A glioma is the most common primary nervous system tumor in adults, accounting for approximately 80% of all malignant primary central nervous system tumors [1], with a poor prognosis and common recurrence after surgery. At present, the standardized treatment scheme is resection in the maximum safe range, combined with concurrent chemoradiotherapy (CCRT) after surgery, and the most commonly used chemotherapy drug is temozolomide [2]. One study found that the use of CCRT significantly increased the incidence of pseudoprogression (psPD) [3]. Due to the spatial heterogeneity of the glioma itself, some patients will suffer recurrence after an operation, and the disease will progress. In the follow-up imaging, such a recurrence manifests as new or increasing abnormal enhancement. Although this is similar to the manifestations of psPD, the prognosis and related clinical strategies of the two are very different. For recurrence, reoperation, electric field treatment, radiotherapy, or chemotherapy may be needed, whereas for psPD, only close follow-up and symptomatic treatment are needed. The differentiation between the two depends on continuous imaging follow-up, but it can take at least four to six months to obtain a reliable differential diagnosis, which can result in a delay in treatment or the implementation of treatment that is inappropriate [4,5]. Clearly, the timely and noninvasive identification of the nature of any abnormal enhancement after glioma surgery is of great guiding significance for clinical treatment.

Although the imaging manifestations of psPD and recurrence are similar, their internal pathological mechanisms are different. The former, which is an enhanced response after radiotherapy is a brain injury that leads to the transient destruction of the blood–brain barrier and an increase in vascular permeability [6]. On the other hand, recurrence is related to the spatial heterogeneity, invasiveness, and vascular proliferation of the tumor itself; and the high expression of protein Ki-67 is related to the invasiveness of the tumor itself and the proliferation of cells and blood vessels [7]. When the tumor involves peritumoral edema, or it is not resected or has a high expression of vascular endothelial growth factor, it can easily recur.

At present, a variety of functional imaging sequences, including diffusion-weighted imaging (DWI), perfusion-weighted imaging (PWI), and magnetic resonance spectroscopy, have great potential for the differentiation between glioma recurrence and psPD [8,9]. Using these imaging methods to observe the molecular diffusion and blood perfusion of suspected abnormal enhancement foci intuitively makes it possible to provide a more objective evaluation for distinguishing recurrence from psPD. However, most of these imaging techniques are based on two-dimensional regions of interest (ROI), and due to the spatial heterogeneity of the tumor itself, it can be difficult to comprehensively evaluate the new abnormal enhancement foci (Fig. 1). In addition, imaging examinations are lengthy and the requirements for equipment performance are high, which can cause conflict with daily clinical needs. Consequently, the use of these techniques is not very popular.

Radiomics is an technology that utilizes computer graphics processing to convert image data into high-dimensional graphic features, such as shape, size, and texture [10,11], and it can be used to evaluate the tumor microenvironment that cannot be seen by the naked eye. In recent years, radiomics technology has made a breakthrough, in that clinical features and molecular markers can now be used as additional input features, in combination with omics features, to construct prediction models. A number of studies have revealed that radiological imaging combined with omics has high predictive ability in tumor diagnosis, grading, phenotypic prediction, and efficacy evaluation [12,13]. At present, the evaluation of an abnormal enhancement focus after glioma surgery mostly concerns T1 enhancement sequencing, which evaluates the enhancement focus itself, but pays little attention to the surrounding area. The internal pathological changes of tumor recurrence and psPD will also cause the morphological changes of abnormal enhancement focus on T1-weighted imaging (T1WI) and T2WI/fluid attenuated inversion recovery (FLAIR) hyperintensive signals around the focus, and the two-way measurement obtained using Response Assessment in Neuro-Oncology (RANO) guidelines may impede an early identification.

This study aimed to evaluate the ability of a multiparametric omics model based on conventional magnetic resonance imaging (MRI) sequencing to distinguish glioma recurrence from psPD after standard treatment, so as to improve the accuracy of diagnosis and provide better guidance for clinical decision-making.