Relationships among slow diffusion coefficient and diffusion-derived ‘vessel density’ signal features, Gadoxetate delayed enhancement, tumor size, tumor grading, and patient survival: an exploratory study of mass-forming intrahepatic cholangiocarcinoma

Introduction

Cholangiocarcinoma represents a group of aggressive tumors arising from the epithelial cells at any point of the biliary tract. According to the site of origin, it can be classified into intrahepatic cholangiocarcinoma (ICC; arising proximally to the second-order bile ducts), perihilar cholangiocarcinoma (originating between second order bile ducts and the insertion of the cystic duct into the common bile duct), and extrahepatic or distal cholangiocarcinoma (below the insertion of common bile duct). ICC is the second most common type of primary hepatic malignancy. ICC can be further classified into mass-forming, periductal- infiltrating, or intraductal growth types. Mass-forming ICC is the most common, accounting for 80% of all cases. Periductal infiltrating ICC constitutes approximately 16% of ICC, while intraductal growing ICC constitutes approximately 6% (1). Surgical resection is the only potentially curative therapy for ICC, resulting in a 5-year survival rate ranging from 15% to 40% (2,3).

ICC is frequently accompanied by a dense desmoplastic stroma that surrounds the malignant ducts and glands of this tumor. The stromal myofibroblasts have a crucial role in accelerating the progression of ICC and in promoting resistance to therapy through interactive autocrine and paracrine signaling pathways that promote malignant cell proliferation, migration, invasiveness, apoptosis resistance and/or epithelial-mesenchymal transition (4,5). For example, a study by Kajiyama et al. (6) reported that scirrhous ICC with a >70% scirrhous area that has fibrous stroma amount greater than that of tumor cells is associated with frequent lymphatic permeation, perineural invasion, and a significantly lower survival, compared with those with non-scirrhous ICC. Mass-forming ICC can show a target sign, consisting of high signal at the periphery of the lesion and low signal intensity in its central parts on diffusion weighted imaging (DWI) (7). The central dark area on DWI reflects fibrosis and necrosis of the tumor, whereas the high signal at peripheral area represents highly cellular and vascular tumor (8,9).

For contrast agent enhanced (CE) imaging, a distinctive feature of ICC, due to the rich cellularity in the peripheral parts of the neoplastic mass, is typically represented by the initial peripheral enhancement in the arterial phase with subsequent centripetal enhancement in the delayed phase (1,7). The delayed phase CE is explained by rich internal fibrous stroma (1,7-9). Literature suggests that hypervascularity in arterial phase CE is associated with better prognosis. Ariizumi et al. (10) reported that, 17.9% (25/140) of their patients with mass-forming ICC exhibited hypervascularity on arterial phase computed tomography (CT); these patients had a significantly higher 5-year survival rate relative to patients with hypovascular ICC (86% vs. 27%). Hypervascular ICCs exhibited less portal vein invasion and intrahepatic metastasis relative to hypovascular ICC. Park et al. (11) reported that the arterial phase CT hypervascular lesions had better recurrence-free survival and post-recurrence prognosis compared to hypovascular CT enhancement. Min et al. (12) examined the pre-operative magnetic resonance imaging (MRI) of 134 ICCs who underwent curative resection. The 5-year risk of death in patients with mass-forming ICC with Gadoxetate diffuse arterial phase hyperenhancement was lower than that of patients with diffuse hypoenhancement or peripheral rim enhancement (5-year risk of death: 5.9%; 87.9%; 59.2%). Patients with diffuse hyperenhancement of mass-forming ICCs had less frequent vascular invasion. On the other hand, the relationship between prognosis and delayed phase CE or Gadoxetate hepatobiliary phase CE remains controversial. With CT delayed phase imaging (4–6 minutes after initiation), Asayama et al. (13) reported that mass-forming ICCs with more than two-thirds of areas with delayed CE were found to have more fibrous stroma during pathologic evaluation and exhibited more frequent perineural invasion and a poorer survival rate after surgery than did those with small areas of delayed CE. Koh et al. (8) categorized the Gadoxetate hepatobiliary phase signal intensity into intermediate or hypointense groups. ICC in the intermediate group were associated with shorter survival time and time to recurrence than the ICC of the hypointense group. The ICCs of the intermediate group showed a tendency for more abundant tumour fibrous stroma than those of the hypointense group. Gabata et al. (9) also noted that the ICCs showing gadopentetate dimeglumine delayed CE corresponded to abundant fibrous stroma. On the other hand, Jin et al. (14) reported that ICCs with gadopentetate dimeglumine arterial diffuse hyper CE or delayed uniform CE (2 min 30 sec to 3 min after injection) had lower preoperative carbohydrate antigen 19-9 (CA19-9) levels, smaller tumor sizes and they were less associated with lymph nodes metastasis, vascular invasion, necrosis or poor tumor differentiation, therefore with higher overall and disease-free survival rates. The uniform CE in delayed MRI increased the detection rate of patients with good prognosis compared to the arterial diffuse hyper enhancement pattern. Kang et al. (15) assessed the enhancement patterns of ICC on the Gadoxetate hepatobiliary phase. They noted that percentage of relative enhancement on hepatobiliary phase was significantly higher in moderately differentiated (66.4%±42.1%) than in poorly differentiated (36.84%±21.5%) tumors and in patients without (59.7%±28.8%) than in those with (24.9%±14.7%) lymph node metastasis. It is likely that the type and dosage of contrast agents and the delay duration between the contrast agent injection and the image acquisition would affect the results, and these two factors should be critically considered when interpreting the results.

Recently two new types of DWI contrast, diffusion-derived ‘vessel density’ (DDVD) and slow diffusion coefficient (SDC), have been tested in the liver. Liver micro-vessels show high-signal when there is no motion probing gradient (b=0 s/mm²) and low-signal when even very low b-value diffusion gradients (such as b=2 s/mm²) are applied. Thus, the signal difference between images when the motion probing gradient is ‘off’ and ‘on’ reflects the extent of functional tissue vessel density, and this metric reflecting tissue perfusion is termed as DDVD (16-18). SDC was proposed to measure tissue slow diffusion (19). In addition to minimizing the perfusion contribution (19,20), SDC also mitigates the ‘T2 shining-through’ effect associated with apparent diffusion coefficient (ADC) and may better reflect true tissue diffusion. In its basic form, SDC is derived from a high b-value DWI image (typically with b-value of 400–500 s/mm²) and a higher b-value DWI image (typically with b-value of 600–800 s/mm²). With the conventional ADC approach, the spleen has been reported to have a much lower ADC than the liver, hepatocellular carcinomas (HCCs) have a lower ADC than liver parenchyma. On the other hand, with SDC analysis, the spleen has a faster diffusion than the liver and HCCs have a faster diffusion than liver parenchyma (19). The liver and spleen have a similar amount of blood perfusion, the spleen is waterier than the liver. HCCs are mostly associated with increased blood supply and increased proportion of arterial blood supply and with edema. It is more reasonable with SDC results that the spleen and HCCs have faster diffusion than liver parenchyma. A combination of SDC and DDVD can provide tissue diffusion and perfusion information initially promised by intravoxel incoherent motion imaging while with a much faster data acquisition and allowing easier pixelwise mapping. We noted that SDC in ICC was positively correlated with DWI signal when b=0 s/mm² (21). Stromal fibrosis is associated with lower DWI signal, thus it is likely that the higher SDC signal ICCs may be associated with less amount of fibrous stroma, this in turn may be associated with a better outcome. An ICC with higher DDVD may also be associated with potential better patient survival due to the better blood supply and less stromal fibrosis. This study evaluates the relationship among SDC and DDVD signal features, Gadoxetate delayed enhancement, tumor size and tumor grading, and survival in patients with mass-forming ICC.

