Beruflich Dokumente
Kultur Dokumente
Research article
A R T I C LE I N FO A B S T R A C T
Keywords: Objective: To investigate and evaluate the value as accuracy of diagnosis different degrees of anemia through the
Anemia unenhanced thoracic computed tomography (CT) scan.
Computed tomography Materials and methods: The CT attenuation of right ventricle (RV) cavity, left ventricular (LV) cavity, inter-
Interventricular septum ventricular septum and difference between the interventricular septum and left ventricle cavity (IVS-LV) were
Hemoglobin
retrospectively analyzed and measured in 317 patients with different degrees of anemia and the normal patients.
Thorax
The hemoglobin (Hb) level was estimated within 24 h. Statistical analysis was made to obtain the best diagnostic
threshold of anemia levels by CT scan, measurement repeatability was assessed using the intra-class correlation
coefficient (ICC) values.
Results: An obvious parallel correlation of hemoglobin concentration and CT attenuation of IVS-LV existed (the
determination coefficient was 0.818; P < .001). Based on receiver operating characteristic (ROC) curves, the
conclusion was that when the threshold of CT attenuation of IVS-LV at (5.5–9.5) HU in male and (4.5–8.5) HU in
female, the specificity for diagnosing mild anemia was low, while the sensitivity was very high (87.1%, 100%,
respectively), if the CT value was among (9.5–13.5) HU in male or (8.5–13.5) HU in female, the sensitivity and
specificity for diagnosing moderate anemia were fine (92.9% and 74.8% in male; 93.9% and 60.0% in female),
and once the CT value was more than 13.5 HU the sensitivity and specificity for diagnosing severe anemia in all
gender were greatly good (94.7% and 83.6% in male; 82.4% and 84.6% in female). The intra-and inter-observer
reliability and reproducibility were good (ICC were all more than 0.99).
Conclusion: The CT attenuation of IVS-LV could predict the severity of anemia with good sensitivity and spe-
cificity, which could add value to clinical practice and more importantly facilitating patient care as hematologic
laboratory investigations are lacking.
Abbreviations: CT, computed tomography; HU, hounsfield unit; RV, right ventricle; LV, left ventricular; IVS-LV, difference between the interventricular septum and
left ventricular cavity; Hb, hemoglobin; DVS, dural venous sinuses; ROIs, regions of interest; ICC, intra-class correlation coefficient; ROC, receiver operating
characteristic; NC, normal control
⁎
Corresponding author.
E-mail address: jnyyfsk@126.com (H. Zhang).
https://doi.org/10.1016/j.ejrad.2018.10.007
Received 14 June 2018; Received in revised form 16 September 2018; Accepted 8 October 2018
0720-048X/ © 2018 Published by Elsevier B.V.
Q.-Q. Zhou et al. European Journal of Radiology 108 (2018) 236–241
Fig. 1. A. One 27-year-old woman with mild anemia (Hb level 99 g/L). LV CT value was 29 HU, RV CT value was 31 HU, septum mean CT value was 376 HU, IVS-LV
CT value was 7 HU. B. One 43-year-old man with moderate anemia (Hb level 88 g/L). LV CT value was 20 HU, RV CT value was 21 HU, septum mean CT value was 32
HU, IVS-LV CT value was 12 HU. C. One 38-year-old woman with severe anemia (Hb level 49.2 g/L). LV CT value was 20 HU, RV CT value was 21 HU, septum mean
CT value was 41 HU, IVS-LV CT value was 21 HU. D. One 29-year-old woman who was normal participant (Hb level 128 g/L), and a long strip was drawn connecting
the depressed areas on the heart near the sternum (top arrow) and the depressed point near the spine (bottom arrow).
LV: left ventricular; RV: right ventricle; IVS-LV: difference between the interventricular septum and left ventricular cavity; Hb: hemoglobin; CT: computed tomo-
graphy.
sign” or “interventricular septum sign”, or they can deduce anemia by information to patient care.
measuring the CT value of dural venous sinuses (DVS). However, none
of the previous studies have demonstrated predictability of severity of
2. Materials and methods
anemia on unenhanced thoracic CT by measuring CT attenuation of
difference between the interventricular septum and left ventricular
2.1. Study population
cavity (IVS-LV), nor have established statistically significant data re-
garding sensitivity and specificity. Even more, they failed to propose
This retrospective single-center study was approved by the local
effective methods when distinguished mild anemic patients from the
medical ethics review boards. A total of 317 consecutive patients (mean
normal. Although Chaudhry et al. [14] had assessed CT attenuation of
age ± standard deviation, 57.66 ± 16.50; 164 females) who pre-
DVS to evaluate the severity of anemia, nevertheless, the consistency of
sented at hospital from January 2016 to March 2017 were enrolled, in
measurement was not so well and the clinical popularizing rate was
addition, all participants performed an unenhanced CT scan and per-
low.
