Time-dependent diffusion MRI of soft tissue tumors: correlations with Ki-67 proliferation status

Introduction

Conventional magnetic resonance imaging (MRI) is a vital imaging tool for the evaluation, local staging, and treatment planning of soft tissue masses in the extremities (1). It is the preferred method for diagnosing soft tissue tumors in clinical practice (2). Previous studies have shown that various MRI techniques assist in the diagnosis and differentiation of soft tissue tumors, as well as in assessing their potential for malignancy (3,4). However, soft tissue tumors are highly heterogeneous and may not be accurately visualized by conventional MRI. Therefore, there is an urgent need for a reliable, non-invasive method to preoperatively evaluate the nature of soft tissue tumors.

Diffusion-weighted imaging (DWI) is a noninvasive MRI technique, and previous studies have demonstrated a significant correlation between the apparent diffusion coefficient (ADC) and the proliferative capacity of soft tissue tumors (5). DWI probes the diffusion of water molecules in biological tissues by applying different diffusion weightings, thereby providing and quantifying fundamental information about tumor components, expressed through ADC values (6). However, DWI cannot provide detailed information about the pathological microstructure of tumors. ADC reflects only an overall measure of water diffusivity, which is determined by multiple microstructural features such as intracellular and extracellular spaces, cell size, permeability, and intrinsic diffusivity (7).

Recently, time-dependent diffusion MRI (td-dMRI), a novel MRI technique, has enhanced the role of MRI in assessing tumor malignancy. This method offers unique advantages by integrating different diffusion weightings and multiple diffusion times within a unified acquisition, allowing the construction of biophysical models for signal analysis (8). Based on these models, the DWI signal can be expressed analytically in relation to microstructural features such as cellularity. By fitting the acquired signals to these analytical expressions, microstructural parameters can be quantitatively derived (7). A study on animals has demonstrated the feasibility of identifying cellular microstructures (9). Recent studies have preliminarily used td-dMRI in vivo to differentiate between malignant and benign head and neck tumors (10), distinguish clinically significant from insignificant prostate cancers (7), histologically classify gliomas (11), and predict breast cancer molecular subtypes and pathological complete response following neoadjuvant chemotherapy (12).

Soft tissue tumors comprise various histopathological subtypes, such as cellular-type, lipomatous-type, and chondroid/myxoid tumors (13). In lipomatous and myxoid tumors, the presence of fatty or myxoid components makes it difficult to exclude these elements during region of interest (ROI) delineation, thereby limiting the accurate identification of purely solid tumor regions (14). In contrast, cellular-type soft tissue tumors are primarily composed of solid tumor cells with relatively homogeneous tissue components. As a result, analyses based on the whole-tumor ROI are less affected by interfering tissues, making imaging evaluation in this subgroup generally more reliable.

At present, the widely used grading systems of soft tissue sarcoma include the National Cancer Institute (NCI) system and the French Federation of Cancer Centers Sarcoma Group (FNCLCC) system (15). However, the current grading systems have certain limitations in evaluating specific types of soft tissue sarcoma and in predicting the prognosis for soft tissue sarcoma (16). The immunohistochemical index Ki-67 is an antigen encoded by a single gene located on chromosome 10. It is expressed during the proliferative phases of the cell cycle and is markedly associated with high mitotic activity. Consequently, it has become one of the most reliable indicators for monitoring cellular proliferative activity (17,18). Therefore, Ki-67 can serve as a marker of tumor proliferation and reflect the extent of tumor cell growth (19). In several studies of soft tissue sarcoma, the Ki-67 proliferation status has enabled the differentiation of low-grade from high-grade sarcomas and the prediction of responses to neoadjuvant radiotherapy (20-22). A Ki-67-labeling index cut-off value of 20% has been proposed as a biomarker for predicting recurrence, metastasis, and patient prognosis (5,23). However, no td-dMRI studies have explored the predictive value of the Ki-67 index in relation to tumor imaging features.

We hypothesized that td-dMRI methods could help characterize microstructural properties that might predict the malignant proliferation capacity of soft tissue tumors. The objective of this study was to assess the correlation between quantitative parameters obtained from td-dMRI and the Ki-67 index, and to ascertain if these parameters could serve as noninvasive predictors of Ki-67 status in soft tissue tumors. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1815/rc).

Methods

Participants

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Research Ethics Committee of the Third Hospital of Hebei Medical University (No. Z2022-059-1). Written informed consent was provided by all participants to undergo td-dMRI in addition to the standard-of-care multiparametric MRI. Between December 2023 and August 2024, 47 consecutive participants with clinically suspected soft tissue tumors were prospectively enrolled at the Third Hospital of Hebei Medical University. The inclusion criteria were as follows: (I) postoperative histopathological verification of soft tissue tumors; (II) completion of a standard td-dMRI examination; and (III) positive immunohistochemical staining for Ki-67 in pathological specimens. The exclusion criteria were as follows: (I) image quality was compromised by severe artifacts; (II) small lesion size (less than 10 mm in short-axis diameters); and (III) a history of treatment or recurrence prior to the MRI examination. In total, 31 participants fulfilled the eligibility criteria, comprising 18 males and 13 females, with a mean age of 52 years (range, 10–76 years). Figure 1 shows a flowchart of participant enrollment. Participants were categorized into two groups according to a Ki-67 cut-off value of 20%. In this study, soft tissue tumors were primarily classified into three types: cellular type, lipomatous, and myxoid tumors.

