Radiomic-Based MRI for Classification of Solitary Brain Metastases Subtypes From Primary Lymphoma of the Central Nervous System.

BACKGROUND: Differential diagnosis of brain metastases subtype and primary central nervous system lymphoma (PCNSL) is necessary for treatment decisions. The application of machine learning facilitates the classification of brain tumors, but prior investigations into primary lymphoma and brain metastases subtype classification have been limited. PURPOSE: To develop a machine-learning model to classify PCNSL, brain metastases with primary lung and non-lung origin. STUDY TYPE: Retrospective. POPULATION: A total of 211 subjects with pathologically confirmed PCNSL or brain metastases (training cohort 168 and testing cohort 43). FIELD STRENGTH/SEQUENCE: A 3.0 T axial contrast-enhanced T1-weighted spin-echo inversion recovery sequence (T1WI-CE), axial T2-weighted fluid-attenuation inversion recovery sequence (T2FLAIR) ASSESSMENT: Several machine-learning models (support vector machine, random forest, and K-nearest neighbors) were built with least absolute shrinkage and selection operator (LASSO) using features from T1WI-CE, T2FLAIR, and clinical. The model with the highest performance in the training cohort was selected to differentiate lesions in the testing cohort. Then, three radiologists conducted a two-round classification (with and without model reference) using images and clinical information from testing cohorts. STATISTICAL TESTS: Five-fold cross-validation was used for model evaluation and calibration. Model performance was assessed based on sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUC). RESULTS: Twenty-five image features were selected by LASSO analysis. Random forest classifier was selected for its highest performance on the training set with an AUC of 0.73. After calibration, this model achieved an accuracy of 0.70 on the testing set. Accuracies of all three radiologists improved under model reference (0.49 vs. 0.70, 0.60 vs. 0.77, 0.58 vs. 0.72, respectively). DATA CONCLUSION: The random forest model based on conventional MRI and clinical data can diagnose PCNSL and brain metastases subtypes (lung and non-lung origin). Model classification can help foster the diagnostic accuracy of specialists and streamline prognostication workflow. EVIDENCE LEVEL: 4 TECHNICAL EFFICACY: Stage 2.

Association of clinical and imaging characteristics with pulmonary function testing in patients with Long-COVID.

Purpose: The purpose of this study is to identify clinical and imaging characteristics associated with post-COVID pulmonary function decline. Methods: This study included 22 patients recovering from COVID-19 who underwent serial spirometry pulmonary function testing (PFT) before and after diagnosis. Patients were divided into two cohorts by difference between baseline and post-COVID follow-up PFT: Decline group (>10 % decrease in FEV1), and Stable group (≤10 % decrease or improvement in FEV1). Demographic, clinical, and laboratory data were collected, as well as PFT and chest computed tomography (CT) at the time of COVID diagnosis and follow-up. CTs were semi-quantitatively scored on a five-point severity scale for disease extent in each lobe by two radiologists. Mann-Whitney U-tests, T-tests, and Chi-Squared tests were used for comparison. P-values <0.05 were considered statistically significant. Results: The Decline group had a higher proportion of neutrophils (79.47 ± 4.83 % vs. 65.45 ± 10.22 %; p = 0.003), a higher absolute neutrophil count (5.73 ± 2.68 × 109/L vs. 3.43 ± 1.74 × 109/L; p = 0.031), and a lower proportion of lymphocytes (9.90 ± 4.20 % vs. 21.21 ± 10.97 %; p = 0.018) compared to the Stable group. The Decline group also had significantly higher involvement of ground-glass opacities (GGO) on follow-up chest CT [8.50 (4.50, 14.50) vs. 3.0 (1.50, 9.50); p = 0.032] and significantly higher extent of reticulations on chest CT at time of COVID diagnosis [6.50 (4.00, 9.00) vs. 2.00 (0.00, 6.00); p = 0.039] and follow-up [5.00 (3.00, 13.00) vs. 2.00 (0.00, 5.00); p = 0.041]. ICU admission was higher in the Decline group than in the Stable group (71.4 % vs. 13.3 %; p = 0.014). Conclusions: This study provides novel insight into factors influencing post-COVID lung function, irrespective of pre-existing pulmonary conditions. Our findings underscore the significance of neutrophil counts, reduced lymphocyte counts, pulmonary reticulation on chest CT at diagnosis, and extent of GGOs on follow-up chest CT as potential indicators of decreased post-COVID lung function. This knowledge may guide prediction and further understanding of long-term sequelae of COVID-19 infection.

