Clinical utility of mono-exponential model diffusion weighted imaging using two b-values compared to the bi- or stretched exponential model for the diagnosis of biliary atresia in infant liver MRI.

PURPOSE: To investigate the clinical utility of mono-exponential model diffusion weighted imaging (DWI) using two b-values compared to the bi- or stretched exponential model to differentiate biliary atresia (BA) from non-BA in pediatric liver magnetic resonance imaging (MRI).
METHODS: Patients who underwent liver MRI with DWI for suspected BA from November 2017 to September 2018 were retrospectively included and divided into BA and non-BA groups. Laboratory results including γ-glutamyl transferase (γGT) were compared between the two groups using the Mann-Whitney U test and Fisher's exact test. The hepatic apparent diffusion coefficient (ADC) 10 using ten b-values and ADC 2 using two b-values were obtained from the mono-exponential model. The slow diffusion coefficient (D), fast diffusion coefficient (D*), and perfusion fraction (f) were obtained from the bi-exponential model. The distributed diffusion coefficient (DDC) and heterogeneity index (α) were measured from the stretched exponential model. Parameters were compared between the two groups using a linear mixed model and diagnostic performance was assessed using the area under the curve (AUC) analysis.
RESULTS: For 12 patients in the BA and five patients in the non-BA group, the ADC 10 (median 0.985 ×10-3 mm2/s vs. 1.332 ×10-3 mm2/s, p = 0.008), ADC 2 (median 0.987 ×10-3 mm2/s vs. 1.335 ×10-3 mm2/s, p = 0.017), D* (median 33.2 ×10-3 mm2/s vs. 55.3 ×10-3 mm2/s, p = 0.021), f (median 13.4%, vs. 22.1%, p = 0.009), and DDC (median 0.889 ×10-3 mm2/s vs. 1.323 ×10-3 mm2/s, p = 0.009) values were lower and the γGT (median 368.0 IU/L vs. 93.5 IU/L, p = 0.02) and α (median 0.699 vs. 0.556, p = 0.023) values were higher in the BA group. The AUC values for γGT (AUC 0.867 95% confidence interval [CI] 0.616-0.984), ADC 10 (AUC 0.963, 95% CI 0.834-0.998), ADC 2 (AUC 0.925, 95% CI 0.781-0.987), f (AUC 0.850, 95% CI 0.686-0.949), and DDC (AUC 0.925, 95% CI 0.781-0.987) were not significantly different, except for the D* and α values.
CONCLUSION: Patients with BA had lower ADC 10, ADC 2, D*, f, and DDC values and higher γGT and α values than those in the non-BA group. The diagnostic performance of ADC 2 using only two b-values showed excellent diagnostic performance and was not significantly different from that of γGT, ADC 10, f, and DDC for diagnosing BA.

Introduction

Biliary atresia (BA) is a rare but fatal cholestatic disease of infants caused by fibro-obliteration of the biliary tree [1]. Early diagnosis and intervention with Kasai portoenterostomy are essential to prevent the progression of liver fibrosis[2]. Delayed diagnosis or complications from intractable cholangitis and portal hypertension could necessitate liver transplantation in children [1,3]; therefore, early and accurate diagnosis of BA is important. For the diagnosis of BA, gray-scale ultrasonography (US), shear wave elastography, hepatobiliary scan, and magnetic resonance imaging (MRI) can be used in addition to the clinical and laboratory tests for neonates presenting with jaundice [2,4]. Even though there are fewer MRI studies compared to US, MRI has proven advantages over US due to its operator-independency, reproducibility and unrestricted field-of-view and it permits visualization of bile ducts and periportal fibrotic masses in BA patients [4-7].

Another advantage of MRI is its quantitative imaging functions. Diffusion weighted imaging (DWI) is a widely used technique; it measures the degree of diffusion of water molecules using a mono-exponential model (MEM) [8]. Separation of pure water molecular diffusion and microvascular perfusion from MEM became possible by adopting multiple b-values, which is now known as the bi-exponential model (BEM) or intravoxel incoherent motion (IVIM) technique [9]. Recently, the stretched exponential model (SEM) was introduced and takes into account the heterogeneous nature of diffusion in different tissues by measuring signal attenuation deviation from mono-exponential values [10].

In previous studies, the apparent diffusion coefficient (ADC) values from MEM were significantly lower in the liver of BA patients compared to patients with neonatal hepatitis [11-13]. A previous study demonstrated that ADC could be a new imaging parameter for assessing liver fibrosis in BA patients [12]. In children with nonalcoholic fatty liver disease, molecular diffusion and perfusion parameters from BEM were differently affected by hepatic steatosis and fibrosis [14]. In a recent study, SEM showed better diagnostic performance for diagnosing hepatic fibrosis in adults [15]. However, to our knowledge, there are no studies investigating the utility of BEM and SEM to diagnose BA in children. In addition, no study has compared the diagnostic performance of various DWI sequences using different models and b-values to reduce the acquisition time of MRI in young infants.

Therefore, the purpose of this study was to assess the clinical utility of mono-exponential model DWI using two b-values compared to the bi- or stretched exponential model to differentiate BA from non-BA in pediatric liver MRI.

Material and methods

Subjects

This retrospective study was approved by the Institutional Review Board of Severance Hospital (Protocol no. 1-2018-0076) and informed consent was waved. Children who underwent liver MRI for suspected BA due to hyperbilirubinemia from November 2017 to September 2018 were included. We excluded children who had undergone liver MRI DWI sequences with different compositions of b values (DWI MRI without using b-values of 0, 25, 50, 75, 100, 150, 200, 400, 600, and 800 s/mm2) or who had other hepatic lesions on MRI.

