An automated liver segmentation in liver iron concentration map using fuzzy c-means clustering combined with anatomical landmark data

Background

To estimate median liver iron concentration (LIC) calculated from magnetic resonance imaging, excluded vessels of the liver parenchyma region were defined manually. Previous works proposed the automated method for excluding vessels from the liver region. However, only user-defined liver region remained a manual process. Therefore, this work aimed to develop an automated liver region segmentation technique to automate the whole process of median LIC calculation.

Methods

553 MR examinations from 471 thalassemia major patients were used in this study. LIC maps (in mg/g dry weight) were calculated and used as the input of segmentation procedures. Anatomical landmark data were detected and used to restrict ROI. After that, the liver region was segmented using fuzzy c-means clustering and reduced segmentation errors by morphological processes. According to the clinical application, erosion with a suitable size of the structuring element was applied to reduce the segmented liver region to avoid uncertainty around the edge of the liver. The segmentation results were evaluated by comparing with manual segmentation performed by a board-certified radiologist.

Results

The proposed method was able to produce a good grade output in approximately 81% of all data. Approximately 11% of all data required an easy modification step. The rest of the output, approximately 8%, was an unsuccessful grade and required manual intervention by a user. For the evaluation matrices, percent dice similarity coefficient (%DSC) was in the range 86–92, percent Jaccard index (%JC) was 78–86, and Hausdorff distance (H) was 14–28 mm, respectively. In this study, percent false positive (%FP) and percent false negative (%FN) were applied to evaluate under- and over-segmentation that other evaluation matrices could not handle. The average of operation times could be reduced from 10 s per case using traditional method, to 1.5 s per case using our proposed method.

Conclusion

The experimental results showed that the proposed method provided an effective automated liver segmentation technique, which can be applied clinically for automated median LIC calculation in thalassemia major patients.

Liver Iron Load Influences Hepatic Fat Fraction in End-Stage Renal Disease Patients on Dialysis: A Proof of Concept Study

Background

Nonalcoholic fatty liver disease (NAFLD) is a spectrum of diseases including steatosis, nonalcoholic steatohepatitis (NASH), cirrhosis, and end-stage liver failure. Hepatic iron accumulation has been linked to hepatic fibrosis severity in NASH and NAFLD. Iron overload induced by parenteral (IV) iron therapy is a potential clinical problem in dialysis patients. We analyzed the hypothetical triggering and aggravating role of iron on NAFLD in patients on dialysis.

Methods

Liver iron concentration (LIC) and hepatic proton density fat fraction (PDFF) were analyzed prospectively in 68 dialysis patients by magnetic resonance imaging (MRI). Follow up of LIC and PDFF was performed in 17 dialysis patients during iron therapy.

Findings

PDFF differed significantly among dialysis patients classified according to LIC: patients with moderate or severe iron overload had increased fat fraction (PDFF: 7.9% (0.5–14.8%)) when compared to those with normal LIC (PDFF: 5% (0.27–11%)) or mild iron overload (PDFF: 5% (0.30–11.6%); P = 0.0049). PDFF correlated with LIC, and ferritin and body mass index. In seven patients monitored during IV iron therapy, LIC and PDFF increased concomitantly (PDFF: initial 2.5%, final 8%, P = 0.0156; LIC: initial 20 μmol/g, final 160 μmol/g: P = 0.0156), whereas in ten patients with iron overload, PDFF decreased after IV iron withdrawal or major dose reduction (initial: 8%, final: 4%; P = 0.0098) in parallel with LIC (initial: 195 μmol/g, final: 45 μmol/g; P = 0.002).

Interpretation

Liver iron load influences hepatic fat fraction in dialysis patients. Iron overload induced by iron therapy may aggravate or trigger NAFLD in dialysis patients.

Trial registration number (ISRCTN)

80100088.

Severe Liver Iron Concentrations (LIC) in 24 Patients with β-Thalassemia Major: Correlations with Serum Ferritin, Liver Enzymes and Endocrine Complications

Introduction

Chronic blood transfusion is the mainstay of care for individuals with β-thalassemia major (BTM). However, it causes iron-overload that requires monitoring and management by long-term iron chelation therapy to prevent endocrinopathies and cardiomyopathies, which can be fatal. Hepatic R2 MRI method (FerriScan®) has been validated as the gold standard for evaluation and monitoring liver iron concentration (LIC) that reflects the total body iron-overload. Although adequate oral iron chelation therapy (OIC) is promising for the treatment of transfusional iron-overload, some patients are less compliant with it, and others suffer from long-term effects of iron overload.

Objective

The aim of our study was to evaluate the prevalence of endocrinopathies and liver dysfunction, in relation to LIC and serum ferritin level, in a selected group of adolescents and young adult BTM patients with severe hepatic iron overload (LIC from 15 to 43 mg Fe/g dry weight).

Patients and Methods

Twenty-four selected BTM patients with severe LIC, due to transfusion-related iron-overload, followed at the Haematology Section, National Centre for Cancer Care and Research, Hamad Medical Corporation of Doha (Qatar), from April 2015 to July 2017, were retrospectively evaluated. The prevalence of short stature, hypogonadism, hypothyroidism, hypoparathyroidism, impaired fasting glucose (IFG), diabetes, and adrenal insufficiency was defined and assessed according to the International Network of Clinicians for Endocrinopathies in Thalassemia (ICET) and American Diabetes Association criteria.

