Musculoskeletal tumors (MSTs) account for about 1 % of malignant tumors in adults and 20 % of tumors in children [1]. Due to the rarity and diversity of its radiological features, the early and accurate diagnosis of MSTs poses a significant challenge to radiologists [2]. Some studies have confirmed that the prognosis of MSTs with higher grade and smaller volume is better, and its early and accurate diagnosis is particularly urgent [3]. Before treatment, patients typically undergo preoperative biopsy and radiologic evaluation, preoperative biopsy is generally considered to be a more reliable diagnosis but is invasive and carries an increased risk of tumor-cell contamination of the surrounding tissue [4,5]. Imaging diagnosis plays a crucial role in determining the presence, location, size, and nature of the tumor [6]. Magnetic resonance imaging (MRI) is widely used for the characterization and staging of MSTs [7]. Although the final diagnosis still depends on pathological examination, functional MRI imaging that provides information about tumor biology has become increasingly attractive with advances in MR Technology.

Diffusion weighted imaging (DWI) is a non-invasive MRI technique, which has been proved to improve the diagnostic accuracy of qualitative detection of musculoskeletal tumors in a number of studies [8,9]. By applying different diffusion weights (q-space) to probe the diffusion behavior of confined water molecules in biological tissues [10], DWI provides basic information about tumor components, which can be quantified and help to distinguish between benign and malignant MSTs using Apparent diffusion coefficient (ADC). However, the utility of ADC analysis of DWI is limited, and the range of ADC values for benign and malignant MSTs overlaps or even does not differ significantly [11]. This is because ADC reflects only the overall water diffusion rate and cannot provide detailed information on the pathological structure of the tumor, such as specific microstructures [12].

In recent years, MR cytometry is an emerging technique that characterizes cellular microstructure by analyzing diffusion signals at multiple diffusion times (t-space), providing unique insights at the cellular level [13]. These properties cannot be measured with Conventional DWI. Specifically, the previously developed IMPULSED (Imaging Microstructural Parameters Using Limited Spectrally Edited Diffusion) method uses a joint acquisition strategy of pulsed gradient spin-echo (PGSE) and oscillating gradient spin-echo (OGSE) sequences, which integrates q-space and t-space to build an analytical biophysical model [14]. Based on this model, DWI signals can be expressed as analytical expressions related to microstructure features such as cell density. The microstructure parameters can be resolved by fitting the acquired signal into the expression [15]. This technique can be implemented on a clinical scanner, and its feasibility has been demonstrated in animal experiments by identifying cellular microstructure [16]. The preliminary study combined OGSE and PGSE sequences to evaluate changes in ADC for distinguishing benign from malignant head and neck tumors in vivo [17]. In recent years, some studies have further obtained applied the IMPULSED method to construct biophysical models for obtaining microstructural parameters and combined these parameters with histopathological analysis to classify pediatric glioma [18], distinguish clinically significant and non-clinically significant prostate cancer [12], predict molecular subtypes of breast cancer, and evaluate response to neoadjuvant chemotherapy [13].

We hypothesized that the MR cytometry method could effectively characterize the microstructure properties of MSTs, and then predict its benign and malignant. This study aims to investigate the diagnostic efficacy of quantitative parameters of MR cytometry in the differential diagnosis of benign and malignant musculoskeletal tumors, and to evaluate whether it is superior to Conventional DWI.