Subjects were divided into two groups according to their operative procedure and pathologic results: BA and non-BA-groups. Diagnosis of BA or non-BA was made through intraoperative cholangiography or liver biopsy. Age in weeks, gender, and laboratory results including aspartate aminotransferase (AST, IU/L), alanine transaminase (ALT, IU/L), total bilirubin (mg/dl), direct bilirubin (mg/dl), alkaline phosphatase (ALP, IU/L), and γ-glutamyl transferase (γGT, IU/L) were assessed within 3 days from the time of MRI examination.

MRI acquisition

Liver MRI was performed with a 1.5 T system (Achieva dStream; Philips Healthcare, Best, the Netherlands) in a pediatric body coil, all of which were performed prior to procedures such as biopsy or cholangiography. MR examinations were performed under sedation administered by a trained pediatric sedation team, and the patients were in free-breathing status. Free-breathing DWI was performed using 10 b-values (0, 25, 50, 75, 100, 150, 200, 400, 600, and 800 s/mm2) with single-shot spin-echo echo-planar imaging (SE EPI) using gradient reversal fat suppression. The used MRI parameters for DWI sequence was as follows; repetition time (TR) 4000 msec, echo time (TE) 90 msec, matrix 128×128, slice thickness 3 mm, flip angle 90°, and number of signal averages 3. The total acquisition time of DWI was 5 minutes 16 seconds.

Diffusion parameters analysis

From the DWI sequence, ADC values can be calculated using the following equation for MEM [8,16]: where S represents the degree of signal attenuation, S0 means signal intensity of the T2-weighted image with no diffusion gradient applied, and b value means degree of diffusion weighting.

Using BEM, D (slow true diffusion from pure water molecular diffusion), D* (fast pseudo-diffusion from microcirculation and perfusion), and f (perfusion fraction) values were obtained using the following equation [17]:

For SEM, heterogeneity of diffusion was assessed by measuring the deviation of diffusion from the mono-exponential behavior using the following equation [10,15]: The distributed diffusion coefficient (DDC) means intravoxel diffusion rate in the presence of heterogeneity of diffusion without separating proton pools into the compartments. The diffusion heterogeneity index (α) demonstrates the degree of deviation of diffusion signal intensities from mono-exponential curve and ranges from 0 to 1 [10]. When the α value is close to 1, the equation follows a mono-exponential curve, and it demonstrates homogeneous diffusion compositions of the proton pools in environment. When the α value is close to 0, it means high intravoxel diffusion heterogeneity [10].

To acquire parametric maps and parameters, we used EXPRESS software version 2.0 (Philips Healthcare, Andover, MA, USA). By loading raw data of DWI sequences in par/rec file formats in this software, multi-parametric maps including ADC 10 using all of the 10 b-values and ADC 2 using two b-values (0, 800 s/mm2), BEM or IVIM with asymptotic fitting (calculated D, then fit D* and f), and SEM were automatically presented. To maintain uniformity, one pediatric radiologist (H.J.S.) who was blind to the final diagnosis selected two representative axial images at the main portal vein level of each patient, and plotted the freehand region-of-interests (ROIs) to cover as much liver parenchyma as possible on the images while avoiding major vessels (Fig 1). Parameters including ADC 10, ADC 2, D, D*, f, DDC, and α values were automatically calculated from each ROI.

Fig 1

Liver diffusion MRI images of a 5-week-old girl in the BA group.

Her initial total/direct bilirubin levels were 5.7/4.7 mg/dl, and her γGT level was 403 IU/L. (A) The ADC with two b-values was 0.939 ×10−3 mm2/s. (B) The D* value was 22.6 ×10−3 mm2/s. (C) The f value was 8%. (D) The DDC value was 0.839 ×10−3 mm2/s. (E) The α value was 0.757. (F) Multi-parametric curves from the mono-, bi-, and stretched exponential models with the x-axis representing b-values and the y-axis representing diffusion-related signal attenuation.

Statistical analysis

Statistical analyses were performed using SPSS version 23 (IBM Corp., Armonk, NY, United States) and MedCalc version 18.2.1 (Ostend, Belgium). Clinical and laboratory results were compared between BA and non-BA groups using Mann-Whitney U and Fisher’s exact tests as appropriate. The area of ROIs (mm2) and diffusion parameters were compared using a linear mixed model (LMM) because two measurements were repeated using two axial images for each patient. The BA diagnostic performances were assessed and compared using area under the curve (AUC) analysis. The optimal cutoff value of each parameter was chosen to maximize the sum of sensitivity and specificity in AUC analysis. To compare the receiver operating characteristic (ROC) curves, the DeLong method was used with the MedCalc program. In addition, the logistic regression test and AUC analysis were performed to analyze diagnostic performances combining ADC 2 and γGT. All data are presented as median values and interquartile ranges. Estimated average values with 95% confidence intervals (CIs) are presented when using LMM. P-values less than 0.05 were considered statistically significant.

Results

Subjects and comparison of clinical and laboratory results

During the study period, a total of 19 children suspected of BA underwent liver MRI. Among them, one child was excluded due to a DWI sequence with different composition of b-values, and one child was excluded because he had multiple liver hemangiomas. Therefore, seventeen patients (M:F = 8:9, mean age 8.8 weeks old, range of 5–16 weeks) were included in this study. Among them, 12 of 17 patients (70.6%) were assigned to the BA group, while five patients (29.4%) were included in the non-BA group because they were confirmed to have neonatal hepatitis (n = 2), total parenteral nutrition induced cholestasis (n = 1), alagille syndrome (n = 1), and Dubin-Johnson syndrome (n = 1). The clinical and laboratory results for the BA and non-BA groups are shown in Table 1. Age and gender were not significantly different between the two groups (p = 1.000, 0.131, respectively). Among laboratory results, only the γGT level was significantly higher in the BA compared to the non-BA group (median 368.0 IU/L vs. 93.5 IU/L, p = 0.020).