Results

Patients’ most common transfusion frequency was every three weeks (70.8%). At the time of LIC measurements, their median age was 21.5 years with a mean age of 21.7 ± 8.0 years. Mean LIC was 32.05 ± 10.53 mg Fe/g dry weight (range: 15 to 43 mg Fe/g dry weight), and mean serum ferritin level was 4,488.6 ± 2,779 μg/L. LIC was correlated significantly with serum ferritin levels (r = 0.512; p = 0.011). The overall prevalence of short stature was 26.1% (6/23), IFG was 16.7% (4/24), sub-clinical hypothyroidism was 14.3% (3/21), hypogonadotropic hypogonadism was 14.3% (2/14), diabetes mellitus was 12.5% (3/24), and biochemical adrenal insufficiency was 6.7% (1/15). The prevalence of hepatitis C positivity was 20.8% (5/24). No case of clinical hypothyroidism, adrenal insufficiency or hypoparathyroidism was detected in this cohort of patients. The prevalence of IFG impaired fasting glucose was significantly higher in BTM patients with very high LIC (>30 mg Fe/g dry liver) versus those with lower LIC (p = 0.044). The prevalence of endocrinopathies was not significantly different between the two groups of patients with LIC above and below 15 mg Fe/g dry weight.

Conclusions

A significant number of BTM patients, with high LIC and endocrine disorders, still exist despite the recent developments of new oral iron chelating agents. Therefore, physicians’ strategies shall optimize early identification of those patients to optimise their chelation therapy and to avoid iron-induced organ damage. We believe that further studies are needed to evaluate if serial measurements of quantitative LIC may predict the risk for endocrine complications. Until these data are available, we recommend a close monitoring of endocrine and other complications, according to the international guidelines.

Quantification of liver iron concentration using the apparent susceptibility of hepatic vessels

Background

The quantification of liver iron concentration (LIC) is important for the monitoring of the body iron level in patients with iron overload. Conventionally, LIC is quantified through R2 or R2* mapping using MRI. In this paper, we demonstrate an alternative approach for LIC quantification through measuring the apparent susceptibility of hepatic vessels using quantitative susceptibility mapping (QSM).

Methods

QSM was performed in the liver region with the iterative susceptibility weighted imaging and mapping (iSWIM) algorithm, using the geometry of the vessels extracted from magnitude images as constraints. The susceptibilities of liver tissue were estimated from the apparent susceptibility of the hepatic veins and then converted to LIC. The accuracy of the proposed method was first validated using simulations, and then confirmed using in vivo data collected on 8 healthy controls and 11 patients at 3T. The effects of data acquisition parameters were studied using simulations, and the LICs estimated using QSM were compared with those estimated using R2* mapping.

Results

Simulation results showed that the use of a 3D data acquisition protocol with higher image resolution led to improved accuracy in LIC quantification using QSM. Both simulations and in vivo data results demonstrated that the LICs estimated using the proposed QSM method agreed well with those estimated using R2* mapping. With the shortest echo time being 2.5ms in the multi-echo gradient echo sequence, simulations results showed that LIC up to 12.45 mg iron/g dry tissue can be quantified using the proposed QSM method. For the in vivo data, the highest LIC measured was 11.32 mg iron/g dry tissue.

Conclusions

The proposed method offers a reliable and flexible way to quantify LIC and has the potential to extend the range of LIC that can be accurately measured using R2* and QSM.

Comparison between different software programs and post-processing techniques for the MRI quantification of liver iron concentration in thalassemia patients

PURPOSE

In magnetic resonance imaging (MRI) relaxometry, various software programs are available to perform R2* measurements and to estimate the liver iron concentration (LIC). The main objective of our study was to compare R2* LIC values, obtained with three different software programs based on specific decay models and calibration curves, with LIC estimates provided by R2-relaxometry (FerriScan).

METHODS

This retrospective study included 15 patients with 15 baseline MRIs and 34 serial examinations. R2* LIC estimates were calculated using the FuncTool, CMRtools/Thalassemia Tools and Quanta Hematology programs. Longitudinal LIC changes (ΔLIC) were calculated using the subset of 34 serial MRIs.

RESULTS

After Bland-Altman analysis on baseline data, Quanta Hematology, which employs the monoexponential-plus-constant fit, produced the lowest mean difference [0.01 ± 0.14 log(mg/gdw)] with the closest limits of agreement. In the longitudinal setting, Quanta Hematology again gave the lowest mean difference between R2 and R2* LIC (0.1 ± 2.6 mg/gdw). Using FerriScan as reference, the value of concordant directional ΔLIC changes was the same for all programs (27/34, 85.7 %).

CONCLUSIONS

R2* LICs are higher than R2 LICs at iron levels <7 mg/gdw, while R2 LIC averages higher than R2* LIC with increasing iron load. The monoexponential-plus-constant model provided the best agreement with R2 LIC estimates.