ipheral blood examination within 24 h. Most of them suffered with
As anemic patients are more likely to be hospitalized, have in-
hematological diseases such as iron deficiency anemia, myelodysplastic
creased lengths of hospital stays, have increased disability resulting in
syndrome, leukemia and lymphoma, and chronic kidney disease in-
greater health care costs, furthermore, the level of hemoglobin has a
cluding chronic nephritis, renal failure, and others were malignant tu-
significant influence of patient's life quality and the prognosis of the
mors, multiple fractures and coronary heart disease, cardiac in-
disease [16–25]. The primary purpose of this study was to evaluate the
sufficiency. According to the hemoglobin level, four subgroups have
strength of correlation between CT attenuation of IVS-LV and patients’
been formed: mild anemia group (n = 60), moderate anemia group
Hb level. We hypothesized that the optimal threshold for diagnosing
(n = 91), severe anemia group (n = 53) and the normal control (NC)
severity of anemia was obtained which will be beneficial to clinical
group (n = 113). Anemia is classified according to the WHO [6] and
treatment and disease prognosis for the patients who cannot tolerate
Chinese criteria of diagnostic anemia [26]: hemoglobin concentration
blood drawing (such as severe coagulopathy) or in emergency thor-
(g/L) at sea level is (90–120) g/L in male defined as mild anemia (fe-
acoabdominal CT examination, such information could add serviceable
male: [90–110] g/L); (60–90) g/L defined as moderate anemia; less
237
Q.-Q. Zhou et al. European Journal of Radiology 108 (2018) 236–241
Table 1
Descriptive analysis of baseline characteristics of the study population and the CT value.
Index NC group Mild anemia group Moderate anemia group Severe anemia group P
n = 113 n = 60 n = 91 n = 53 value
Age (years) 57.00 ± 14.67 57.40 ± 17.72 59.88 ± 15.77 55.55 ± 19.73 .439
No. of women 52/113 (46.0 %) 29/60 (48.3%) 49/91 (53.8%) 34/53(64.2%) .159
Hb level (g/L) 133.49 ± 12.06 101.63 ± 7.20 73.03 ± 8.05 53.56 ± 5.60 .000*
RV CT value 36.27 ± 4.76 29.13 ± 4.10 24.08 ± 5.10 21.25 ± 4.10 .000*
LV CT value 35.20 ± 4.65 28.97 ± 4.50 23.08 ± 5.12 19.96 ± 4.20 .000*
Septum CT value 38.47 ± 4.46 37.00 ± 4.17 36.76 ± 5.06 36.23 ± 3.84 .007
IVS-LV CT value 3.27 ± 2.75 8.03 ± 2.72 13.56 ± 2.94 16.26 ± 2.46 .000*
Note—Unless otherwise indicated, data are means ± standard deviations, and numbers in parentheses are percentages.
LV: left ventricular; RV: right ventricle; IVS-LV: difference between the interventricular septum and left ventricular cavity; Hb: hemoglobin; CT: computed tomo-
graphy; NC: normal control.
* P < .001 The difference was statistically significant in comparison with all subgroups.
than 60 g/L defined as severe anemia. The exclusion criteria were as parameters and the least significant difference (LSD) multiple com-
follows: (1) the subjects were aged under 15 years; (2) had diseases that parison method was employed to compare differences among variables.
affect cardiac density such as vasculitis lesions, glycogen accumulation, Categorical variables were assessed by the Pearson’s χ2 test, with the
or hepatic hemochromatosis; (3) had a history of iodine allergy test or Fisher exact test used when n < 40 or T < 1. To identify the optimal
transfusion high protein recently; (4) performed enhanced scan with diagnostic parameter, scatter plot and simple linear regression analysis
contrast medium before the CT scan; (5) images have motion or metal were performed in the study groups. Receiver operating characteristic
(calcification) streak artifacts. (ROC) curve analysis was performed to determine the threshold CT
value for diagnosing the severity of anemia, and Medcalc Statistical
software version 15.6.1 (Medcalc Software bvba, Ostend, Belgium) was
2.2. CT scan and data measurement
used to compare the statistical significance of the ROC areas among
these groups. Measurement repeatability was assessed using intraclass
All non-contrast thoracoabdominal CT studies were volume scans
correlation coefficient (ICC) values. A P value less than 0.05 was con-
performed on a 16-detector row CT scanner (Brilliance,Philips,
sidered to represent the significance threshold. All statistical analyses
Eindhoven, the Netherlands) with a chest scan from the thoracic inlet to
were performed with SPSS software version 21.0 (International
the diaphragm, and abdomen scan from the top of the diaphragm to the
Business Machines Corporation, Chicago Illinois, United States). All
anterior superior iliac spine level. The scans were performed after
graphics were created using GraphPad Prism version 5.00 for Windows
breath-hold. Scan parameters for this protocol were tube voltage
(GraphPad Software, San Diego California, United States).