Table 1

Comparison of clinical and laboratory results between BA and non-BA groups.

	BA (n = 12)	non-BA (n = 5)	p-value
Age (weeks)	8 (6, 12)	8 (5, 11)	1.000
Gender (M:F)	4:8	4:1	0.131’
AST (IU/L)	168.5 (112.8, 374.3)	72.5 (35.0, 148.3)	0.114
ALT (IU/L)	101.5 (53.8, 173.5)	27.5 (15.8, 103.0)	0.206
Total bilirubin (mg/dl)	8.0 (5.7, 9.1)	5.75 (3.8, 11.4)	0.527
Direct bilirubin (mg/dl)	6.0 (4.6, 6.9)	3.20 (3.0, 7.8)	0.291
ALP (IU/L)	649.5 (465.8, 795.3)	616.5 (354.5, 1006.0)	0.752
γGT (IU/L)	368.0 (185.8, 545.0)	93.5 (86.5, 136.5)	0.020a

Abbreviations: BA = Biliary atresia, AST = Aspartate aminotransferase, ALT = Alanine transaminase, ALP = Alkaline phosphatase, γGT = γ-glutamyl transferase.

Values are presented as median (interquartile ranges).

P-values by Mann-Whitney U test (‘Fisher’s exact test).

a P-value < 0.05.

Comparison of diffusion parameters

Table 2 shows the results of diffusion parameters for the BA and non-BA groups calculated from MEM, BEM, and SEM. The ROI area was not significantly different between the two groups (median 1215.5 mm2 vs. 1256.2 mm2, p = 0.810). From MEM, the ADC 10 (median 0.985 ×10−3 mm2/s vs. 1.332 ×10−3 mm2/s, p = 0.008) and the ADC 2 (median 0.987 ×10−3 mm2/s vs. 1.335 ×10−3 mm2/s, p = 0.017) were significantly lower in the BA group compared to the non-BA group. From BEM, D* (median 33.2 ×10−3 mm2/s vs. 55.3 ×10−3 mm2/s, p = 0.021) and f values (median 13.4%, vs. 22.1%, p = 0.009) were significantly lower in the BA group. However, the D value, which represented true water molecular diffusion, was not significantly different between the two groups (0.835 ×10−3 mm2/s vs. 1.025 ×10−3 mm2/s, p = 0.115). From SEM, DDC was significantly lower in the BA group (median 0.889 ×10−3 mm2/s vs. 1.323 ×10−3 mm2/s, p = 0.009), and the α value was significantly higher in the BA group (median 0.699 vs. 0.556, p = 0.023).

Table 2

Comparison of MRI parameters between BA and non-BA groups.

	BA (n = 12)	non-BA (n = 5)	p-value
Area of ROIs (mm2)	1215.5 (1023.5–1407.5)	1256.2 (958.7–1553.7)	0.810
ADC 10 (10−3 mm2/s)	0.985 (0.854–1.117)	1.332 (1.128–1.535)	0.008 a
ADC 2 (10−3 mm2/s)	0.987 (0.837–1.136)	1.335 (1.104–1.567)	0.017 a
D (10−3 mm2/s)	0.835 (0.703–0.966)	1.025 (0.821–1.228)	0.115
D* (10−3 mm2/s)	33.2 (23.1–43.3)	55.3 (39.7–71.0)	0.021 a
f (%)	13.4 (10.1–16.6)	22.1 (17.1–27.2)	0.007 a
DDC (10−3 mm2/s)	0.889 (0.721–1.057)	1.323 (1.062–1.584)	0.009 a
α	0.699 (0.634–0.765)	0.556 (0.454–0.658)	0.023 a

Abbreviations: BA = Biliary atresia, ROI = Region-of-interest, ADC = Apparent diffusion coefficient, D = Slow true diffusion from pure water molecular diffusion, D* = Fast pseudo-diffusion from microcirculation and perfusion, f = Perfusion fraction, DDC = Distributed diffusion coefficient, α = Heterogeneity index.

Values are presented as estimated average values with 95% confidence intervals based on a linear mixed model.

a P-value < 0.05.

Assessment and comparison of diagnostic performances

The diagnostic performances of significant results in laboratory and diffusion parameters are summarized in Table 3. The γGT value over 188 IU/L showed an AUC value of 0.867 (95% CI 0.616–0.984) for diagnosing BA. The ADC 10 value of ≤ 1.158 ×10−3 mm2/s showed an AUC value of 0.963, while The ADC 2 value of ≤ 1.165 ×10−3 mm2/s showed an AUC value of 0.925. In BEM, a D* value of ≤ 28.6 ×10−3 mm2/s showed an AUC value of 0.771, and an f value of ≤ 14.3% showed an AUC value of 0.850. In SEM, a DDC value of ≤ 1.256 ×10−3 mm2/s showed an AUC value of 0.925, while an α value over 0.680 had an AUC value of 0.788.

Table 3

Diagnostic performance of parameters for the differentiation of BA from non-BA.