Methods

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Data of 24 mass-forming ICC cases were prospectively acquired in two centers with institutional ethical approvals and informed consent obtained from individual participants (Center 1 with ethical approval number of S1709442250839: Sun Yat-Sen Memorial Hospital, Guangzhou, China; and Center 2 with ethical approval number of KY2025058, Fifth Affiliated Hospital of Anhui Medical University, Fuyang, China). The pre-treatment liver DWI was based on a single-shot spin-echo type echo-planar sequence, and the default spectral pre-saturation technique was used for fat suppression (Table 1). All cases had histopathological diagnosis. Follow-up data were available in 20 cases from Center 1 (n=9 alive till the last follow-up, n=10 died during the follow-up with known death date, and 1 died shortly after ICC diagnosis and exact death date unknown).

We were able to conduct additional histopathological grading for 14 cases from Center 1. ICCs were graded into three categories: ‘light/moderately malignant ICC’ (group A, this being the least aggressive grade, n=4), ‘moderately malignant ICC’ (group B, n=6), and ‘highly malignant ICC’ (group C, n=4) (22-24). For the grading, we re-analyzed postoperative pathology reports and tissue sections of enrolled patients. To assist in the classification of the aggressiveness of the tumors, we applied a semi-quantitative histopathological risk score (HRS) system, covering recognized prognostic indicators. Core morphological risk factors (assigned 1 point for each factor) were: vascular/lymphatic invasion, perineural invasion, tumor capsule/marginal invasion, satellite lesions, regional lymph node metastasis, poor differentiation, aggressive growth pattern (predominantly nested/solid structures, lacking glandular structures), definite large-scale tumor necrosis, and strongly positive stromal fibrosis (promoting connective tissue proliferation in the stroma); if microscopic or intraoperative findings confirm distant/multiple vascular tumor emboli, distant satellite lesions, or distant lymph node metastasis, the an additional 1 point was added for each abnormality. Immunophenotypic and microenvironment indicators (assigned 0.5 point for each factor) were: high Ki-67 expression (>50%), abnormal p53 expression (completely negative with no expression, or >70% diffusely strongly positive expression suggesting a mutant type), significant lack of tumor stromal inflammatory cell infiltration (suggesting immune tolerance/“cold tumor” phenotype), and positive Epstein-Barr virus (EBV) infection. The cumulative HRS score was calculated, and patients were then divided into three groups. ‘Light/moderately malignant ICC’ had an HRS of ≤2.5. These tumors were moderately to well-differentiated, with glandular growth structures preserved microscopically. The borders and capsule structure were usually relatively intact, and there were no definite adjacent satellite lesions or vascular emboli. ‘Moderately malignant HCC’ tended to have an HRS of 3–5.5. These tumors show localized capsule invasion or localized breakthrough, with a single adjacent satellite lesion or microvascular emboli, the surgical margins remained negative, and no definite tumor metastasis was found in the adjacent lymph nodes dissected intraoperatively. ‘Highly malignant HCC’ tended to have an HRS of ≥6. These tumors often exhibited extensive invasion beyond the border or capsule, frequently accompanied by multiple satellite lesions, microvascular emboli, poor differentiation, and nested growth structures. Intraoperatively, lymph node metastases were commonly observed, and often with p53 mutation expression. Since a full HRS assessment could not be completed for the patients without surgical resection of the tumor, our histopathological classification of the tumors was assisted by this HRS, but not totally dependent on HRS.

SDC weighted maps (n=24) were derived from the equation (19):

SDC=S(b1)−S(b2)b2−b1[unit:arbitraryunit(au)/s][1]

where b₁ and b₂ refer to a high b-value and a higher b-value, respectively; S(b₁) and S(b₂) denote the DWI signal-intensity acquired at the high b-value and the higher b-value, respectively. SDC was calculated with b=500 and 800 s/mm² images for Center 1 data, and with b=400 and 600 s/mm² images for Center 2 data.

DDVD weighted maps (only available in n=16 ICCs) were calculated from b=0 and 10 s/mm² images, and derived from the equation (16,17):

DDVD=S(b0)−S(b10)[unit:au/pixel][2]

where b₀ and b₁₀ refer b=0 and 10 s/mm², respectively; S(b₁₀) and S(b₁₀) denote the DWI signal-intensity acquired at b=0 and 10 s/mm², respectively.

ADC weighted maps (n=24) were calculated according to

ADC=ln(S(b1)/S(b2))b2−b1[unit:×10−3mm2/s][3]

where b₂ and b₁ refers to b=800 and 0 s/mm², respectively, for Center 1 data, and where b₂ and b₁ refers to b=600 and 0 s/mm², respectively, for Center 2 data. S(b₁) and S(b₂) denote the DWI signal-intensity acquired at the b-factor value of b=0 and 800 s/mm², respectively, for Center 1 data; and at the b-factor value of b=0 and 600 s/mm², respectively, for Center 2 data.

The image analysis and scoring were conducted by two readers in consensus (a specialist radiologist and one senior radiology trainee). A semi-quantitative score (SQS) was taken on DDVD map and SDC map (Figure 1) (21,25,26). The scoring was based on the dominant solid part of the lesion. The signal of the background liver, spleen, kidneys, and liquid/vessels were used as the references. To facilitate visual assessment, the SDC color scale was adjusted so that the kidneys appeared as a non-saturated reddish-orange, typically rendering liquid signal as saturated red on this scale. Relative to the adjacent liver signal, tumor SDC and DDVD signals were assigned to six categories: low signal (scored as ‘0’), iso-signal (scored as ‘1’), slightly high signal (scored as ‘1.5’), high signal (scored as ‘2’), higher signal (scored as ‘2.5’), and markedly high signal (scored as ‘3’). On SDC, ‘high signal’ was typically the expected spleen signal, and ’slightly high signal meant the signal higher than the background liver but notably lower than the spleen signal; ‘higher signal’ was signal notably higher than spleen signal, but lower than liquid signal and might be slightly higher or close to the kidney signal; and ‘markedly high signal’ was usually close to the liquid or blood vessel signal. Note that, liquid signal was saturated red according to our color scaling, thus an SQS of 3.0 for a solid lesion does not mean its SDC value was absolutely as high as that of liquid. On DDVD map, DDVD ‘higher signal’ (score: 2.5) is uncommon (none in the current study), and would be higher than DDVD signal of the kidneys; ‘markedly high signal’ was usually blood vessel signal, artifacts for stomach liquid also commonly show ‘markedly high signal’ (21,25,26).

