An MRI-based joint model of radiomics and spatial distribution differentiates autoimmune encephalitis from low-grade diffuse astrocytoma

Detection of antibodies in suspected autoimmune encephalitis diseases using machine learning

In our study, we aim to predict the antibody serostatus of patients with suspected autoimmune encephalitis (AE) using machine learning based on pre-contrast T2-weighted MR images acquired at symptom onset. A confirmation of seropositivity is of great importance for a reliable diagnosis in suspected AE cases. The cohort used in our study comprises 98 patients diagnosed with AE. 57 of these patients had previously tested positive for autoantibodies associated with AE. In contrast, no antibodies were detected in the remaining 41 patients. A manual bilateral segmentation of the hippocampus was performed using the open-source software 3D Slicer on T2-weighted MR-images. Subsequently, 107 Radiomics features were extracted from each T2-weighted MR image utilizing the open source PyRadiomics software package. Our study cohort was randomly divided into training and independent test data. Five conventional machine learning algorithms and a neural network were tested regarding their ability to differentiate between seropositive and seronegative patients. All performance values were determined based on independent test data. Our final model includes six features and is based on a Lasso regression. Using independent test data, this model yields a mean AUC of 0.950, a mean accuracy of 0.892, a mean sensitivity of 0.892 and a mean specificity of 0.891 in predicting antibody serostatus in patients with suspected AE. Our results show that Radiomics-based machine learning is a very promising method for predicting serostatus of suspected AE patients and can thus help to confirm the diagnosis. In the future, such methods could facilitate and accelerate the diagnosis of AE even before the results of specific laboratory tests are available, allowing patients to benefit more quickly from a reliable treatment strategy.

Prediction of Seropositivity in Suspected Autoimmune Encephalitis by Use of Radiomics: A Radiological Proof-of-Concept Study

In this study, we sought to evaluate the capabilities of radiomics and machine learning in predicting seropositivity in patients with suspected autoimmune encephalitis (AE) from MR images obtained at symptom onset. In 83 patients diagnosed with AE between 2011 and 2022, manual bilateral segmentation of the amygdala was performed on pre-contrast T2 images using 3D Slicer open-source software. Our sample of 83 patients contained 43 seropositive and 40 seronegative AE cases. Images were obtained at our tertiary care center and at various secondary care centers in North Rhine-Westphalia, Germany. The sample was randomly split into training data and independent test data. A total of 107 radiomic features were extracted from bilateral regions of interest (ROIs). Automated machine learning (AutoML) was used to identify the most promising machine learning algorithms. Feature selection was performed using recursive feature elimination (RFE) and based on the determination of the most important features. Selected features were used to train various machine learning algorithms on 100 different data partitions. Performance was subsequently evaluated on independent test data. Our radiomics approach was able to predict the presence of autoantibodies in the independent test samples with a mean AUC of 0.90, a mean accuracy of 0.83, a mean sensitivity of 0.84 and a mean specificity of 0.82, with Lasso regression models yielding the most promising results. These results indicate that radiomics-based machine learning could be a promising tool in predicting the presence of autoantibodies in suspected AE patients. Given the implications of seropositivity for definitive diagnosis of suspected AE cases, this may expedite diagnostic workup even before results from specialized laboratory testing can be obtained. Furthermore, in conjunction with recent publications, our results indicate that characterization of AE subtypes by use of radiomics may become possible in the future, potentially allowing physicians to tailor treatment in the spirit of personalized medicine even before laboratory workup is completed.

Multiple diffusion metrics in differentiating solid glioma from brain inflammation

Background and purpose

The differential diagnosis between solid glioma and brain inflammation is necessary but sometimes difficult. We assessed the effectiveness of multiple diffusion metrics of diffusion-weighted imaging (DWI) in differentiating solid glioma from brain inflammation and compared the diagnostic performance of different DWI models.

Materials and methods

Participants diagnosed with either glioma or brain inflammation with a solid lesion on MRI were enrolled in this prospective study from May 2016 to April 2023. Diffusion-weighted imaging was performed using a spin-echo echo-planar imaging sequence with five b values (500, 1,000, 1,500, 2000, and 2,500 s/mm2) in 30 directions for each b value, and one b value of 0 was included. The mean values of multiple diffusion metrics based on diffusion tensor imaging (DTI), diffusion kurtosis imaging (DKI), mean apparent propagator (MAP), and neurite orientation dispersion and density imaging (NODDI) in the abnormal signal area were calculated. Comparisons between glioma and inflammation were performed. The area under the curve (AUC) of the receiver operating characteristic curve (ROC) of diffusion metrics were calculated.

Results

57 patients (39 patients with glioma and 18 patients with inflammation) were finally included. MAP model, with its metric non-Gaussianity (NG), shows the greatest diagnostic performance (AUC = 0.879) for differentiation of inflammation and glioma with atypical MRI manifestation. The AUC of DKI model, with its metric mean kurtosis (MK) are comparable to NG (AUC = 0.855), followed by NODDI model with intracellular volume fraction (ICVF) (AUC = 0.825). The lowest value was obtained in DTI with mean diffusivity (MD) (AUC = 0.758).

Conclusion

Multiple diffusion metrics can be used in differentiation of inflammation and solid glioma. Non-Gaussianity (NG) from mean apparent propagator (MAP) model shows the greatest diagnostic performance for differentiation of inflammation and glioma.

MRI-based clinical-radiomics nomogram to predict early neurological deterioration in isolated acute pontine infarction: a two-center study in Northeast China

OBJECTIVE

To predict the appearance of early neurological deterioration (END) among patients with isolated acute pontine infarction (API) based on magnetic resonance imaging (MRI)-derived radiomics of the infarct site.

METHODS

544 patients with isolated API were recruited from two centers and divided into the training set (n = 344) and the verification set (n = 200). In total, 1702 radiomics characteristics were extracted from each patient. A support vector machine algorithm was used to construct a radiomics signature (rad-score). Subsequently, univariate and multivariate logistic regression (LR) analysis was adopted to filter clinical indicators and establish clinical models. Then, based on the LR algorithm, the rad-score and clinical indicators were integrated to construct the clinical-radiomics model, which was compared with other models.

RESULTS

A clinical-radiomics model was established, including the 5 indicators rad-score, age, initial systolic blood pressure, initial National Institute of Health Stroke Scale, and triglyceride. A nomogram was then made based on the model. The nomogram had good predictive accuracy, with an area under the curve (AUC) of 0.966 (95% confidence interval [CI] 0.947-0.985) and 0.920 (95% [CI] 0.873-0.967) in the training and verification sets, respectively. According to the decision curve analysis, the clinical-radiomics model showed better clinical value than the other models. In addition, the calibration curves also showed that the model has excellent consistency.

CONCLUSION

The clinical-radiomics model combined MRI-derived radiomics and clinical metrics and may serve as a scoring tool for early prediction of END among patients with isolated API.