Figure 1 Flowchart illustrating selection of the study population. MRI, magnetic resonance imaging.

Image acquisition

All images were obtained on a 3T MRI scanner (MAGNETOM Vida; Siemens Healthineers, Forchheim, Germany). td-dMRI integrates results from pulsed gradient spin echo (PGSE) and oscillating gradient spin echo (OGSE) encoding through a specialized research sequence. Specifically, OGSE sequences with oscillation frequencies of 25 Hz (effective diffusion time: 7.1 ms; b values: 0, 400, 800 s/mm²) and 40 Hz (effective diffusion time: 4.6 ms; b values: 0, 200, 460 s/mm²) employing a cosine-trapezoidal modulation, were used along with a conventional PGSE (0 Hz) sequence (effective diffusion time: 70 ms; b values: 0, 400, 800 s/mm²) to capture diffusion signals at varying diffusion times. Other acquisition parameters were kept consistent for both sequences: three orthogonal diffusion directions; repetition time/echo time =5,500 ms/123 ms; field-of-view = 220×220 mm²; slice number =14; and slice thickness =4 mm. The total scan time for the td-dMRI protocol was approximately 7.5 minutes. Conventional MRI examinations were typically performed before td-dMRI, with contrast-enhanced sequences applied as required. Both conventional and enhanced sequences played a key role in delineating the ROIs.

Image analysis

To obtain microstructural parameters, a two-compartment model known as Imaging Microstructural Parameters Using Limited Spectrally Edited Diffusion (IMPULSED) (9) was used. The microstructural parameters derived from this model included cell diameter (d), intracellular volume fraction (v_in, the relative proportion of water molecules within the intracellular space), cellularity, and extracellular diffusivity (D_ex, the effective diffusivity of water molecules within the extracellular space). A radiologist with a decade of experience, unaware of the participant data, manually delineated the ROIs within soft tissue tumors on each slice using diffusion-weighted images and excluded border voxels to avoid partial volume effects. Figure 2A illustrates the delineation of the ROI on a case image. For each tumor ROI, fitted microstructural parameters (cellularity, cell diameter, intracellular volume fraction and extracellular diffusivity), ADC values at various diffusion times [ADC at 0 Hz, 25 Hz, and 40 Hz (ADC_{0 Hz}, ADC_{25 Hz}, ADC_{40 Hz})], the change in ADC (cADC), and the relative change in ADC (rcADC) were calculated and averaged. Figure 2 illustrates a representative td-dMRI map of a soft tissue tumor that has been histopathologically verified.

Figure 2 Histopathologically confirmed aggressive fibromatosis involving the left buttock of an 18-year-old woman (case No. 25). (A) Each parametric map displays the lesion, and white contours denote the tumor areas manually outlined. (B) Hematoxylin and eosin staining confirmed the mass as an aggressive fibromatosis (×100). (C) Immunohistochemical staining for Ki-67 showed approximately 5% nuclear positivity (×200). ADC, apparent diffusion coefficient; ADC_{0 Hz}, ADC at 0 Hz; ADC_{25 Hz}, ADC at 25 Hz; ADC_{40 Hz}, ADC at 40 Hz; d, cell diameter; D_ex, extracellular diffusivity; IMPULSED, Imaging Microstructural Parameters Using Limited Spectrally Edited Diffusion; T1w, T1-weighted; T2w, T2-weighted; v_in, intracellular volume fraction.

ADC values were computed using Eq. [1], in which S0 and S1 denote the signal intensities obtained from DWI a low (b0) and high (b1) b values, respectively. The time dependence of ADC diffusion was evaluated by calculating changes in ADC between the OGSE and the PGSE sequences, then determining their ratio relative to the ADC of the PGSE sequence. The corresponding cADC and rcADC were computed according to Eqs. [2,3] presented below (24).