Figure 1 Examples of SDC/DDVD visual scoring and visual ‘% delay CE’ estimation in four ICC cases (arrows). The tumor morphology is shown with T2WI (A1,B1,C1,D1). (A2) The tumor has markedly high SDC signal (SQS: 3.0), being similar to the GI fluid signal and higher than the kidney signal. (B2) The tumor has high SDC signal (SQS: 2.0), being similar to the spleen signal. (C2) The tumor has only slightly high SDC signal (SQS: 1.5). (D2) The tumors had markedly high SDC signal (SQS: 3.0), being much higher than the spleen signal. One DDVD maps, ICC shows iso-signal (SQS: 1.0) on (A3,B3,C3) [there likely be necrotic areas in (A3) which show artificial high signal], and show high signal (SQS: 2.0) on (D3) (higher than the liver signal and being similar to the spleen signal). The proportion of visual ‘% delayed CE’ was estimated to be 70%, 67%, 20%, and 85% on (A4,B4,C4,D4), respectively. ‘% delay CE’, percentage area of contrast agent enhanced during delayed phase; DDVD, diffusion-derived ‘vessel density’; GI, gastrointestinal tract; ICC, intrahepatic cholangiocarcinoma; SDC, slow diffusion coefficient; SQS, semi-quantitative score; T2WI, T2-weighted imaging.

In addition to the SQS, SDC and ADC values were quantitatively measured in the MRI section showing the largest tumor size. The ICC tumor’s values were normalized by the adjacent liver value as ln(SDC_ICC/SDC_liver) and (ADC_ICC/ADC_liver).

Delayed phase MRI scan (n=20) was obtained a median delay of 3 min 51 sec (range: 2 min 52 sec to 4 min 20 sec) after the injection of standard dose of Gadoxetate (0.025 mmol/kg). Four cases did not have such a delayed MRI scan. Percentage area of CE during delayed phase (‘% delay CE’) was estimated visually in consensus by two readers (Figure 1) and also measured quantitatively (Figure 2). Visually, the percentage area of contrast enhanced tumor area was estimated and compared to the total tumor area. For quantification, the spleen CE signal during delayed phase, which was typically lower than the liver CE signal, was used as the reference signal. ‘% delay CE’ was tumor lesion pixel number proportionally above a threshold relative to the spleen CE signal intensity, and thresholds of 70%, 75%, 80%, 85%, 90%, 95%, 100%, 105%, 110%, and 115% of the spleen CE signal intensity were tested (Figure 2). It is noted that, while reflecting the same physiological phenomenon, visual ‘% delay CE’ and quantitative ‘% delay CE’ measured differently. Visual ‘% delay CE’ estimated the proportion of enhanced tumor area relative to the total tumor area, while quantitative ‘% delay CE’ measured the proportion of pixels higher than a certain threshold [i.e., (pixel number of those above the threshold)/(total tumor pixel number)]. Visual ‘% delay CE’ estimation and quantitative ‘% delay CE’ measurement were conducted independently by different raters.

Figure 2 Testing of quantitative ‘% delay CE’. (A1) A Gadoxetate delayed phase MRI. (A2-A10) Firstly a ROI was drawn on A1 for the tumor and the tumor was segmented, then the mean spleen signal during the delayed phase was used as the reference signal, pixels inside the tumor were detained when their signal intensities were higher than the ‘mean spleen signal ×0.70’ (A2), or higher than the ‘mean spleen signal ×0.75’ (A3), or higher than the ‘mean spleen signal ×0.80’ (A4), or higher than the ‘mean spleen signal ×0.85’ (A5), or higher than the ‘mean spleen signal ×0.90’ (A6), or higher than the ‘mean spleen signal ×0.95’ (A7), or higher than the ‘mean spleen signal ×1.00’ (A8), or higher than the ‘mean spleen signal ×1.05’ (A9), or higher than the ‘mean spleen signal ×1.10’ (A10). Quantitative ‘% delay CE’ is: (pixel number of those above the threshold)/(total tumor pixel number). Considering all the cases in the current study, pixels inside the tumor were detained when their signal intensities were higher than the mean spleen signal (A8) was considered a good compromise (see the “Results” section). Adopting the result of (A8), quantitative ‘% delay CE’ was 80% for this case. According to visual estimation, ‘% delay CE’ was 70% for this case (Figure 1, A4). ‘% delay CE’, percentage area of contrast agent enhanced during delayed phase; MRI, magnetic resonance imaging; ROI, region of interest.

Tumor size (presenting size when the pre-treatment MRI scan was taken) was measured on the section showing the largest size by the largest diameter (unit in cm) or by region of interest (ROI) based area (unit in mm²).

For statistical analysis, data were processed using GraphPad Prism (San Diego, CA, USA). Comparisons were performed using Kruskal-Wallis test or Mann-Whitney test. Pearson correlation and Spearman correlation were used to test the strength and direction of linear relationships between two variables. Kaplan-Meier analysis was used to investigate the relationship between SDC SQS (≥2.5 vs. ≤2.0) and patient survival during the follow-up. As this was an exploratory study, a P<0.05 was considered statistically significant, and P<0.10 was considered having a trend of significance.

Results

The degree of tumor malignancy was negatively correlated with quantitative ‘% delay CE’ with a Spearman r_sof 0.703 (P=0.008) (Figure 3A). A trend was noted that better differentiated tumors had a smaller size, though without statistical significance with the sample size in this study (Figure 3B).

Figure 3 Relationship between quantitative ‘% delay CE’ and tumor grading (A), and relationship between tumor size (area in mm² measured at the largest section) and tumor grading (B). ‘% delay CE’, percentage area of contrast agent enhanced during delayed phase; r_s, Spearman r_s.