MRI-Based Prediction of Clinical Improvement Following Ventricular Shunt Placement for Normal Pressure Hydrocephalus (NPH): Development and Evaluation of an Integrated Multi-Sequence Machine Learning Algorithm.

BACKGROUND AND PURPOSE: Symptoms of normal pressure hydrocephalus (NPH) are sometimes refractory to shunt placement, with limited ability to predict improvement for individual patients. We evaluated an MRI-based artificial intelligence method to predict post-shunt NPH symptom improvement. MATERIALS AND METHODS: NPH patients who underwent magnetic resonance imaging (MRI) prior to shunt placement at a single center (2014-2021) were identified. Twelve-month post-shunt improvement in modified Rankin Scale (mRS), incontinence, gait, and cognition were retrospectively abstracted from clinical documentation. 3D deep residual neural networks were built on skull stripped T2-weighted and fluid attenuated inversion recovery (FLAIR) images. Predictions based on both sequences were fused by additional network layers. Patients from 2014-2019 were used for parameter optimization, while those from 2020-2021 were used for testing. Models were validated on an external validation dataset from a second institution (n=33). RESULTS: Of 249 patients, n=201 and n=185 were included in the T2-based and FLAIR-based models according to imaging availability. The combination of T2-weighted and FLAIR sequences offered the best performance in mRS and gait improvement predictions relative to models trained on imaging acquired using only one sequence, with AUROC values of 0.7395 [0.5765-0.9024] for mRS and 0.8816 [0.8030-0.9602] for gait. For urinary incontinence and cognition, combined model performances on predicting outcomes were similar to FLAIR-only performance, with AUROC values of 0.7874 [0.6845-0.8903] and 0.7230 [0.5600-0.8859]. CONCLUSIONS: Application of a combined algorithm using both T2-weighted and FLAIR sequences offered the best image-based prediction of post-shunt symptom improvement, particularly for gait and overall function in terms of mRS. ABBREVIATIONS: NPH = normal pressure hydrocephalus; iNPH = idiopathic NPH; sNPH = secondary NPH; AI = artificial intelligence; ML = machine learning; CSF = cerebrospinal fluid; AUROC = area under the receiver operating characteristic; FLAIR = fluid attenuated inversion recovery; BMI = body mass index; CCI = Charlson Comorbidity Index; SD = standard deviation; IQR = interquartile range.

Head and neck tumor segmentation convolutional neural network robust to missing PET/CT modalities using channel dropout.

Objective. Radiation therapy for head and neck (H&N) cancer relies on accurate segmentation of the primary tumor. A robust, accurate, and automated gross tumor volume segmentation method is warranted for H&N cancer therapeutic management. The purpose of this study is to develop a novel deep learning segmentation model for H&N cancer based on independent and combined CT and FDG-PET modalities.Approach. In this study, we developed a robust deep learning-based model leveraging information from both CT and PET. We implemented a 3D U-Net architecture with 5 levels of encoding and decoding, computing model loss through deep supervision. We used a channel dropout technique to emulate different combinations of input modalities. This technique prevents potential performance issues when only one modality is available, increasing model robustness. We implemented ensemble modeling by combining two types of convolutions with differing receptive fields, conventional and dilated, to improve capture of both fine details and global information.Main Results. Our proposed methods yielded promising results, with a Dice similarity coefficient (DSC) of 0.802 when deployed on combined CT and PET, DSC of 0.610 when deployed on CT, and DSC of 0.750 when deployed on PET.Significance. Application of a channel dropout method allowed for a single model to achieve high performance when deployed on either single modality images (CT or PET) or combined modality images (CT and PET). The presented segmentation techniques are clinically relevant to applications where images from a certain modality might not always be available.