ADC=ln(S0/S1)/(b1−b0)[1]

ADC=ADC40Hz−ADC0Hz[2]

rcADC=(ADC40Hz−ADC0Hz)/ADC0Hz×100%[3]

Immunohistochemical analysis

Biopsies were conducted by a board-certified musculoskeletal radiologist. Multiple specimens (at least three) were obtained using core needle biopsies and an automatic biopsy gun under image guidance, depending on the tumor’s location and characteristics. All pathological sections were independently reviewed by two senior pathologists, each with more than a decade of professional experience, without access to the MRI findings. The formalin-fixed, paraffin-embedded tissues were stained with hematoxylin and eosin (H&E) for histological examination and were also subjected to immunohistochemical analysis using antibodies against Ki-67. Brown nuclear staining was considered indicative of Ki-67 positivity. Ki-67-positive tumor cells were identified through the enumeration of at least 1,000 tumor cells within over five high-power fields characterized by their uniform distribution. The Ki-67 index was recorded as the proportion of Ki-67-positive tumor cells. The participants were divided into high-proliferation (Ki67 >20%) and low-proliferation (Ki67 ≤20%) groups.

Statistical analysis

The Shapiro-Wilk test was applied to assess whether the data were normally distributed. Quantitative parameters derived from td-dMRI were expressed as mean ± standard deviation. To compare normally distributed data (d, D_ex, ADC_{40 Hz}, and ADC_{0 Hz}), a t-test was conducted. For comparisons of non-normally distributed data (Ki-67, cellularity, v_in, ADC_{25 Hz}, cADC, and rcADC), the Mann-Whitney U test was used. The correlations between the Ki-67 index and td-dMRI parameters were assessed within each group. The correlations between the Ki-67 index and td-dMRI parameters were assessed within each group. Pearson correlation analysis was used for normally distributed data, whereas Spearman correlation analysis was utilized for non-normally distributed data, to investigate the relationships of the Ki-67 index with key td-dMRI parameters.

Receiver operating characteristic (ROC) curves were used to assess the efficacy of td-dMRI parameters in differentiation between low and high Ki-67 statuses. The area under the curve (AUC), sensitivity, specificity, and Youden’s index were calculated. AUC values were categorized as follows: low (0.5< AUC ≤0.7), moderate to good (0.7< AUC ≤0.9), and very good to excellent (0.9< AUC ≤1). The correlation strength was classified as very strong (0.8< |r| ≤1.0), strong (0.6< |r| ≤0.8), medium (0.4< |r| ≤0.6), low (0.2< |r| ≤0.4), or very low to negligible (0.0< |r| ≤0.2) (5). The parameter exhibiting the highest AUC and correlation coefficient was chosen as the most effective metric for discriminating participants into low and high Ki-67 values. Statistical analyses were performed using SPSS 27.0 (IBM Corp., Armonk, NY, USA) and MedCalc 22.1.0 (MedCalc Software, Ostend, Belgium). A P value <0.05 was considered statistically significant.

Results

Participant characteristics

In total, 31 participants [mean age, 52.1±19.9 years, with an interquartile range (IQR) of 44.8–59.4 years] with soft tissue tumors were recruited for this study. The median interval between MRI examination and biopsy was 4 days (range, 3–7 days). Among all the participants, there were 13 females and 18 males, with 6 participants having tumors located in the trunk and 25 participants having tumors located in the limbs. No significant differences were observed in age, gender, or the location of soft tissue tumors between the two groups (all P>0.05, Table 1). Histologic examination classified all participants’ tumors into 11 benign and 20 malignant soft-tissue tumors. Among them, there were 5 myxoid soft tissue tumors, 3 lipomatous soft tissue tumors (including 1 myxoid liposarcoma), and the remaining 24 cases were cellular soft tissue tumors. Table 2 shows the distribution of soft tissue tumor diagnoses according to Ki-67 proliferation status and malignancy. There were 16 participants with a pathologically obtained Ki-67 >20%, and 15 participants with Ki-67 ≤20%.

Comparison of microstructural features of different Ki-67 groups

The values of D_ex, ADC_{40 Hz}, ADC_{25 Hz}, and ADC_{0 Hz} were significantly higher in participants with low Ki-67 values (≤20%) compared to those with high Ki-67 values (>20%) (P<0.001). However, participants with low Ki-67 values (≤20%) exhibited significantly lower cellularity, v_in, cADC, and rcADC values compared to those with high Ki-67 values (>20%). There was no significant difference in the d between high Ki-67 values and low Ki-67 values (P>0.05). These results are summarized in Table 3 and illustrated in Figure 3.

Figure 3 Box plots illustrating differences in parameters derived from time-dependent diffusion MRI between the Ki-67 high- and low-proliferation groups. *, P<0.01; **, P<0.001. Results with P<0.05 are not marked. Dots represent individual data points, boxes show standard deviations, and lines indicate medians. ADC, apparent diffusion coefficient; ADC_{0 Hz}, ADC at 0 Hz; ADC_{25 Hz}, ADC at 25 Hz; ADC_{40 Hz}, ADC at 40 Hz; cADC, change in ADC; d, cell diameter; D_ex, extracellular diffusivity; MRI, magnetic resonance imaging; rcADC, relative change in ADC; v_in, intracellular volume fraction.