Based on visual assessment, both tumor SDC SQS and DDVD SQS were moderately and positively correlated with visual ‘% delay CE’, with a Spearman r_s of 0.504 (P=0.02) and 0.549 (P=0.03), respectively (Figure 4A,4B). The product of ‘(SDC SQS) × (DDVD SQS)’ was positively correlated with visual ‘% delay CE’ with an improved Spearman r_s of 0.635 (P=0.009) (Figure 4C). There was no apparent correlation between SDC SQS and tumor size (Figure 4D), and also no apparent correlation between DDVD SQS and tumor size (Figure 4E). The tumor size was negatively correlated with visual ‘% delay CE’, with a Spearman r_s of −0.381 (P=0.10) (Figure 4F). Adjusted for lesion size, correlation strength between SDC SQS and visual ‘% delay CE’ was Spearman r_s=0.347 (P=0.15), correlation strength between DDVD SQS and visual ‘% delay CE’ was Spearman r_s=0.425 (P=0.11). Thus, the correlation between SDC SQS and visual ‘% delay CE’, and the correlation between DDVD SQS and visual ‘% delay CE’, were not dominated by tumor size.

Figure 4 Correlations between visual ‘% delay CE’ and SDC SQS (A), between visual ‘% delay CE’ and DDVD SQS (B), between visual ‘% delay CE’ and product of ‘(SDC SQS) × (DDVD SQS)’ (C), between SDC SQS and lesion size (one dimensional diameter in cm at the largest section) (D), between DDVD SQS and lesion size (E), and between ‘% delay CE’ and tumor size (F). ‘% delay CE’, percentage area of contrast agent enhanced during delayed phase; DDVD, diffusion-derived ‘vessel density’; r_s, Spearman r_s; SDC, slow diffusion coefficient; SQS, semi-quantitative score.

The correlation between ln(SDC_ICC/SDC_liver) and quantitative ‘% delay CE’ is shown in Figure 5. Ln(SDC_ICC/SDC_liver) was moderately and positively correlated with quantitative ‘% delay CE’. Correlation with quantitative ‘% delay CE’ by relative thresholds of 90%, 95%, 100%, 105%, and 110% spleen CE signal derived Pearson r of 0.441 (P=0.052), of 0.470 (P=0.04), of 0.481 (P=0.03) of 0.490 (P=0.03), and 0.480 (P=0.03), respectively. The correlation between visually estimated SDC SQS and quantitative ‘% delay CE’ with the threshold of 100% spleen CE signal derived Pearson r of 0.387 (P=0.09) and Spearman r_s of 0.322 (P=0.17), respectively.

Figure 5 Correlations between ‘% delay CE’ and quantitative ln(SDC_ICC/SDC_liver) for the thresholds of 90% (A), 95% (B), 100% (C), 105% (D), 110% (E) of the spleen CE signal intensity, and between quantitative ‘% delay CE’ and visual SDC SQS (F). ‘% delay CE’, percentage area of contrast agent enhanced during delayed phase; CE, contrast agent enhanced; ICC, intrahepatic cholangiocarcinoma; r, Pearson r; ROI, region of interest; SDC, slow diffusion coefficient; SQS, semi-quantitative score.

All patients with follow-up (n=20) were from Center 1. The alive patients (n=9) had a median follow-up of 839 days [95% confidence interval (CI): 498–942], while the dead patients (n=11) had a median survival of 255 days (95% CI: 120–821) (Figure 6A). Alive patients had a mean age of 59 years (range: 48–71 years) and four cases were female; dead patients had a mean age of 55.2 years (range: 40–76 years) and six cases were female. The median tumor size was larger for the dead patient group but without statistical significance (1,508 vs. 2,563 mm²) (Figure 6B). The alive group (n=9) had a SDC SQS higher than the dead group (n=11, mean: 2.89 vs. 2.23, P=0.007) (Figure 6C). The dead group had 4 cases with SDC SQS ≥2.5, three of them had a large tumor size (>4,000 mm²), and an additional patient had an advanced age of 76 years. DDVD SQS was higher in the alive group than in the dead group, but without statistical significance achieved in this study (P=0.29) (Figure 6D). The product of ‘(SDC SQS) × (DDVD SQS)’ was also higher in the alive group (Figure 6E). Figure 6F and Figure 7 suggested that ICCs with a high SDC score (≥2.5) and small/intermediate presenting tumor size (<4,000 mm² in the largest section) were associated with a better survival potential. ICCs with lower SDC score (≤2.0) were likely associated with a poor prognosis regardless of the presenting tumor size. Of note, one patient had an initial SDC score of 3.0 with a tumor size of 21,437.5 mm² at the largest cross section was followed up for 19 months and did not show tumor recurrence at the latest follow-up.

Figure 6 Differences between the alive group and the dead group in follow-up days (A), tumor size (B), SDC SQS (C), DDVD SQS (D), product of ‘(SDC SQS) × (DDVD SQS)’ (E), and the relationship between tumor size and SDC SQS for the alive group and the dead group (F). In (B-D), green and blue colors indicate two cases with large tumor size, and brown color indicates a case with advanced aged (76 years). In (F), the brown arrow indicates a patient aged 76 years, and the green arrow indicates a tumor with size of 21,437.5 mm² and this patient was followed up for 19 months and did not show tumor recurrence at the latest follow-up. All patients with follow-up data were from Center 1 (Sun Yat-Sen Memorial Hospital). ‘% delay CE’, percentage area of contrast agent enhanced during delayed phase; DDVD, diffusion-derived ‘vessel density’; MRI, magnetic resonance imaging; SDC, slow diffusion coefficient; SQS, semi-quantitative score.

Figure 7 Kaplan-Meier survival analysis according to patients with tumor SDC SQS ≥2.5 or ≤2.0. Patients having a tumor SDC SQS ≥2.5 were more likely to survive during the follow-up in this study. All patients with follow-up data are from Center 1 (Sun Yat-Sen Memorial Hospital). SDC, slow diffusion coefficient; SQS, semi-quantitative score.

All alive patients (n=9) had SDC SQS ≥2.5, 8 of them were treated surgically after ICC diagnosis, and then followed by chemotherapy. One was a relatively recent patient, and no surgery was conducted (yet). For the dead patients (n=11), 5 had surgery treatment followed by chemotherapy, 1 case had radiofrequency ablation and intra-arterial chemotherapy, 1 case had intra-arterial chemotherapy, and 3 cases had only standard chemotherapy, 1 case was not further treated in the authors’ institution. Since all these 20 patients with follow-up were initially cared at Center 1 and the same treatment guidelines were followed, these data show ICC with SDC SQS ≥2.5 were more likely to be candidate for surgical removal of the tumor.

The majority of the ICCs had a higher ADC value than liver parenchyma [i.e., (ADC_ICC/ADC_liver) >1]. There was no statistically significant correlation between (ADC_ICC/ADC_liver) and histological malignancy (Figure 8A). (ADC_ICC/ADC_liver) was not correlated with quantitative ‘% delay CE’ (Figure 8B), nor with visual ‘% delay CE’ (Figure 8C). There was no statistically significant correlation between (ADC_ICC/ADC_liver) and patient survival (Figure 8D).