	Cutoff values	Sensitivity (%)	Specificity (%)	AUC (95% CI)
γGT (IU/L)	> 188	75 (42.8–94.5)	100 (47.8–100)	0.867 (0.616–0.984)
ADC 10 (10−3 mm2/s)	≤ 1.158	83.3 (62.6–95.3)	100 (69.2–100)	0.963 (0.834–0.998)
ADC 2 (10−3 mm2/s)	≤ 1.165	79.2 (57.8–92.9)	100 (69.2–100)	0.925 (0.781–0.987)
D* (10−3 mm2/s)	≤ 28.6	58.3 (36.6–77.9)	90 (55.5–99.7)	0.771 (0.595–0.897)
f (%)	≤ 14.3	66.7 (44.7–84.4)	100 (69.2–100)	0.850 (0.686–0.949)
DDC (10−3 mm2/s)	≤ 1.256	95.8 (78.9–99.9)	80 (44.4–97.5)	0.925 (0.781–0.987)
α	> 0.68	62.5 (40.6–81.2)	90.0 (55.5–99.7)	0.788 (0.614–0.908)

Abbreviations: BA = Biliary atresia, AUC = Area-under-the-curve, CI = confidence interval, γGT = γ-glutamyl transferase, ADC = Apparent diffusion coefficient, D* = Fast pseudo-diffusion from microcirculation and perfusion, f = Perfusion fraction, DDC = Distributed diffusion coefficient, α = Heterogeneity index.

When comparing ROC curves, ADC 10 showed higher diagnostic performances compared with D* and α (p = 0.025 and 0.033, respectively), while all other parameters showed no significant differences for diagnosing BA (Table 4). The diagnostic performance of ADC 2 using only two b-values showed excellent diagnostic performance and was not significantly different from that of γGT, ADC 10, f, and DDC for diagnosing BA. When we obtain the diagnostic performance of combining ADC 2 and γGT together, the AUC value was 0.987 with a 95% CI of 0.960–1.000.

Table 4

P-values comparing ROC curves for the differentiation of BA from non-BA.

	ADC 10	ADC 2	D*	f	DDC	α
γGT	0.125	0.380	0.340	0.847	0.389	0.396
ADC 10	.	0.195	0.025 a	0.073	0.181	0.033 a
ADC 2	.	.	0.083	0.286	1.000	0.130
D*	.		.	0.416	0.075	0.859
f	.	.		.	0.280	0.378
DDC	.	.	.		.	0.125

Abbreviations: BA = Biliary atresia, ROC = Receiver operating characteristic, γGT = γ-glutamyl transferase, ADC = Apparent diffusion coefficient, D* = Fast pseudo-diffusion from microcirculation and perfusion, f = Perfusion fraction, DDC = Distributed diffusion coefficient, α = Heterogeneity index.

a P-value < 0.05.

Discussion

In this study, only the γGT value (>188 IU/L) was a significant clinical and laboratory finding that differentiated BA from non-BA groups. However, most of the diffusion parameters including ADC 10, ADC 2, D*, f, and DDC values were significantly lower and the α value was significantly higher in the livers of those in the BA group compared to those in the non-BA group. Only the D value showed no significant difference suggesting that pure water molecular diffusion and vascular perfusion affected liver diffusion differently in the BA and non-BA groups. The diagnostic performances of most of the significant parameters were good to excellent, with the exception of the D* and α values. In addition, diagnostic performance of ADC 2 with b-values of 0 and 800 s/mm2 was not significantly different from that of other parameters including ADC 10, f, DDC, and γGT; this shows that an ADC value with two b-values can be used to differentiate BA and non-BA while reducing image acquisition time in diffusion MRI of pediatric livers.

There are very few studies that have investigated DWI applied to liver MRIs for BA patients. In 2011 and 2015, two studies demonstrated that hepatic ADC values were significantly lower in children with BA compared to those with normal livers and that there was a negative correlation with the degree of liver fibrosis in BA patients [11,12]. Peng et al. mentioned that ADC values could also be used to predict the degree of liver fibrosis in postoperative patients [12]. In 2016, Liu et al. first utilized diffusion tensor imaging (DTI) to differentiate BA from non-BA and demonstrated that fractional anisotropy from DTI showed no significant differences between the two groups, while ADC values with 0 and 1000 s/mm2 had significantly lower values in BA patients [13]. The diagnostic performance of ADC value was demonstrated by an AUC value of 0.805 with a cutoff value of 1.317 ×10−3 mm2/s with a sensitivity of 75% and specificity of 82% [13]. To our knowledge, our study is the first attempt to utilize MEM, BEM, and SEM DWI to differentiate BA from non-BA. Our results are consistent with other studies that showed a lower ADC value in the BA group; however we used a control group with non-BA patients instead of normal children as was done in other studies. In addition, our diagnostic performances using these three models were higher than those in previous study.

BEM studies in adults with liver cirrhosis showed that either D* or f was significantly lower in cirrhotic livers than in healthy livers, while D showed frequent trends for no significant difference or poor correlation with fibrosis level [18,19]. One recent study on pediatric liver BEM, D*, and f values were significantly decreased in liver fibrosis [20]. We can suggest that decreased values in perfusion related parameters of BA patients could be from portal hypertension and decreased portal perfusion from increased collagen fibers and activated stellate cells in fibrotic livers [19]. Decreased microperfusion parameters from BEM were also noted in BA patients after receiving Kasai operation [20]. However, BA was known to have hepatic arteriopathy and increased hepatic arterial flow on color Doppler US could be used for the diagnosis of BA [21]. Effects of decreased portal perfusion, hepatic arteriopathy, and cholestasis in the BA patients before receiving operation were not fully understood yet. Therefore, further studies to know the reason for decreased perfusion parameters in BA patients are needed with pathologic correlation.