Diagnostic performance of microstructural parameters derived from td-dMRI

ROC curves for several significant parameters predicting low versus high Ki-67 statuses are displayed in Figure 4; the corresponding diagnostic characteristics are presented in Table 4. Among all parameters, the rcADC exhibited the highest AUC [AUC =0.91, 95% confidence interval (CI): 0.82–1.00] for distinguishing between low and high Ki-67 states, with a sensitivity of 81.3% and a specificity of 86.7%. In addition, among the four microstructural parameters derived from td-dMRI, the v_in had the highest AUC value of 0.85 (95% CI: 0.72–0.99), with a sensitivity of 93.8% and a specificity of 66.7% for distinguishing between low and high Ki-67 states. Except for d, the AUC values of the other parameters were higher than 0.7 (P<0.05, all). The predictive thresholds for these parameters are detailed in Table 4.

Figure 4 ROC curves illustrating the diagnostic performance of time-dependent diffusion MRI parameters in differentiating low-proliferation (Ki-67 index ≤20%) and high-proliferation (Ki-67 index >20%) groups with respect to soft tissue tumors. ADC, apparent diffusion coefficient; ADC_{0 Hz}, ADC at 0 Hz; ADC_{25 Hz}, ADC at 25 Hz; ADC_{40 Hz}, ADC at 40 Hz; cADC, change in ADC; d, cell diameter; D_ex, extracellular diffusivity; MRI, magnetic resonance imaging; rcADC, relative change in ADC; ROC, receiver operating characteristic; v_in, intracellular volume fraction.

Correlation with immunohistochemistry results in all soft tissue tumors

The Pearson correlation coefficient was used to evaluate the correlation strength among D_ex, ADC_{40 Hz}, ADC_{0 Hz}, and Ki-67 values (Table 5). Spearman correlation coefficients were used to evaluate the strength of the correlations between Ki-67 values and each of the following: cellularity, v_in, ADC_{25 Hz}, cADC, and rcADC (Table 6). Figure 5 illustrates the relationship between td-dMRI parameters and Ki-67 values. Both rcADC and v_in showed strong positive correlations with Ki-67 values [r_rcADC =0.68 (95% CI: 0.41–0.83), r_vin =0.61 (95% CI: 0.32–0.80); all P<0.001]. Cellularity and cADC exhibited moderate positive correlations with Ki-67 values [r_celluarity =0.54 (95% CI: 0.22–0.76), r_cADC =0.54 (95% CI: 0.21–0.75); all P=0.002]. ADC_{0 Hz} and ADC_{25 Hz} demonstrated moderate negative correlations with Ki-67 values [r_{ADC0 Hz} =−0.52 (95% CI: −0.20 to −0.74), r_{ADC25 Hz} =−0.42 (95% CI: −0.06 to −0.68); P=0.003, P=0.020]. D_ex and ADC_{40 Hz} displayed weak negative correlations with Ki-67 values [r_Dex =−0.37 (95% CI: −0.02 to −0.64), r_{ADC40 Hz} =−0.37 (95% CI: −0.02 to −0.64); P=0.038, P=0.040]. These differences were statistically significant.

Figure 5 Scatter plots illustrating correlation between parameters derived from time-dependent diffusion MRI and Ki-67 values. ADC, apparent diffusion coefficient; ADC_{0 Hz}, ADC at 0 Hz; ADC_{25 Hz}, ADC at 25 Hz; ADC_{40 Hz}, ADC at 40 Hz; cADC, change in ADC; D_ex, extracellular diffusivity; MRI, magnetic resonance imaging; rcADC, relative change in ADC; v_in, intracellular volume fraction.

Correlation with immunohistochemistry results in cellular-type soft tissue tumors

Spearman correlation coefficients were used to assess the strength of the associations between the Ki-67 index and the quantitative parameters of td-dMRI in cellular-type soft tissue tumors (Table 7). rcADC, v_in, cADC, and cellularity showed a highly significant positive correlation with the Ki-67 index in cellular-type soft tissue tumors [r_rcADC =0.74 (95% CI: 0.46–0.88), P<0.001; r_vin =0.72 (95% CI: 0.44–0.87), P<0.001; r_cADC =0.67 (95% CI: 0.36–0.85), P<0.001; r_cellularity =0.61 (95% CI: 0.26–0.82), P=0.002]. ADC_{0 Hz} exhibited a strong negative correlation with the Ki-67 index [r_{ADC0 Hz} =−0.66 (95% CI: −0.34 to −0.84), P<0.001]. ADC_{25 Hz}, ADC_{40 Hz}, and D_ex demonstrated moderate negative correlations with the Ki-67 index [r_{ADC25 Hz} =−0.50 (95% CI: −0.11 to −0.76), P=0.013; r_{ADC40 Hz} =−0.48 (95% CI: −0.09 to −0.75), P=0.016; r_Dex = −0.46 (95% CI: −0.06 to −0.74), P=0.023].

