Clinical data and MRI features-based nomogram for differentiation of central nervous system infection and central nervous system involvement in hematological malignancy.

Abstract

Central nervous system leukemia (CNSL) and central nervous system infection (CNSI) are the most important complications in patients with acute leukemia (AL). However, the differential diagnosis could represent a major challenge since the two disorders are all heterogeneous entities with overlapping clinical characteristics and radiological appearances. In this paper, we conduct a retrospective study to develop a model based on clinical data and magnetic resonance imaging (MRI) to distinguish CNSL from CNSI. A total of 108 patients with AL who underwent cranial MRI between January 2020 and December 2023 in our hospital were included. Univariate and multivariate logistic regression analyses were used to determine the independent predictors. A nomogram was developed based on the predictors, and the performance of the nomogram was evaluated by the area under the receiver operating characteristic (ROC) curve. The validation cohort was used to test the predictive model. Hyperleukocytosis at initial diagnosis, marrow state, fever, conscious disturbance, coinfection in other sites and MRI (parenchyma type) were identified as independent factors. A nomogram was constructed and the discrimination was presented as AUC = 0.947 (95% CI 0.9105-0.984). Calibration of the nomogram showed that the predicted probability matched the actual probability well.