One recent study utilized SEM to assess liver fibrosis in adults [15]. They demonstrated that D*, f, DDC, and α values were significantly decreased as the degree of liver fibrosis increased, while ADC and D did not show any differences [15]. Our results correspond with this study showing significantly lower D*, f and DDC values in the BA group. One of the interesting points in our study was that the α value was significantly higher in the BA group compared to the non-BA group. This might be because of differences between our study’s samples compared to the other study. We could suggest that increased α values in BA group might be from relative homogeneity in decreased hepatic diffusion in the BA group compared to heterogeneous diffusion tendencies in the non-BA group, while higher α value represents decreased heterogeneity of diffusion rates. Our study showed the clinical potential of α value as a new quantitative MRI parameter for assessing diffusion heterogeneity. Because of the lack of studies applying SEM in pediatric liver and for BA patients, more studies are needed to validate this finding.

The D* showed relatively lower diagnostic performance, which was in accord with other studies that showed larger variation in D* measurement among BEM parameters [15,19,22]. The ADC 2 with two b-values showed excellent diagnostic performance and the diagnostic performance was not significantly different from that of other significant parameters in this study, suggesting that image acquisition time for obtaining MEM in pediatric liver can be reduced. In addition, this study demonstrated the potential utility of a new objective imaging parameter using MRI for the diagnosis of BA. Although γGT levels can be used to diagnose BA, imaging studies are usually recommended for the evaluation of infants with cholestasis due to overlaps in the laboratory results from other diseases [23]. In addition, MRI is more objective than US, which is a great advantage and research has emphasized the additive role of MRI for reducing the false-negative and false-positive results of US [4,5]. In addition, MRI can be used not only for morphologic evaluation, but to assess fibrosis in BA patients using DWI [12]. Usually the multimodality approach including laboratory and imaging studies is needed to accurately diagnose BA. Therefore, this study showed the potential utility of ADC and different DWI models as objective and quantitative imaging markers for the diagnosis of BA in conjunction with conventional laboratory results. Because this was a preliminary study with a small number of subjects, further investigation dealing with the additive role of different DWI models to conventional diagnostic methods is needed in the future.

Our study has several limitations. First, because it was retrospective, there was small number of included patients, especially in the non-BA group. Second, we did not evaluate the diagnostic performance of parameters according to the degree of liver fibrosis in BA patients or the prognostic value of the parameters after the Kasai operation. We did not assess whether parameters had additive effects for diagnosing BA on conventional ultrasonography or MRI because we wanted to focus on demonstrating feasibility and utility of diffusion models in pediatric liver MRI and for BA patients. Further studies are needed that utilize these diffusion models to diagnose and predict prognostic values in BA patients. In addition, a multi-parametric approach that combines diffusion parameters with γGT, which showed equivalent diagnostic performance to diffusion parameters in our study, may also yield good performance and should be considered in future studies.

Conclusions

The MEM, BEM, and SEM DWI were all feasible for pediatric liver MRI. Patients in the BA group had significantly lower ADC 10, ADC 2, D*, f, and DDC values and higher γGT and α values in the liver than those in the non-BA group. The diagnostic performance of ADC 2 using only two b-values was excellent and was not significantly different from the diagnostic performances of γGT value and other diffusion parameters for diagnosing BA. Further prospective and larger studies are necessary for the clinical application of diffusion parameters in BA patients.

Anonymized data of this study.

(PDF)

Click here for additional data file.

13 Sep 2019

PONE-D-19-21097

Usefulness of mono-, bi-, and stretched exponential model diffusion weighted imaging for the differentiation of biliary atresia and non-biliary atresia in pediatric liver MRI

PLOS ONE

Dear Dr Hyun Joo Shin

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please address the issues raised by the reviewers listed below.

We would appreciate receiving your revised manuscript by Oct 28 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Gregory Tiao, M.D.

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2.  We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

* In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

* We will update your Data Availability statement on your behalf to reflect the information you provide.

Additional Editor Comments (if provided):

The authors present an interesting manuscript but need to address the issues raised by the reviewers

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: PONE-D-19-21097: Usefulness of mono-, bi-, and stretched exponential model diffusion weighted imaging for the differentiation of biliary atresia and non-biliary atresia in pediatric liver MRI

This manuscript demonstrates feasibility and utility of multi-b-value diffusion imaging in infant population for differentiation of biliary atresia and non-biliary atresia and provides further insights into the diffuse changes in liver. One important finding of the study as written in discussion is “The ADC 2 with two b-values showed excellent diagnostic performance and was comparable with other significant parameters in this study, suggesting that image acquisition time for obtaining MEM in pediatric liver can be reduced.” Performing MRI in infants is especially challenging due to time constraints imposed by the need for sedation/anesthesia. In that light this reviewer would recommend authors to keep the primary focus of the manuscript (and Title) something like “Clinical utility of two b-values compared to mono-, bi-, and stretched exponential model diffusion weighted imaging for the differentiation of biliary atresia and non-biliary atresia in infant liver MRI”.

The systematic comparison of using 2 (0 and 800) versus 10 (0, 25, 50, 75 100, 150, 200, 400, 600, 800) b values and using mono-, bi-, and stretched exponential models to derive different measures of diffusion and pseudo-diffusion is an important endeavor. As authors have described in the discussion prognostic value of the additive quantitative information from ADC mono, f, and DDC cannot be inferred from this specific study. Additionally, in routine clinical practice added value of ADC 2 (with considerations for sedation and expensive MRI) over γGT, itself may need justification.