Discussion

This study investigated the feasibility of using td-dMRI to predict the Ki-67 status of soft tissue tumors. The results showed that td-dMRI parameters (especially rcADC and v_in) were strongly positively correlated with the Ki-67 index, with significant diagnostic performance (AUCs of 0.91 and 0.85, respectively). The results suggest that td-dMRI has significant potential for noninvasively assessing tumor proliferative activity and patient prognosis.

Ki-67 proliferation status has been reported to correlate with ADC, pure diffusion coefficient (D), volume transfer constant (K_trans), and rate constant (K_ep), with ADC in particular showing a significant negative correlation with Ki-67, which may be related to restricted water diffusivity in highly proliferative tumors (5,20,25-27). This is consistent with our findings, in which ADC_{0 Hz} demonstrated a moderate negative correlation with the Ki-67 index.

However, ADC reflects only the overall diffusivity of water molecules and does not directly capture microstructural features at the cellular level, such as intracellular and extracellular spaces, cell size, membrane permeability, or intrinsic diffusivity (7). In our study, td-dMRI-derived parameters showed stronger correlations with Ki-67. Specifically, rcADC (r=0.68) and v_in (r=0.61) demonstrated strong positive correlations with the Ki-67 index. Cellularity (r=0.54) and cADC (r=0.54) exhibited moderate positive correlations. ADC_{0 Hz} (r=−0.52) and ADC_{25 Hz} (r=−0.42) showed moderate negative correlations with Ki-67, whereas D_ex (r=−0.37) and ADC_{40 Hz} (r=−0.37) displayed only weak negative correlations. Overall, these correlations were lower than those reported by Zhan et al. (5) (r_ADC =−0.71), which may be attributed to the heterogeneity of our sample population.

To address this issue, we further performed subgroup analyses focusing exclusively on cellular type soft tissue tumors and found that the correlations of all parameters were enhanced, with rcADC (r=0.74) and v_in (r=0.72) showing particularly stronger associations. We speculate that fat-rich lipomatous tumors, which typically present with low ADC, and water-rich myxoid tumors, which exhibit high ADC, may dilute the relationships between diffusion parameters and Ki-67 in pooled analyses (28,29). In contrast, cellular type tumors are characterized by densely packed cells and minimal extracellular matrix, which reduces confounding effects from non-cellular components and allows td-dMRI parameters to more reliably reflect proliferation-related microstructural features.

Previous studies have demonstrated that conventional ADC has good diagnostic performance in predicting high- and low-proliferation soft tissue tumors, demonstrating an AUC of 0.74, with 77.8% sensitivity and 69.6% specificity (30). In our study, rcADC, cADC, cellularity, and v_in all exhibited high diagnostic performance in differentiating the high-proliferation group from the low-proliferation group, with rcADC showing the highest AUC of 0.91 and sensitivity and specificity of 81.3% and 86.7%, respectively. The values of rcADC, cADC, cellularity, and v_in were significantly higher in the high-proliferation group than in the low-proliferation group. Lima et al. applied td-dMRI to distinguish benign from malignant head and neck tumors and reported that malignant tumors exhibited significantly higher rcADC values compared with benign tumors (10). This may be because the use of rcADC to assess the solid components of tumors reduces the influence of necrosis and edema, thereby providing a more accurate representation of tumor heterogeneity compared with global assessments obtained from conventional DWI (10,31). These findings are consistent with our results.

Additionally, highly malignant tumor tissues are characterized by active cell proliferation and densely packed cells, which leads to increased cellularity, reduced extracellular space, and lower extracellular diffusivity (32,33). Previous studies have demonstrated a strong association between the v_in obtained from td-dMRI and the fraction of nuclei observed in pathological examinations (7). In highly malignant tumors, the nuclei are often enlarged and exhibit an increased nuclear fraction (34-36). These characteristics of tumor microstructure are consistent with our findings.

The D_ex, ADC_{0 Hz}, ADC_{25 Hz}, and ADC_{40 Hz} values were significantly higher in the low-proliferation group compared to the high-proliferation group. This may be due to the fact that tumors characterized by high proliferation rates exhibited more diffusion-impeding microstructures, including cell membranes and fibers, as well as increased tissue density. These features resulted in lower ADC (across all diffusion times) in high-proliferation tumors compared to low-proliferation tumors (31). Increased expression of Ki-67 activates tumor cell proliferation, resulting in tightly arranged cells and decreased extracellular space, which restricts water diffusion. The restriction is reflected by decreased ADC (5,30).

No significant difference in parameter d was observed between the low- and high-proliferation groups. This may result from the inherent heterogeneity of soft tissue tumors (37). Compared with low-malignancy tumors, highly malignant soft tissue tumors exhibit greater variability in cell size. The overlapping cell diameters between benign and malignant soft tissue tumors limit the ability of cell diameter measurements to distinguish tumor malignancy (38).