Figure 8 Potential relationship between (ADC_ICC/ADC_liver) and tumor differentiation (A), between (ADC_ICC/ADC_liver) and quantitative ‘% delay CE’ (B), between (ADC_ICC/ADC_liver) and visual ‘% delayed CE’ (C), and between (ADC_ICC/ADC_liver) and patient survival (D). There is a weak trend that the ‘highly malignant ICC’ have a lower (ADC_ICC/ADC_liver), but statistical significance was not achieved with the samples in this study. The color signs in (D) are the same as in Figure 6. ‘% delay CE’, percentage area of contrast agent enhanced during delayed phase; ADC, apparent diffusion coefficient; CE, contrast agent enhanced; ICC, intrahepatic cholangiocarcinoma; r, Pearson r; r_s, Spearman r_s; ROI, region of interest.

Discussion

This study shows ‘% delay CE’, which was scanned on average 3 min 51 sec after Gadoxetate injection, was positively correlated with better tumor differentiation (Figure 3A). This is consistent with the concept that hypervascular ICC has a better prognosis (10-12), as a higher delayed phase CE is also likely associated with tumor hypervascularity. Our results differ from the CT results described by Asayama et al. (13), where ICCs with delayed CT CE seen in more than two-thirds of areas exhibited more frequent perineural invasion and a poorer survival rate after surgery than did those with delayed CE seen in smaller areas. In the study reported by Asayama et al., delayed phase scanning begun 4–6 minutes after initiation of the injection, whereas in the current study delayed scan was obtained at a median of 3 min 51 sec (range: 2 min 52 sec to 4 min 20 sec) after the injection. As we noted earlier, the type and dosage of contrast agents and the delay duration between the contrast agent injection and the image acquisition would likely affect the results. Our results may concur with the results described by Jin et al. (14), where ICCs with a Magnevist delayed uniform CE had smaller tumor sizes and they were less associated with lymph nodes metastasis, vascular invasion, necrosis or poor tumor differentiation. Kang et al. (15) also described that percentage of relative enhancement on Gadoxetate hepatobiliary phase images was significantly higher in moderately differentiated tumors than in poorly differentiated tumors. As expected, a weak trend was noted in this study that ‘light/moderately malignant’ tumors had a smaller tumor size than those of the tumors of more aggressive grade (Figure 3B), indicating that tumors of more aggressive grade had a faster growth.

ICCs with higher SDC signal may be associated with more cellularity and less amount of fibrous stroma, and this in turn may associate with a better outcome (Figure 9). The result of Figure 4A shows a higher SDC SQS was associated with higher ‘% delay CE’, and this in turn would reflect more tumor cellularity and better tumor differentiation. As expected, Figure 4B shows higher DDVD SQS was also associated with higher ‘% delay CE’, reflecting that Gadoxetate delayed phase CE is positively associated with tumor vascularity. Figure 4C shows a product of ‘(SDC SQS) × (DDVD SQS)’ further improved the correlation strength between such a metric and ‘% delay CE’. This study did not show an apparent correlation between SDC SQS and tumor size, also no apparent correlation between DDVD SQS and tumor size. It could be expected that when a tumor is large, the ‘% delay CE’ would be more difficult to be high. Figure 4F indeed shows such a trend (i.e., the larger tumor size, the less ‘% delay CE’), but the correlation did not appear to be strong.

Figure 9 Tentative explanation of the relationship between tumor fibrosis and vascularity, % CE (median of 4 min after 0.025 mmol/kg Gadoxetate), DDVD, and SDC, and patient survival. CE, contrast agent enhanced; DDVD, diffusion-derived ‘vessel density’; SDC, slow diffusion coefficient.

We tried to see whether a quantitative approach could improve the correlation strength between SDC signal and ‘% delay CE’. The results of this study show, by quantitative measurement, the correlation strength of SDC signal and ‘% delay CE’ remained modest with a Pearson r of around 0.5, being broadly similar to the visual assessment results (Figure 5). These results confirmed the reliability of visual assessment. However, a comparison of Figure 5C and Figure 5F shows, when the Y-axis was both ‘100% spleen CE signal’ threshold and the X-axis was in quantitative ln(SDC_ICC/SDC_liver) or in visual SDC SQS, correlation strength improved from 0.387 to 0.481 and P value improved from 0.09 to 0.03 favoring results with quantitative ln(SDC_ICC/SDC_liver). Thus, quantitative measurement could modestly improve the correlation strength. However, in radiological practice, visual scoring is a more time-efficient approach.

The most interesting results of this study are shown with Figures 6,7. The alive patients had a median follow-up of 839 days, while the dead patients had a median survival of 255 days. Figure 6C shows the alive group had SDC SQS higher than the dead group (P=0.007). It is noteworthy that, the dead group had 4 cases with SDC SQS ≥2.5, three of them had large tumor sizes, and an additional patient had an advanced age of 76 years. Figure 6D shows a trend that the alive group also had a DDVD SQS higher than the dead group, but without statistical significance (P=0.45). That the alive group has higher DDVD SQS is consistent with that hypervascular ICC has a better prognosis (10-12). Figure 6F suggests that an ICC patient might survive if the SDC SQS was high and the tumor size was not large (e.g., <4,000 mm² in the largest section). On the other hand, an ICC patient might not survive if the SDC SQS was ≤2.0 regardless of its size. Thus, upon further validation, SDC may be an important prognostic factor for mass-forming ICC.

