteria. We obtain the same result when training the same classifiers solely on complete perfusion data. CONCLUSIONS: Our findings suggest that ensemble classifiers (i.e., random forests and boost classifiers) show promising results in predicting tumour progression earlier than established radiological criteria and should be further investigated.

Assessing Treatment Response Through Generalized Pharmacokinetic Modeling of DCE-MRI Data.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) enables the quantification of contrast leakage from the vascular tissue by using pharmacokinetic (PK) models. Such quantitative analysis of DCE-MRI data provides physiological parameters that are able to provide information of tumor pathophysiology and therapeutic outcome. Several assumptive PK models have been proposed to characterize microcirculation in the tumoral tissue. In this paper, we present a comparative study between the well-known extended Tofts model (ETM) and the more recent gamma capillary transit time (GCTT) model, with the latter showing initial promising results in the literature. To enhance the GCTT imaging biomarkers, we introduce a novel method for segmenting the tumor area into subregions according to their vascular heterogeneity characteristics. A cohort of 11 patients diagnosed with glioblastoma multiforme with known therapeutic outcome was used to assess the predictive value of both models in terms of correctly classifying responders and nonresponders based on only one DCE-MRI examination. The results indicate that GCTT model's PK parameters perform better than those of ETM, while the segmentation of the tumor regions of interest based on vascular heterogeneity further enhances the discriminatory power of the GCTT model.

A Proposed Paradigm Shift in Initializing Cancer Predictive Models with DCE-MRI Based PK Parameters: A Feasibility Study.

Glioblastoma multiforme is the most aggressive type of glioma and the most common malignant primary intra-axial brain tumor. In an effort to predict the evolution of the disease and optimize therapeutical decisions, several models have been proposed for simulating the growth pattern of glioma. One of the latest models incorporates cell proliferation and invasion, angiogenic net rates, oxygen consumption, and vasculature. These factors, particularly oxygenation levels, are considered fundamental factors of tumor heterogeneity and compartmentalization. This paper focuses on the initialization of the cancer cell populations and vasculature based on imaging examinations of the patient and presents a feasibility study on vasculature prediction over time. To this end, pharmacokinetic parameters derived from dynamic contrast-enhanced magnetic resonance imaging using Toft's model are used in order to feed the model. K (trans) is used as a metric of the density of endothelial cells (vasculature); at the same time, it also helps to discriminate distinct image areas of interest, under a set of assumptions. Feasibility results of applying the model to a real clinical case are presented, including a study on the effect of certain parameters on the pattern of the simulated tumor.

The use of rapid prototyped implants to simulate knee joint abnormalities for in vitro testing: a validation study with replica implants of the native trochlea.

To investigate the biomechanical effect of skeletal knee joint abnormalities, the authors propose to implant pathologically shaped rapid prototyped implants in cadaver knee specimens. This new method was validated by replacing the native trochlea by a replica implant on four cadaver knees with the aid of cadaver-specific guiding instruments. The accuracy of the guiding instruments was assessed by measuring the rotational errors of the cutting planes (on average 3.01° in extension and 1.18° in external/internal rotation). During a squat and open chain simulation, the patella showed small differences in its articulation with the native trochlea and the replica trochlea, which could partially be explained by the rotational errors of the implants. This study concludes that this method is valid to investigate the effect of knee joint abnormalities with a replica implant as a control condition to account for the influence of material properties and rotational errors of the implant.

Data analysis and tissue type assignment for glioblastoma multiforme.

Glioblastoma multiforme (GBM) is characterized by high infiltration. The interpretation of MRSI data, especially for GBMs, is still challenging. Unsupervised methods based on NMF by Li et al. (2013, NMR in Biomedicine) and Li et al. (2013, IEEE Transactions on Biomedical Engineering) have been proposed for glioma recognition, but the tissue types is still not well interpreted. As an extension of the previous work, a tissue type assignment method is proposed for GBMs based on the analysis of MRSI data and tissue distribution information. The tissue type assignment method uses the values from the distribution maps of all three tissue types to interpret all the information in one new map and color encodes each voxel to indicate the tissue type. Experiments carried out on in vivo MRSI data show the feasibility of the proposed method. This method provides an efficient way for GBM tissue type assignment and helps to display information of MRSI data in a way that is easy to interpret.