Paper is very well written paper. Although this is a limited sample study, the findings of the study are very promising and encouraging for further more rigorous clinical validation. Authors have done a very good job of putting their contribution in this field in right perspective in the discussion session. Overall, discussion is very well written and has good balance. However, the points from the discussion do not reflect in the conclusion. The findings do not support claim of “new imaging parameters for the diagnosis of BA”. The γGT should be mentioned in the conclusion. Same applies for the Abstract of the manuscript.

Few remarks,

Title

Title may benefit with a qualifier

Clinical utility of two b-values compared to mono-, bi-, and stretched exponential model diffusion weighted imaging for the differentiation of biliary atresia and non-biliary atresia in infant liver MRI

Abstract

The values and differentiation power of should be reported. See comments above.

Introduction

A line referring to consideration for added time for extra sequences whould add to the context of the work (ADC 2 v/s ADC mono)

Materials and Methods

L 86: Consider changing different liver MRI DWI sequences to “composition of b values”

L 129: ADC “2” refers to ADC using 2 b values, thus ADC 1 may be better referred to as ADC mono or something like that.

Discussion

Overall, discussion is very well written and has good balance.

Figures.

Please include appropriate values and units for all the color bar

Reviewer #2: PONE-D-19-21097

• Very small study – should be considered a pilot investigation

• Study is innovative, with novel use of quantitative MRI parameters for detecting BA

• Results are compelling and deserving of a larger study

• Scattered corrections, additions are suggested to improve this manuscript

Abstract:

1. Define all abbreviations, please

2. Methods – what statistical test was used to compare groups – student’s t-test or Mann-Whitney U?

3. Re: Results – 1st sentence, consider providing actual results/p-values

4. Conclusion… can help in diagnosis of BA, perhaps… will certainly be some false positives and negatives

5. Keywords: change child to infant?; use “magnetic resonance imaging” and “diffusion-weighted imaging” as keywords

Introduction

6. MRI is not good for bile ducts in babies, in general… note, US actually has considerably better spatial resolution that MRI

7. When you discuss ultrasound, it may be worth separating gray-scale from shear wave elastography… both have been shown to add value in the diagnosis of BA

8. Any a priori hypothesis?

Methods

9. What is meant by different DWI sequences? Clearly list inclusion and exclusion criteria

10. Were labs really at exact time of MRI exam? Or did you use values closest to MRI?

11. Philips Healthcare MRI is out of Best, the Netherlands

12. Was DWI free breathing or respiratory-triggered/navigator gated?

13. What type of fat suppression was employed?

14. Who performed image analyses? Were they blinded to BA vs. non-BA diagnosis?

15. Why were linear mixed models needed to compare BA to non-BA pts?

Results

16. How was optimal ROC cut-off value chosen? Should you maximize sensitivity over specificity?

17. Does statistical analysis section indicate how ROC AUC values were compared?

18. Can you combine multiple MRI parameters and GGT to get even better performance (so called multi-parametric approach, use logistic regression which will generate an ROC AUC)

Discussion

19. OK

Conclusion

20. Emphasize that additional larger, prospective investigations are needed

References

21. OK

Figures

22. Fig 1F – can x-axis be labeled every 100? Is y-axis signal attenuation or signal intensity?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Review_PONE-D-19-21097_ASP.docx

Click here for additional data file.

20 Oct 2019

Responses to reviewers

Dear Editor,

Thank you for your considerate review and suggestions for the revision of our manuscript entitled “Usefulness of mono-, bi-, and stretched exponential model diffusion weighted imaging for the differentiation of biliary atresia and non-biliary atresia in pediatric liver MRI”. We have reviewed the suggestions made by the Reviewers and have done our best to revise the manuscript accordingly. Please find our responses below.

Reviewer’s Comments to Author:

Reviewer #1:

1. This manuscript demonstrates feasibility and utility of multi-b-value diffusion imaging in infant population for differentiation of biliary atresia and non-biliary atresia and provides further insights into the diffuse changes in liver. One important finding of the study as written in discussion is “The ADC 2 with two b-values showed excellent diagnostic performance and was comparable with other significant parameters in this study, suggesting that image acquisition time for obtaining MEM in pediatric liver can be reduced.” Performing MRI in infants is especially challenging due to time constraints imposed by the need for sedation/anesthesia. In that light this reviewer would recommend authors to keep the primary focus of the manuscript (and Title) something like “Clinical utility of two b-values compared to mono-, bi-, and stretched exponential model diffusion weighted imaging for the differentiation of biliary atresia and non-biliary atresia in infant liver MRI”.

� Thank you for your detailed comments. We agree with the Reviewer’s recommendations and have revised the title, purpose, and conclusion to reflect the Reviewer’s comments.

2. The systematic comparison of using 2 (0 and 800) versus 10 (0, 25, 50, 75 100, 150, 200, 400, 600, 800) b values and using mono-, bi-, and stretched exponential models to derive different measures of diffusion and pseudo-diffusion is an important endeavor. As authors have described in the discussion prognostic value of the additive quantitative information from ADC mono, f, and DDC cannot be inferred from this specific study. Additionally, in routine clinical practice added value of ADC 2 (with considerations for sedation and expensive MRI) over γGT, itself may need justification.