For our results in this study with limited data, ADC was not correlated with ‘% delay CE’ (Figure 8B,8C), and not correlated with survival (Figure 8D). The literature results of ICC’s ADC results have been conflicting. Lewis et al. (27) noted that a lower ADC was associated with a higher degree of malignancy. Sheng et al. (28) reported ICCs with low ADC values showed a shorter progression-free survival than those with high values. Similarly, Yamada et al. (29) also noted that 5-year overall survival rate in the low ADC ICC group was significantly worse than in the high ADC ICC group. On the other hand, Zhou et al. (30,31) reported that higher ADC values were positively correlated with microvascular invasion and lymph node metastasis. Lee et al. (32) divided ICC patients into those in whom less than one-third of the tumor showed diffusion restriction (group 1) and those in whom more than one third of the tumor showed diffusion restriction (group 2). The 1- and 3-year overall survival rates were 77% and 26%, respectively, in group 1, and 92% and 67%, respectively, in group 2. Thus, a larger area of diffusion restriction is associated with better survival. These paradoxical results are at least partially moderated by the ‘T2 shine-through’ effect for ADC, as ADC is non-linearly associated with the T2 value of the tissues (33,34). In the current study, despite the statistical non-significancy, a trend was still noted that the highly malignant ICCs (group C) tended to show a lower ADC (Figure 7A). In addition to the more reasonableness with SDC compared with ADC as noted in the introduction (19,20), SDC has showed advantage over ADC in a number of clinical applications. For example, patients with advanced liver fibrosis show an abnormal regulation of extracellular fluid volume, resulting in the accumulation of sodium and water retention, ascites, oedema and pleural effusion. Ascites commonly occurs in diseases causing sinusoidal portal hypertension such as liver cirrhosis (i.e., stage 4 liver fibrosis). Despite the increased water content in cirrhotic livers, studies reported a decrease in ADC in liver cirrhosis (35); on the other hand, liver cirrhosis has an increased SDC (36). Based on T2-weighted imaging (T2WI) signal and the three metrics of DDVD, SDC, and ADC, we applied a scoring scheme termed ‘LiverMss-FNH’ to evaluate liver solid mass. In two studies totaling 25 focal nodular hyperplasia (FNH) and 132 liver malignant tumors, it was shown that LiverMss ≥3.0 suggests the possibility of a liver mass being FNH, and LiverMss ≥4 can strongly favor the diagnosis for FNH (25,26). A typical SDC signal, i.e., being iso-signal or slightly high signal relative to the liver signal (but not high signal), has an odds ratio of 38 in favor of liver FNH over liver malignant tumors (25,26), while a typical ADC SQ score, i.e., being iso-signal or slightly high signal (but not low signal), has an odds ratio of only around 8 in favor of liver FNH over liver malignant tumors (assuming malignant tumors typically have a low ADC) (25,26). In a study of 63 patients with diffuse gliomas [30 isocitrate dehydrogenase (IDH)-mutant and 33 IDH-wildtype], IDH mutant negative tumors had SDC_b500b750 value of 0.339±0.055 au/s, IDH mutant positive tumors had SDC_b500b750 value of 0.437±0.097 au/s, with area under receiver operating characteristic curve (AUROC) of 0.828 for separation. IDH mutant negative tumors had ADC_b0b1000 value of 0.985±0.235 mm²/s, IDH mutant positive tumors had ADC_b0b1000 value of 1.290×10⁻³±0.381×10⁻³ mm²/s, with AUROC of 0.760 for separation (37,38). Thus, SDC as a biomarker offers a better lesion differentiation power than ADC for IDH mutation status. In a recent analysis of Ki-67 labeling index (LI) of 22 HCCs, SDC had a receiver operating characteristic area under the curve (ROAUC) of 0.825 (P=0.01) for the classification of Ki-67 low vs. high expression (Ki-67 LI <20% vs. ≥20%), ADC had a ROAUC of 0.708 (P=0.10) for the classification of Ki-67 low vs. high expression, and a combination of SDC and ADC had a ROAUC of 0.958 (P<0.001) for the classification of Ki-67 low vs. high expression (39).

This study has many limitations. The most important limitation is the small sample size, thus the results in study should be considered as exploratory or hypothesis-generating. The histopathological grading data and the follow-up data were from Center 1 only. Many of the statistical comparisons were only marginally significant, or a trend was shown but statistical significance was not achieved. The small sample size also limited our ability to conduct more comprehensive analyses. In addition to the biology of the tumors, the survival of the patients is also affected by other factors such as the general condition of the patients, adequacy of the surgery and post-surgery treatment, these could not be well controlled in this study. However, our data suggests that ICCs with SDC SQS ≥2.5 were more likely to be candidates for surgical removal of the tumor. This study evaluated delayed CE only. The relationship between high arterial CE and better ICC tumor differentiation has been well documented (10-12). However, there remains controversy in ICC tumor differentiation vs. delayed phase CE or hepatobiliary phase CE (13-15). Compared with standard extra-cellular gadolinium agents, arterial phase images using Gadoxetate are associated with weaker arterial enhancement (40,41). Compared with arterial phase CE, delayed phase CE or hepatobiliary phase CE is less affected by the contrast agent injection rate and the circulation speed of the study subjects, offering a more stable assessment. Jin et al. (14) described that the Magnevist delayed phase CE increased the detection rate of patients with good prognosis compared to the arterial CE pattern. The DWI data in this study were not acquired with a breathhold, while misalignment between b=0 and 10 s/mm² images, or between b=500 (or 400) and b=800 (or 600) s/mm² images, can lead to erroneous DDVD/SDC signal. The number of excitations (NEX) was only with Center 1, this led to 6 ICCs’ DDVD maps being not usable (all Center 1 cases). Moreover, we have recently described that there is a higher degree of difference in DWI image distortions between b=0 s/mm² DWI image and b=10 s/mm² DWI image for DDVD mapping than between the high b-value and higher b-value DWI images for SDC mapping (42). The image distortion associated with echo-planar imaging (EPI) acquisition may also be partially overcome by turbo spin echo DWI. Compared with traditional EPI DWI used in the current study, turbo spin echo DWI can reduce image geometric distortions (43-45). This approach can be tested in the future. In this study, SDC was calculated with b₁=500 and b₂=800 s/mm² images for Center 1 data, with b₁=400 and b₂=600 s/mm² images for Center 2 data. According to liver IVIM (Intravoxel Incoherent Motion) observations, once b-values is ≥400 s/mm², the relationship between DWI signal decay and b-value increase follows a linear pattern except when DWI signal is contaminated by noises (46). The DWI scan parameters differed among the two centers though this is unlikely to have affected the SDC conclusions of this study. It is anticipated that the value of ADC_b0b800 would be only slightly higher than ADC_b0b600, and contribution of b-value difference to ‘(ADC_ICC/ADC_liver)’ will be very small. For Figure 7A and Figure 7D results, only Center 1 data were analysed.

Conclusions

In conclusion, higher DDVD SQS and higher SDC SQS are associated with higher ‘% delay CE’, which may reflect higher perfusion and less fibrous stroma in ICC. A higher ‘% delay CE’ is associated with a better tumor differentiation. A higher SDC SQS is associated with a better survival potential. ICCs with SDC SQS ≤2.0 are likely associated with a poor prognosis regardless of the initial presenting tumor size. Due to the small sample size in our study, there is a strong need for larger, standardized validation studies to strengthen our conclusions.

Acknowledgments

None.

Footnote

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0225/dss

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0225/prf

Funding: This was supported by Guangdong Medical Association Clinical Research Fund (ID 2025YX-B1002).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0225/coif). F.Y.X. reports that the metric SDC is associated with a pending patent application (co-inventorship: F.Y.X.). Y.X.J.W. reports that he is the founder of Yingran Medicals Ltd., which develops medical image-based diagnostics software. The metric DDVD is associated with a granted China patent (No. ZL201910125747.2, inventorship). The metric SDC is associated with a pending patent application (inventorship). 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. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Data of 24 mass-forming ICC cases were prospectively acquired in two centers with institutional ethical approvals and informed consent obtained from individual participants (Center 1 with ethical approval number of S1709442250839: Sun Yat-Sen Memorial Hospital, Guangzhou, China; and Center 2 with ethical approval number of KY2025058, Fifth Affiliated Hospital of Anhui Medical University, Fuyang, China).