Additionally, we observed that tumor ADC at different frequencies decreased as diffusion time increased, which is consistent with previous findings (10,39). However, their correlations with Ki-67 were lower than those of cellularity (r=0.54) and v_in (r=0.61). This observation suggests that, compared with measured ADC metrics, td-dMRI-derived microstructural parameters provide superior predictive value for Ki-67 status. Although tissue heterogeneity and model assumptions could pose certain limitations, td-dMRI offers a theoretical basis for more detailed quantification of tumor microstructure.

This study had several limitations. First, Ki-67 histochemistry specimens (not whole-mount) were collected by sampling a small area of the tumor, whereas td-dMRI measurements were averaged over the entire tumor region. Consequently, these two data sources may not be closely aligned. Second, the sample size was relatively small, and the cohort included a histologically heterogeneous spectrum of soft-tissue tumors, which may have affected the stability of the associations between imaging parameters and proliferative markers. Future multicenter studies with larger and more homogeneous populations are needed to further validate the reproducibility and diagnostic robustness of td-dMRI across different tumor subtypes. Third, our histological findings require further validation, potentially through comparison with those obtained via high-gradient strength systems that utilize higher oscillating frequencies.

Conclusions

We have demonstrated that td-dMRI, specifically the rcADC parameter, can noninvasively predict soft tissue tumor proliferative activity preoperatively. These findings further validate the translational potential of this technique as a noninvasive tool for guiding preoperative tumor grading, treatment planning, and prognostic assessment.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1815/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1815/dss