� When diagnosing biliary atresia, laboratory results including γGT can be used, but imaging modalities such as ultrasonography and MRI also have important roles because laboratory results of BA overlap with those of other diseases [1]. MRI is more objective compared to ultrasonography and research has emphasized the additive role of MRI for reducing the false-negative and false-positive results of ultrasonography [2,3]. In addition, not only for morphologic imaging, MRI can be used to assess the degree of fibrosis in biliary atresia patients using DWI [4]. This study showed that ADC and different DWI models have the potential to be objective and quantitative imaging markers for the diagnosis of BA in conjunction with conventional laboratory results. Because this study is a preliminary study with a small number of subjects, further investigation dealing with the additive role of different DWI models to conventional diagnostic methods is needed. We added a related explanation to the Discussion in this revision with the relevant references.

3. Paper is very well written paper. Although this is a limited sample study, the findings of the study are very promising and encouraging for further more rigorous clinical validation. Authors have done a very good job of putting their contribution in this field in right perspective in the discussion session. Overall, discussion is very well written and has good balance. However, the points from the discussion do not reflect in the conclusion. The findings do not support claim of “new imaging parameters for the diagnosis of BA”. The γGT should be mentioned in the conclusion. Same applies for the Abstract of the manuscript.

� Thank you for your encouraging comments and we have revised the Abstract, Discussion, and our conclusions to better reflect the findings of our study.

Title

4. Title may benefit with a qualifier

Clinical utility of two b-values compared to mono-, bi-, and stretched exponential model diffusion weighted imaging for the differentiation of biliary atresia and non-biliary atresia in infant liver MRI

� We agree with the Reviewer that the title can be more narrowed down and revised it accordingly.

Abstract

5. The values and differentiation power of should be reported. See comments above.

� The requested data were added to the Abstract.

Introduction

6. A line referring to consideration for added time for extra sequences whould add to the context of the work (ADC 2 v/s ADC mono)

� Thank you for your insightful comment. We added a brief mention of this advantage to the Introduction.

Materials and Methods

7. L 86: Consider changing different liver MRI DWI sequences to “composition of b values”

� We revised not only the phrase in question but the overall sentence, because the prior expression was thought to be too ambiguous.

8. L 129: ADC “2” refers to ADC using 2 b values, thus ADC 1 may be better referred to as ADC mono or something like that.

� Thank you for pointing this out as we realized that the terms themselves were causing confusion. The mono-exponential model was used to obtain ADC 1 and ADC 2 with different b-values. To avoid confusion on these two ADC values, we changed ‘ADC 1’ to ‘ADC 10’ throughout the manuscript to indicate that the ADC value was obtained from 10 b-values.

Discussion

9. Overall, discussion is very well written and has good balance.

Figures.

10. Please include appropriate values and units for all the color bar

� Thank you for your comment. The units in each map were depicted by the software and we could not change the way the color bars or the units were displayed on the maps. However, in the manuscript, we presented each parameter value in widely accepted units. Therefore, we added each unit around the color bar for better clarity. In addition, the annotation in Figure 1F was revised (ADC 1 was changed to ADC 10).

Reviewer #2:

• Very small study – should be considered a pilot investigation

• Study is innovative, with novel use of quantitative MRI parameters for detecting BA

• Results are compelling and deserving of a larger study

• Scattered corrections, additions are suggested to improve this manuscript

�Thank you for your supportive comments and hope our revisions will be found satisfactory.

Abstract:

1. Define all abbreviations, please

� We wrote out all abbreviations in the Abstract as requested.

2. Methods – what statistical test was used to compare groups – student’s t-test or Mann-Whitney U?

� The Mann-Whitney U test and Fisher’s exact test were used. We added these details to the Abstract.

3. Re: Results – 1st sentence, consider providing actual results/p-values

� We added the actual results and p-values as requested.

4. Conclusion… can help in diagnosis of BA, perhaps… will certainly be some false positives and negatives

� In light of Reviewer 1’s comments, we changed our conclusion to emphasize the diagnostic performance of ADC 2 with two b-values. DWI parameters could also lead to false-positive or false-negative results, but this preliminary study could not cover all of these parameters. Of its many traits, MRI is beneficial or advantageous because of its objectiveness and because it results in fewer false-positive or false-negative cases compared to ultrasonography [2]. Further investigation with a larger number of subjects is needed to validate the DWI models and conventional imaging modalities following this study. We added this content to the Abstract and the Discussion as well.

5. Keywords: change child to infant?; use “magnetic resonance imaging” and “diffusion-weighted imaging” as keywords

� We changed the Keywords as suggested.

Introduction

6. MRI is not good for bile ducts in babies, in general… note, US actually has considerably better spatial resolution that MRI

� We truly appreciate your comment. We agree that US has good spatial resolution in infants. However, a previous study showed that MRI is not inferior to US when assessing findings such as triangular cord thickness, visibility of common bile duct and abnormality of gallbladder, which are used to diagnose biliary atresia [3]. In addition, MRI has strengths such as operator independency, reproducibility, and unrestricted FOV, unlike US [2]. We added and revised our explanation in the Introduction to reflect these previous studies.

7. When you discuss ultrasound, it may be worth separating gray-scale from shear wave elastography… both have been shown to add value in the diagnosis of BA

� We totally agree with the Reviewer and have changed the sentence accordingly.

8. Any a priori hypothesis?

� There has been several papers on the utility of ADC using 2 b-values for the diagnosis of BA, while there are currently no studies on the use of bi- or stretched exponential models for the diagnosis of BA. Instead, there have been a few studies using the bi- or stretched exponential models for diagnosis or grading of fibrosis in children and adults. Since we thought that the ADC values obtained by using 2 b-values decrease due to fibrotic change in BA patients, we tried to investigate the diagnosis of BA using the mono, bi- and stretched exponential models. Also, if several of the model parameters had significant results, we wanted to identify the most efficient parameter. We added further explanations to the Introduction.