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. Seo N, Kim DY, Choi JY. Cross-Sectional Imaging of Intrahepatic Cholangiocarcinoma: Development, Growth, Spread, and Prognosis. AJR Am J Roentgenol 2017;209:W64-75. [Crossref] [PubMed]
2. Hu LS, Zhang XF, Weiss M, et al. Recurrence Patterns and Timing Courses Following Curative-Intent Resection for Intrahepatic Cholangiocarcinoma. Ann Surg Oncol 2019;26:2549-57. [Crossref] [PubMed]
3. Lee AJ, Chun YS. Intrahepatic cholangiocarcinoma: the AJCC/UICC 8th edition updates. Chin Clin Oncol 2018;7:52.
4. Sirica AE. The role of cancer-associated myofibroblasts in intrahepatic cholangiocarcinoma. Nat Rev Gastroenterol Hepatol 2011;9:44-54. [Crossref] [PubMed]
5. Gentilini A, Pastore M, Marra F, et al. The Role of Stroma in Cholangiocarcinoma: The Intriguing Interplay between Fibroblastic Component, Immune Cell Subsets and Tumor Epithelium. Int J Mol Sci 2018;19:2885. [Crossref] [PubMed]
6. Kajiyama K, Maeda T, Takenaka K, et al. The significance of stromal desmoplasia in intrahepatic cholangiocarcinoma: a special reference of 'scirrhous-type' and 'nonscirrhous-type' growth. Am J Surg Pathol 1999;23:892-902. [Crossref] [PubMed]
7. Kovač JD, Janković A, Đikić-Rom A, et al. Imaging Spectrum of Intrahepatic Mass-Forming Cholangiocarcinoma and Its Mimickers: How to Differentiate Them Using MRI. Curr Oncol 2022;29:698-723. [Crossref] [PubMed]
8. Koh J, Chung YE, Nahm JH, et al. Intrahepatic mass-forming cholangiocarcinoma: prognostic value of preoperative gadoxetic acid-enhanced MRI. Eur Radiol 2016;26:407-16. [Crossref] [PubMed]
9. Gabata T, Matsui O, Kadoya M, et al. Delayed MR imaging of the liver: correlation of delayed enhancement of hepatic tumors and pathologic appearance. Abdom Imaging 1998;23:309-13. [Crossref] [PubMed]
10. Ariizumi S, Kotera Y, Takahashi Y, et al. Mass-forming intrahepatic cholangiocarcinoma with marked enhancement on arterial-phase computed tomography reflects favorable surgical outcomes. J Surg Oncol 2011;104:130-9. [Crossref] [PubMed]
11. Park HM, Jang HY, Lee DE, et al. Prognostic impact of tumor vascularity on CT in resectable intrahepatic cholangiocarcinoma. HPB (Oxford) 2022;24:359-69. [Crossref] [PubMed]
12. Min JH, Kim YK, Choi SY, et al. Intrahepatic Mass-forming Cholangiocarcinoma: Arterial Enhancement Patterns at MRI and Prognosis. Radiology 2019;290:691-9. [Crossref] [PubMed]
13. Asayama Y, Yoshimitsu K, Irie H, et al. Delayed-phase dynamic CT enhancement as a prognostic factor for mass-forming intrahepatic cholangiocarcinoma. Radiology 2006;238:150-5. [Crossref] [PubMed]
14. Jin KP, Sheng RF, Yang C, et al. Combined arterial and delayed enhancement patterns of MRI assist in prognostic prediction for intrahepatic mass-forming cholangiocarcinoma (IMCC). Abdom Radiol (NY) 2022;47:640-50. [Crossref] [PubMed]
15. Kang Y, Lee JM, Kim SH, et al. Intrahepatic mass-forming cholangiocarcinoma: enhancement patterns on gadoxetic acid-enhanced MR images. Radiology 2012;264:751-60. [Crossref] [PubMed]
16. Wáng YXJ. Living tissue intravoxel incoherent motion (IVIM) diffusion MR analysis without b=0 image: an example for liver fibrosis evaluation. Quant Imaging Med Surg 2019;9:127-33. [Crossref] [PubMed]
17. Xiao BH, Huang H, Wang LF, et al. Diffusion MRI Derived per Area Vessel Density as a Surrogate Biomarker for Detecting Viral Hepatitis B-Induced Liver Fibrosis: A Proof-of-Concept Study. SLAS Technol 2020;25:474-83. [Crossref] [PubMed]
18. Ma FZ, Xiao BH, Wáng YXJ. MRI signal simulation of liver DDVD (diffusion derived 'vessel density') with multiple compartments diffusion model. Quant Imaging Med Surg 2025;15:1710-8. [Crossref] [PubMed]
19. Xu FY, Xiao BH, Wáng YXJ. The rationale for proposing a magnetic resonance slow diffusion metric and its proof-of-concept testing showing spleen parenchyma and hepatocellular carcinoma have faster diffusion than liver parenchyma. Quant Imaging Med Surg 2025;15:3792-806. [Crossref] [PubMed]
20. Zheng GP, Xu FY, Luo JQ, et al. The impacts of time of echo and time of repetition on slow diffusion coefficient (SDC) measure and the potential age- and gender-related variations in liver and spleen SDC values: studies of healthy volunteers. Quant Imaging Med Surg 2026;16:410. [Crossref] [PubMed]
21. Wáng YXJ, Zheng CJ, Xu XY, et al. Diffusion-derived ‘vessel density’ (DDVD) and slow diffusion coefficient (SDC) signal features of mass-forming intrahepatic cholangiocarcinoma compared to those of hepatocellular carcinoma and liver metastases: preliminary results. Rofo 2026;198:914-22. [Crossref] [PubMed]
22. Budau KL, Sigel CS, Bergmann L, et al. Prognostic Impact of Tumor Budding in Intrahepatic Cholangiocellular Carcinoma. J Cancer 2022;13:2457-71. [Crossref] [PubMed]
23. Wu G, Chen X, Luo R, et al. Histopathologic Grading of Residual Tumor Predicts Survival of Intrahepatic Cholangiocarcinoma Patients Treated With Neoadjuvant Therapy: Major Pathologic Response and Its Clinical Significance. Am J Surg Pathol 2025;49:578-87. [Crossref] [PubMed]
24. Gao S, Sun W, Zhang Y, et al. Correlation analysis of MR elastography and Ki-67 expression in intrahepatic cholangiocarcinoma. Insights Imaging 2023;14:204. [Crossref] [PubMed]
25. Li CY, Zheng CJ, Duan XH, et al. Separation of liver focal nodular hyperplasia (FNH) and liver malignant tumors by a combination of T2-weighted imaging signal and three diffusion magnetic resonance metrics of diffusion-derived vessel density, slow diffusion coefficient, and apparent diffusion coefficient. J Gastrointest Oncol 2026;17:20. [Crossref] [PubMed]