Funding: This study was supported by the Science and Technology Program of Hebei Province (No. 22377751D) and Hebei Natural Science Foundation (No. H2025206262).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1815/coif). Y.J. and Mengzhu Wang are employees of MR Research Collaboration Team, Siemens Healthineers Ltd., Beijing, China. T.F. is employed by, owns stocks of and holds patents filed by Siemens Healthineers AG. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Research Ethics Committee of the Third Hospital of Hebei Medical University (No. Z2022-059-1) and informed consent was taken from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Riley GM, Steffner R, Kwong S, Chin A, Boutin RD. MRI of Soft-Tissue Tumors: What to Include in the Report. Radiographics 2024;44:e230086. [Crossref] [PubMed]
2. Robba T, Chianca V, Albano D, Clementi V, Piana R, Linari A, Comandone A, Regis G, Stratta M, Faletti C, Borrè A. Diffusion-weighted imaging for the cellularity assessment and matrix characterization of soft tissue tumour. Radiol Med 2017;122:871-9. [Crossref] [PubMed]
3. Bruno F, Arrigoni F, Mariani S, Splendiani A, Di Cesare E, Masciocchi C, Barile A. Advanced magnetic resonance imaging (MRI) of soft tissue tumors: techniques and applications. Radiol Med 2019;124:243-52. [Crossref] [PubMed]
4. de Castro Luna R, Kumar NM, Fritz J, Ahlawat S, Fayad LM. MRI evaluation of soft tissue tumors: comparison of a fast, isotropic, 3D T2-weighted fat-saturated sequence with a conventional 2D T2-weighted fat-saturated sequence for tumor characteristics, resolution, and acquisition time. Eur Radiol 2022;32:8670-80. [Crossref] [PubMed]
5. Zhan J, Hao D, Wang D, Yue B, Zhou R, Tian N, Liu M, Gao C, Xu W, Cui J. Standard diffusion-weighted, intravoxel incoherent motion, and dynamic contrast-enhanced MRI of musculoskeletal tumours: correlations with Ki67 proliferation status. Clin Radiol 2021;76:941.e11-8. [Crossref] [PubMed]
6. Amornsiripanitch N, Bickelhaupt S, Shin HJ, Dang M, Rahbar H, Pinker K, Partridge SC. Diffusion-weighted MRI for Unenhanced Breast Cancer Screening. Radiology 2019;293:504-20. [Crossref] [PubMed]
7. Wu D, Jiang K, Li H, Zhang Z, Ba R, Zhang Y, Hsu YC, Sun Y, Zhang YD. Time-Dependent Diffusion MRI for Quantitative Microstructural Mapping of Prostate Cancer. Radiology 2022;303:578-87. [Crossref] [PubMed]
8. Jiang X, Li H, Xie J, Zhao P, Gore JC, Xu J. Quantification of cell size using temporal diffusion spectroscopy. Magn Reson Med 2016;75:1076-85. [Crossref] [PubMed]
9. Xu J, Jiang X, Li H, Arlinghaus LR, McKinley ET, Devan SP, Hardy BM, Xie J, Kang H, Chakravarthy AB, Gore JC. Magnetic resonance imaging of mean cell size in human breast tumors. Magn Reson Med 2020;83:2002-14. [Crossref] [PubMed]
10. Iima M, Yamamoto A, Kataoka M, Yamada Y, Omori K, Feiweier T, Togashi K. Time-dependent diffusion MRI to distinguish malignant from benign head and neck tumors. J Magn Reson Imaging 2019;50:88-95. [Crossref] [PubMed]
11. Zhang H, Liu K, Ba R, Zhang Z, Zhang Y, Chen Y, Gu W, Shen Z, Shu Q, Fu J, Wu D. Histological and molecular classifications of pediatric glioma with time-dependent diffusion MRI-based microstructural mapping. Neuro Oncol 2023;25:1146-56. [Crossref] [PubMed]
12. Wang X, Ba R, Huang Y, Cao Y, Chen H, Xu H, Shen H, Liu D, Huang H, Yin T, Wu D, Zhang J. Time-Dependent Diffusion MRI Helps Predict Molecular Subtypes and Treatment Response to Neoadjuvant Chemotherapy in Breast Cancer. Radiology 2024;313:e240288. [Crossref] [PubMed]
13. Rottmann D, Abdulfatah E, Pantanowitz L. Molecular testing of soft tissue tumors. Diagn Cytopathol 2023;51:12-25. [Crossref] [PubMed]
14. Yue Y, Liu Y, Song L, Chen X, Wang Y, Wang Z. MRI findings of low-grade fibromyxoid sarcoma: a case report and literature review. BMC Musculoskelet Disord 2018;19:65. [Crossref] [PubMed]
15. Carbone F, Pizzolorusso A, Di Lorenzo G, Di Marzo M, Cannella L, Barretta ML, Delrio P, Tafuto S. Multidisciplinary Management of Retroperitoneal Sarcoma: Diagnosis, Prognostic Factors and Treatment. Cancers (Basel) 2021;13:4016. [Crossref] [PubMed]
16. Gamboa AC, Gronchi A, Cardona K. Soft-tissue sarcoma in adults: An update on the current state of histiotype-specific management in an era of personalized medicine. CA Cancer J Clin 2020;70:200-29. [Crossref] [PubMed]
17. Bruno PS, Arshad A, Gogu MR, Waterman N, Flack R, Dunn K, Darie CC, Neagu AN. Post-Translational Modifications of Proteins Orchestrate All Hallmarks of Cancer. Life (Basel) 2025;15:126. [Crossref] [PubMed]
18. Remnant L, Kochanova NY, Reid C, Cisneros-Soberanis F, Earnshaw WC. The intrinsically disorderly story of Ki-67. Open Biol 2021;11:210120. [Crossref] [PubMed]
19. Peng Z, Zhao T, Gao P, Zhang G, Wu X, Tian H, Qu M, Tan X, Zhang Y, Zhao X, Qi X. Tumor-Derived Extracellular Vesicles Enable Tumor Tropism Chemo-Genetherapy for Local Immune Activation in Triple-Negative Breast Cancer. ACS Nano 2024;18:30943-56. [Crossref] [PubMed]
20. Lee JH, Yoon YC, Seo SW, Choi YL, Kim HS. Soft tissue sarcoma: DWI and DCE-MRI parameters correlate with Ki-67 labeling index. Eur Radiol 2020;30:914-24. [Crossref] [PubMed]