Methods

9. What is meant by different DWI sequences? Clearly list inclusion and exclusion criteria

� DWI MRI that were obtained without using the 10 b values mentioned in the Methods (0, 25, 50, 75, 100, 150, 200, 400, 600, and 800 s/mm2) were excluded. We revised the sentence because the prior expression was thought to be too ambiguous in light of the Reviewer’s comments.

10. Were labs really at exact time of MRI exam? Or did you use values closest to MRI?

� Lab tests were performed within 3 days of the MRI examination. We added this detail to the Materials and Methods.

11. Philips Healthcare MRI is out of Best, the Netherlands

� Thank you for pointing out this error. We immediately edited the MRI information.

12. Was DWI free breathing or respiratory-triggered/navigator gated?

� Free breathing was the chosen method and we added this detail to the Materials and Methods.

13. What type of fat suppression was employed?

� Gradient reversal fat suppression was employed and this detail was also added to the Materials and Methods.

14. Who performed image analyses? Were they blinded to BA vs. non-BA diagnosis?

� One experienced pediatric radiologist who was blinded to the final diagnosis performed the image analyses. We added this information to the Materials and Methods.

15. Why were linear mixed models needed to compare BA to non-BA pts?

� As ROIs were drawn on each of the two representative axial images of the liver separately, two values were obtained for each parameter. We used the linear mixed model to use each of the repeated measurements, instead of using median or mean values, for a more accurate analysis. We added further explanations on this to the Methods.

Results

16. How was optimal ROC cut-off value chosen? Should you maximize sensitivity over specificity?

� The optimal cutoff value was chosen to maximize q the sum of sensitivity and specificity. We added explanations to the Materials and Methods.

17. Does statistical analysis section indicate how ROC AUC values were compared?

� For the comparison of ROC curves, the DeLong method was used with the MedCalc program. We added this detail to the Statistical Analysis section.

18. Can you combine multiple MRI parameters and GGT to get even better performance (so called multi-parametric approach, use logistic regression which will generate an ROC AUC)

� Thank you for your perceptive comment. Using logistic regression, we obtained the AUC value of 0.987 (95% CI 0.960-1.000) for combined ADC 2 and γGT. We presented this result in the Materials and Methods and Results. As we could not compare the diagnostic performances of multiparametric results combining MRI parameters, laboratory tests, and even conventional imaging findings in this study, further research following this preliminary study is needed for validation. We added a statement on the necessity for such future research to the Discussion as well.

Discussion

19. OK

Conclusion

20. Emphasize that additional larger, prospective investigations are needed

� Because our study is of retrospective design and small in size, we agree that larger prospective studies are necessary before diffusion parameters can be clinically applied in pediatric liver MRI. We added these details to the Conclusion.

References

21. OK

Figures

22. Fig 1F – can x-axis be labeled every 100? Is y-axis signal attenuation or signal intensity?

� The software for analyzing each parametric map automatically generated the graph (Fig 1F) and we could not change the scales of the X-axis. The X-axis represents b-values (0, 25, 50, 75, 100, 150, 200, 400, 600, and 800 s/mm2) and the Y-axis represents diffusion-related signal attenuation. A higher number of b-values makes more pronounced diffusion-related signal attenuation. We added this information to the Figure legend.

Thank you once again for your time and efforts in reviewing our manuscript.

References

1. Sun S, Chen G, Zheng S, Xiao X, Xu M, Yu H, et al. Analysis of clinical parameters that contribute to the misdiagnosis of biliary atresia. J Pediatr Surg. 2013;48: 1490-1494. doi:10.1016/j.jpedsurg.2013.02.034 PMID:23895960

2. Kim YH, Kim MJ, Shin HJ, Yoon H, Han SJ, Koh H, et al. MRI-based decision tree model for diagnosis of biliary atresia. Eur Radiol. 2018. doi:10.1007/s00330-018-5327-0 PMID:29476221

3. Han SJ, Kim MJ, Han A, Chung KS, Yoon CS, Kim D, et al. Magnetic resonance cholangiography for the diagnosis of biliary atresia. J Pediatr Surg. 2002;37: 599-604 PMID:11912518

4. Peng SS, Jeng YM, Hsu WM, Yang JC, Ho MC. Hepatic ADC map as an adjunct to conventional abdominal MRI to evaluate hepatic fibrotic and clinical cirrhotic severity in biliary atresia patients. Eur Radiol. 2015;25: 2992-3002. doi:10.1007/s00330-015-3716-1 PMID:25921590

Submitted filename: DDC response to reviewer 20191019.docx

Click here for additional data file.

4 Dec 2019

Clinical utility of mono-exponential model diffusion weighted imaging using two b-values compared to the bi- or stretched exponential model for the diagnosis of biliary atresia in infant liver MRI

PONE-D-19-21097R1

Dear Dr. Shin,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Gregory Tiao, M.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

The authors have addressed all the concerns raised during the review

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: Thank you for addressing the reviewers' comments. I have no remaining concerns. Your study, while small, has interesting results/conclusions and should pave the way for future investigations.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

10 Dec 2019

PONE-D-19-21097R1

Clinical utility of mono-exponential model diffusion weighted imaging using two b-values compared to the bi- or stretched exponential model for the diagnosis of biliary atresia in infant liver MRI

Dear Dr. Shin:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Gregory Tiao

Academic Editor

PLOS ONE