26. Xu XY, Zeng WT, Sun MH, et al. Reliable diagnosis of liver focal nodular hyperplasia by a combination of T2 weighted signal and three diffusion magnetic resonance metrics of diffusion-derived 'vessel density', slow diffusion coefficient, and apparent diffusion coefficient: a validation study of two centers' data. Quant Imaging Med Surg 2026;16:80. [Crossref] [PubMed]
27. Lewis S, Besa C, Wagner M, et al. Prediction of the histopathologic findings of intrahepatic cholangiocarcinoma: qualitative and quantitative assessment of diffusion-weighted imaging. Eur Radiol 2018;28:2047-57. [Crossref] [PubMed]
28. Sheng R, Sun W, Huang X, et al. Apparent Diffusion Coefficient MRI Shows Association With Early Progression of Unresectable Intrahepatic Cholangiocarcinoma With Combined Targeted-Immunotherapy. J Magn Reson Imaging 2023;57:275-84. [Crossref] [PubMed]
29. Yamada S, Morine Y, Imura S, et al. Prognostic prediction of apparent diffusion coefficient obtained by diffusion-weighted MRI in mass-forming intrahepatic cholangiocarcinoma. J Hepatobiliary Pancreat Sci 2020;27:388-95. [Crossref] [PubMed]
30. Zhou Y, Wang X, Xu C, et al. Mass-forming intrahepatic cholangiocarcinoma: Can diffusion-weighted imaging predict microvascular invasion? J Magn Reson Imaging 2019;50:315-24. [Crossref] [PubMed]
31. Zhou Y, Zhou G, Gao X, et al. Apparent diffusion coefficient value of mass-forming intrahepatic cholangiocarcinoma: a potential imaging biomarker for prediction of lymph node metastasis. Abdom Radiol (NY) 2020;45:3109-18. [Crossref] [PubMed]
32. Lee J, Kim SH, Kang TW, et al. Mass-forming Intrahepatic Cholangiocarcinoma: Diffusion-weighted Imaging as a Preoperative Prognostic Marker. Radiology 2016;281:119-28. [Crossref] [PubMed]
33. Wáng YXJ, Ma FZ. A tri-phasic relationship between T2 relaxation time and magnetic resonance imaging (MRI)-derived apparent diffusion coefficient (ADC). Quant Imaging Med Surg 2023;13:8873-80. [Crossref] [PubMed]
34. Wáng YXJ. An explanation for the triphasic dependency of apparent diffusion coefficient (ADC) on T2 relaxation time: the multiple T2 compartments model. Quant Imaging Med Surg 2025;15:3779-91. [Crossref] [PubMed]
35. Wáng YXJ, Wang X, Wu P, et al. Topics on quantitative liver magnetic resonance imaging. Quant Imaging Med Surg 2019;9:1840-90. [Crossref] [PubMed]
36. Xu FY, Hu GW, Zheng CJ, et al. Increased Brownian Motion of Water Molecules in Livers With Advanced Fibrosis: Preclinical and Clinical Observations With Magnetic Resonance Imaging (MRI). Mol Imaging 2026; [Crossref] [PubMed]
37. Zheng WY, Shi YT, He YY, et al. A novel MRI diffusion metric 'slow diffusion coefficient' (SDC) for diagnosing isocitrate dehydrogenase (IDH) genotype in diffuse gliomas: initial promising results. BMC Med Imaging 2025;25:427. [Crossref] [PubMed]
38. Zheng WY, Shi YT, He YY, et al. Improvement of magnetic resonance imaging diffusion metric 'slow diffusion coefficient (SDC)' over apparent diffusion coefficient (ADC) for isocitrate dehydrogenase (IDH) genotyping in diffuse gliomas. Quant Imaging Med Surg 2026;16:107. [Crossref] [PubMed]
39. Yang NJ, Zeng WT, Xiang HY, et al. Classification of hepatocellular carcinoma Ki-67 expression status by a simple combination of magnetic resonance slow diffusion coefficient (SDC) and apparent diffusion coefficient (ADC): initial promising results. Quant Imaging Med Surg 2026;16:500. [Crossref] [PubMed]
40. Tamada T, Ito K, Sone T, et al. Dynamic contrast-enhanced magnetic resonance imaging of abdominal solid organ and major vessel: comparison of enhancement effect between Gd-EOB-DTPA and Gd-DTPA. J Magn Reson Imaging 2009;29:636-40. [Crossref] [PubMed]
41. Davenport MS, Viglianti BL, Al-Hawary MM, et al. Comparison of acute transient dyspnea after intravenous administration of gadoxetate disodium and gadobenate dimeglumine: effect on arterial phase image quality. Radiology 2013;266:452-61. [Crossref] [PubMed]
42. Sun MH, Li CY, Xu XY, et al. The role of number-of-excitations (NEX) for the usability of free-breathing acquired diffusion-derived ‘vessel density’ (DDVD) and slow diffusion coefficient (SDC) maps in separating liver cysts, hemangiomas, and solid masses with a focus on small lesions. J Gastrointest Oncol 2026;17:168. [Crossref]
43. Yoshizako T, Yoshida R, Asou H, et al. Comparison between turbo spin-echo and echo planar diffusion-weighted imaging of the female pelvis with 3T MRI. Acta Radiol Open 2021;10:2058460121994737. [Crossref] [PubMed]
44. Mori N, Mugikura S, Miyashita M, et al. Turbo Spin-echo Diffusion-weighted Imaging Compared with Single-shot Echo-planar Diffusion-weighted Imaging: Image Quality and Diagnostic Performance When Differentiating between Ductal Carcinoma in situ and Invasive Ductal Carcinoma. Magn Reson Med Sci 2021;20:60-8. [Crossref] [PubMed]
45. Liu Y, Hu L, Qian L, et al. Intravoxel incoherent motion diffusion-weighted imaging in nasopharyngeal carcinoma: comparison of the turbo spin-echo and echo-planar imaging techniques. Quant Imaging Med Surg 2024;14:9101-11. [Crossref] [PubMed]
46. Chevallier O, Zhou N, He J, et al. Removal of evidential motion-contaminated and poorly fitted image data improves IVIM diffusion MRI parameter scan-rescan reproducibility. Acta Radiol 2018;59:1157-67. [Crossref] [PubMed]