21. Tanaka K, Hasegawa T, Nojima T, Oda Y, Mizusawa J, Fukuda H, Iwamoto Y. Prospective evaluation of Ki-67 system in histological grading of soft tissue sarcomas in the Japan Clinical Oncology Group Study JCOG0304. World J Surg Oncol 2016;14:110. [Crossref] [PubMed]
22. Kershaw L, Forker L, Roberts D, Sanderson B, Shenjere P, Wylie J, Coyle C, Kochhar R, Manoharan P, Choudhury A. Feasibility of a multiparametric MRI protocol for imaging biomarkers associated with neoadjuvant radiotherapy for soft tissue sarcoma. BJR Open 2021;3:20200061. [Crossref] [PubMed]
23. Suo S, Yin Y, Geng X, Zhang D, Hua J, Cheng F, Chen J, Zhuang Z, Cao M, Xu J. Diffusion-weighted MRI for predicting pathologic response to neoadjuvant chemotherapy in breast cancer: evaluation with mono-, bi-, and stretched-exponential models. J Transl Med 2021;19:236. [Crossref] [PubMed]
24. Kamimura K, Kamimura Y, Nakano T, Hasegawa T, Nakajo M, Yamada C, Akune K, Ejima F, Ayukawa T, Ito S, Nagano H, Takumi K, Nakajo M, Uchida H, Tabata K, Iwanaga T, Imai H, Feiweier T, Yoshiura T. Differentiating brain metastasis from glioblastoma by time-dependent diffusion MRI. Cancer Imaging 2023;23:75. [Crossref] [PubMed]
25. Iima M, Kataoka M, Honda M, Le Bihan D. Diffusion-Weighted MRI for the Assessment of Molecular Prognostic Biomarkers in Breast Cancer. Korean J Radiol 2024;25:623-33. [Crossref] [PubMed]
26. Xiao Z, Zhong Y, Tang Z, Qiang J, Qian W, Wang R, Wang J, Wu L, Tang W, Zhang Z. Standard diffusion-weighted, diffusion kurtosis and intravoxel incoherent motion MR imaging of sinonasal malignancies: correlations with Ki-67 proliferation status. Eur Radiol 2018;28:2923-33. [Crossref] [PubMed]
27. Wu L, Ding L, Lin Y, Ou Y, Chen Y, Tang Y, Lin Y. Combination amide proton transfer imaging with diffusion-weighted imaging for differentiating tumor characteristics and assessing Ki-67 expression in soft tissue tumors. Magn Reson Imaging 2025;123:110490. [Crossref] [PubMed]
28. Lee S, Lee SY, Jung JY, Nam Y, Jeon HJ, Jung CK, Shin SH, Chung YG. Ensemble learning-based radiomics with multi-sequence magnetic resonance imaging for benign and malignant soft tissue tumor differentiation. PLoS One 2023;18:e0286417. [Crossref] [PubMed]
29. Benhabib H, Brandenberger D, Lajkosz K, Demicco EG, Tsoi KM, Wunder JS, Ferguson PC, Griffin AM, Naraghi A, Haider MA, White LM. MRI Radiomics Analysis in the Diagnostic Differentiation of Malignant Soft Tissue Myxoid Sarcomas From Benign Soft Tissue Musculoskeletal Myxomas. J Magn Reson Imaging 2025;61:2630-41. [Crossref] [PubMed]
30. Zhang K, Dai Y, Liu Y, Tao J, Pan Z, Xie L, Wang S. Soft tissue sarcoma: IVIM and DKI parameters correlate with Ki-67 labeling index on direct comparison of MRI and histopathological slices. Eur Radiol 2022;32:5659-68. [Crossref] [PubMed]
31. Maekawa T, Hori M, Murata K, Feiweier T, Kamiya K, Andica C, Hagiwara A, Fujita S, Koshino S, Akashi T, Kamagata K, Wada A, Abe O, Aoki S. Differentiation of high-grade and low-grade intra-axial brain tumors by time-dependent diffusion MRI. Magn Reson Imaging 2020;72:34-41. [Crossref] [PubMed]
32. Yuan J, Xie D, Fang S, Meng F, Wu Y, Shan D, Shao N, Wang B, Tian Z, Wang Y, Xu C, Chen X. Qualitative and quantitative MRI analysis of alveolar soft part sarcoma: correlation with histological grade and Ki-67 expression. Insights Imaging 2024;15:142. [Crossref] [PubMed]
33. He L, Qin Y, Hu Q, Liu Z, Zhang Y, Ai T. Quantitative characterization of breast lesions and normal fibroglandular tissue using compartmentalized diffusion-weighted model: comparison of intravoxel incoherent motion and restriction spectrum imaging. Breast Cancer Res 2024;26:71. [Crossref] [PubMed]
34. Deng S, Wu Z, Wu Y, Zhang W, Li J, Dai N, Zhang B, Yan J. Meta-Analysis of the Correlation between Apparent Diffusion Coefficient and Standardized Uptake Value in Malignant Disease. Contrast Media Mol Imaging 2017;2017:4729547. [Crossref] [PubMed]
35. Zhou T, Qiao B, Peng B, Liu Y, Gong Z, Kang M, He Y, Pang C, Dai Y, Sheng M. Predicting histological grade in pediatric glioma using multiparametric radiomics and conventional MRI features. Sci Rep 2024;14:13683. [Crossref] [PubMed]
36. Zink D, Fischer AH, Nickerson JA. Nuclear structure in cancer cells. Nat Rev Cancer 2004;4:677-87. [Crossref] [PubMed]
37. Sedaghat S, Schmitz F, Meschede J, Sedaghat M. Systematic analysis of post-treatment soft-tissue edema and seroma on MRI in 177 sarcoma patients. Surg Oncol 2020;35:218-23. [Crossref] [PubMed]
38. Shashni B, Ariyasu S, Takeda R, Suzuki T, Shiina S, Akimoto K, Maeda T, Aikawa N, Abe R, Osaki T, Itoh N, Aoki S. Size-Based Differentiation of Cancer and Normal Cells by a Particle Size Analyzer Assisted by a Cell-Recognition PC Software. Biol Pharm Bull 2018;41:487-503. [Crossref] [PubMed]
39. Iima M, Nobashi T, Imai H, Koyasu S, Saga T, Nakamoto Y, Kataoka M, Yamamoto A, Matsuda T, Togashi K. Effects of diffusion time on non-Gaussian diffusion and intravoxel incoherent motion (IVIM) MRI parameters in breast cancer and hepatocellular carcinoma xenograft models. Acta Radiol Open 2018;7:2058460117751565. [Crossref] [PubMed]