Changed cerebral function and morphology serve as neuroimaging evidence for subclinical type 2 diabetic polyneuropathy.

Introduction: Central and peripheral nervous systems are all involved in type 2 diabetic polyneuropathy mechanisms, but such subclinical changes and associations remain unknown. This study aims to explore subclinical changes of the central and peripheral and unveil their association. Methods: A total of 55 type-2 diabetes patients consisting of symptomatic (n = 23), subclinical (n = 12), and no polyneuropathy (n = 20) were enrolled in this study. Cerebral morphology, function, peripheral electrophysiology, and clinical information were collected and assessed using ANOVA and post-hoc analysis. Gaussian random field correction was used for multiple comparison corrections. Pearson/Spearman correlation analysis was used to evaluate the association of the cerebral with the peripheral. Results: When comparing the subclinical group with no polyneuropathy groups, no statistical differences were shown in peripheral evaluations except amplitudes of tibial nerves. At the same time, functional connectivity from the orbitofrontal to bilateral postcentral and middle temporal cortex increased significantly. Gray matter volume of orbitofrontal and its functional connectivity show a transient elevation in the subclinical group compared with the symptomatic group. Besides, gray matter volume in the orbitofrontal cortex negatively correlated with the Neuropathy Symptom Score (r = -0.5871, p < 0.001), Neuropathy Disability Score (r = -0.3682, p = 0.009), and Douleur Neuropathique en 4 questions (r = -0.4403, p = 0.003), and also found correlated positively with bilateral peroneal amplitude (r > 0.4, p < 0.05) and conduction velocities of the right sensory sural nerve(r = 0.3181, p = 0.03). Similarly, functional connectivity from the orbitofrontal to the postcentral cortex was positively associated with cold detection threshold (r = 0.3842, p = 0.03) and negatively associated with Neuropathy Symptom Score (r = -0.3460, p = 0.01). Discussion: Function and morphology of brain changes in subclinical type 2 diabetic polyneuropathy might serve as an earlier biomarker. Novel insights from subclinical stage to investigate the mechanism of type 2 diabetic polyneuropathy are warranted.

Functional Connectivity Density for Radiation Encephalopathy Prediction in Nasopharyngeal Carcinoma.

The diagnostic efficiency of radiation encephalopathy (RE) remains heterogeneous, and prediction of RE is difficult at the pre-symptomatic stage. We aimed to analyze the whole-brain resting-state functional connectivity density (FCD) of individuals with pre-symptomatic RE using multivariate pattern analysis (MVPA) and explore its prediction efficiency. Resting data from NPC patients with nasopharyngeal carcinoma (NPC; consisting of 20 pre-symptomatic RE subjects and 26 non-RE controls) were collected in this study. We used MVPA to classify pre-symptomatic RE subjects from non-RE controls based on FCD maps. Classifier performances were evaluated by accuracy, sensitivity, specificity, and area under the characteristic operator curve. Permutation tests and leave-one-out cross-validation were applied for assessing classifier performance. MVPA was able to differentiate pre-symptomatic RE subjects from non-RE controls using global FCD as a feature, with a total accuracy of 89.13%. The temporal lobe as well as regions involved in the visual processing system, the somatosensory system, and the default mode network (DMN) revealed robust discrimination during classification. Our findings suggest a good classification efficiency of global FCD for the individual prediction of RE at a pre-symptomatic stage. Moreover, the discriminating regions may contribute to the underlying mechanisms of sensory and cognitive disturbances in RE.