Background: The radiological differential diagnosis between tumor recurrence and radiation-induced necrosis (ie, pseudoprogression) is of paramount importance in the management of glioma patients.Objective: This research aims to develop a deep learning methodology for automated differentiation of tumor recurrence from radiation necrosis based on routine magnetic resonance imaging (MRI) scans.Methods: In this retrospective study, 146 patients who underwent radiation therapy after glioma resection and presented with suspected recurrent lesions at the follow-up MRI examination were selected for analysis. Routine MRI scans were acquired from each patient, including T1, T2, and gadolinium-contrast-enhanced T1 sequences. Of those cases, 96 (65.8%) were confirmed as glioma recurrence on postsurgical pathological examination, while 50 (34.2%) were diagnosed as necrosis. A light-weighted deep neural network (DNN) (ie, efficient radionecrosis neural network [ERN-Net]) was proposed to learn radiological features of gliomas and necrosis from MRI scans. Sensitivity, specificity, accuracy, and area under the curve (AUC) were used to evaluate performance of the model in both image-wise and subject-wise classifications. Preoperative diagnostic performance of the model was also compared to that of the state-of-the-art DNN models and five experienced neurosurgeons.Results: DNN models based on multimodal MRI outperformed single-modal models. ERN-Net achieved the highest AUC in both image-wise (0.915) and subject-wise (0.958) classification tasks. The evaluated DNN models achieved an average sensitivity of 0.947 (SD 0.033), specificity of 0.817 (SD 0.075), and accuracy of 0.903 (SD 0.026), which were significantly better than the tested neurosurgeons (P=.02 in sensitivity and P<.001 in specificity and accuracy).Conclusions: Deep learning offers a useful computational tool for the differential diagnosis between recurrent gliomas and necrosis. The proposed ERN-Net model, a simple and effective DNN model, achieved excellent performance on routine MRI scans and showed a high clinical applicability.

Objective. This study aimed to develop an automatic and accurate method for severity assessment and localization of coronary artery disease (CAD) based on an optically pumped magnetometer magnetocardiography (MCG) system. Approach. We proposed spatiotemporal features based on the MCG one-dimensional signals, including amplitude, correlation, local binary pattern, and shape features. To estimate the severity of CAD, we classified the stenosis as absence or mild, moderate, or severe cases and extracted a subset of features suitable for assessment. To localize CAD, we classified CAD groups according to the location of the stenosis, including the left anterior descending artery (LAD), left circumflex artery (LCX), and right coronary artery (RCA), and separately extracted a subset of features suitable for determining the three CAD locations. Main results. For CAD severity assessment, a support vector machine (SVM) achieved the best result, with an accuracy of 75.1%, precision of 73.9%, sensitivity of 67.0%, specificity of 88.8%, F1-score of 69.8%, and area under the curve of 0.876. The highest accuracy and corresponding model for determining locations LAD, LCX, and RCA were 94.3% for the SVM, 84.4% for a discriminant analysis model, and 84.9% for the discriminant analysis model. Significance . The developed method enables the implementation of an automated system for severity assessment and localization of CAD. The amplitude and correlation features were key factors for severity assessment and localization. The proposed machine learning method can provide clinicians with an automatic and accurate diagnostic tool for interpreting MCG data related to CAD, possibly promoting clinical acceptance.

The advent of optically pumped magnetometers (OPMs) facilitates the development of on-scalp magnetoencephalography (MEG). In particular, the triaxial OPM emerged recently, making simultaneous measurements of all three orthogonal components of vector fields possible. The detection of triaxial magnetic fields improves the interference suppression capability and achieves higher source localization accuracy using fewer sensors. The source localization accuracy of MEG is based on the accurate co-registration of MEG and MRI. In this study, we proposed a triaxial co-registration method according to combined principal component analysis and iterative closest point algorithms for use of a flexible cap. A reference phantom with known sensor positions and orientations was designed and constructed to evaluate the accuracy of the proposed method. Experiments showed that the average co-registered position errors of all sensors were approximately 1 mm and average orientation errors were less than 2.5° in the X - and Y orientations and less than 1.6° in the Z orientation. Furthermore, we assessed the influence of co-registration errors on the source localization using simulations. The average source localization error of approximately 1 mm reflects the effectiveness of the co-registration method. The proposed co-registration method facilitates future applications of triaxial sensors on flexible caps.

Magnetoencephalography (MEG) based on optically pumped magnetometers (OPM-MEG) has shown better flexibility in sensor configuration compared with the conventional superconducting quantum interference devices-based MEG system while being better suited for all-age groups. However, this flexibility presents challenges for the co-registration of MEG and magnetic resonance imaging (MRI), hindering adoption. This study presents a toolbox called OMMR, developed in Matlab, that facilitates the co-registration step for researchers and clinicians. OMMR integrates the co-registration methods of using the electromagnetic digitization system and two types of optical scanners (the structural-light and laser scanner). As the first open-source co-registration toolbox specifically for OPM-MEG, the toolbox aims to standardize the co-registration process and set the ground for future applications of OPM-MEG.

Objective: Early detection and prediction of Alzheimer's Disease are paramount for elucidating neurodegenerative processes and enhancing cognitive resilience. Structural Magnetic Resonance Imaging (sMRI) provides insights into brain morphology, while resting-state Magnetoencephalography (rsMEG) elucidates functional aspects. However, inherent disparities between these multimodal neuroimaging modalities pose challenges to the effective integration of multimodal features.Approach: To address these challenges, we propose a deep learning-based multimodal classification framework for Alzheimer's disease, which harnesses the fusion of pivotal features from sMRI and rsMEG to augment classification precision. Utilizing the BioFIND dataset, classification trials were conducted on 163 Mild Cognitive Impairment cases and 144 cognitively Healthy Controls.Results: The study findings demonstrate that the InterFusion method, combining sMRI and rsMEG data, achieved a classification accuracy of 0.827. This accuracy significantly surpassed the accuracies obtained by rsMEG only at 0.710 and sMRI only at 0.749. Moreover, the evaluation of different fusion techniques revealed that InterFusion outperformed both EarlyFusion with an accuracy of 0.756 and LateFusion with an accuracy of 0.801. Additionally, the study delved deeper into the role of different frequency band features of rsMEG in fusion by analyzing six frequency bands, thus expanding the diagnostic scope.Discussion: These results highlight the value of integrating resting-state rsMEG and sMRI data in the early diagnosis of Alzheimer's disease, demonstrating significant potential in the field of neuroscience diagnostics.