120 kV, automatic tube current setting (30–500 mA), rotation time of
0.4 s, and axial slice thickness of 0.5 mm with reconstruction interval of
1.25 mm. Unenhanced CT images were evaluated on a preset standard 3. Results
mediastinum window (window width, 350 HU; window level, 40 HU).
Two experienced pectoral radiologists (Y.S.Y. and B.B.D.; pectoral The study cohort contained 164 (51.74%) female and 153(48.26%)
radiologist with 12 and 10 years' experience, respectively) who were male participants aged 57.66 ± 16.49 years (mean ± standard de-
blinded to the subjects’ characteristics independently read and mea- viation; range, 16–90 years). The NC group (n = 113; age,
sured the CT value (unit HU) of patient's left ventricle (LV), right 57.00 ± 14.67 years; 52 females), mild anemia group (n = 60; age,
ventricle (RV), and the ventricular septal [10] on a Picture Archiving 57.40 ± 17.72 years; 29 females), moderate anemia group (n = 91;
and Communication System (PACS) (ZLsoft, Chongqing, China). The age, 59.88 ± 15.77 years; 49 females), and severe anemia group
average of two measurements was considered as the result, and the CT (n = 53; age, 55.55 ± 19.73 years; 34 females) were matched for age
attenuation difference between the interventricular septum and left and gender (P > .05). After ANOVA analysis and LSD multiple com-
ventricle (IVS-LV) was calculated. parison, Hb level, the CT value of RV, LV and IVS-LV showed significant
Regions of interest (ROIs) were placed on the LV cavity, RV cavity, difference in comparison with all subgroups (all P < .001) (Table 1).
and interventricular septum. The sizes of the ROIs differed from each A pectoral radiologist (Y.S.Y.) measured all the parameters using the
other based on the difference in size of the underlying, irregularly long PACS once and repeated the measurement a week later, while another
strips was measured on interventricular septum and small rounds were pectoral radiologist (B.B.D.) obtained another measurement. The intra-
taken on ventricular cavity approximately 100 mm2 for the areas and inter-observer ICC values were close to 1 (Table 2), showing the
(Fig. 1). We choose the level avoiding artifacts from the spine as the high reliability and good repeatability of CT measurements obtained
heart measurement levels. In cases where difficulty was encountered using the PACS.
during delineation of the interventricular septum, a long strip con- In all subjects, the scatter points of Hb level, RV CT value, LV CT
necting the depressed area on the heart margin near the sternum (top value and IVS-LV CT value were all linearly distributed around the
arrow) and the depressed point near the spine or descending aorta diagonal. There was a positive linear correlation between Hb level and
(bottom arrow) was drawn, and the possible CT value of the inter- RV/LV CT value (R2 = 0.646 and 0.664 respectively; P < .001). Hb
ventricular septum was measured along this line (Fig. 1D)
Table 2
Comparison of intra- and inter-observer measurement reliability and repeat-
2.3. Statistical analysis
ability.
Demographic information, measurement data and clinical char- ICC 95%CI P value
acteristics (age, gender and Hb level) were reported as mean ±
intra-observer 0.997 0.996–0.997 .573
standard deviation and compared between the four groups. The per- inter-observer 0.993 0.991–0.995 .577
formance of the Kolmogorov-Smirnov test to detect normal distribution.
One-way analysis of variance (ANOVA) was used for continuous ICC: intraclass correlation coefficient.
238
Q.-Q. Zhou et al. European Journal of Radiology 108 (2018) 236–241
sensitivity and specificity were 93.9% and 60.0%, and when the CT
value was greater than 13.5 HU in diagnosing severe anemia, the di-
agnostic sensitivity was 82.4% and the specificity was 84.6%.
4. Discussion
Fig. 3. Comparison of receiver operating characteristic (ROC) curves of IVS-LV CT value diagnosing various degrees of anemia in male patients (A) and female
patients (B).
IVS-LV: difference between the interventricular septum and left ventricular cavity; CT: computed tomography.
239
Q.-Q. Zhou et al. European Journal of Radiology 108 (2018) 236–241
240
Q.-Q. Zhou et al. European Journal of Radiology 108 (2018) 236–241
biopsy in hematologic patients with lung infection, Zhonghua Xue Ye Xue Za Zhi 33 M. Cetin, Diagnostic quality of CT pulmonary angiography in pulmonary throm-
(8) (2012) 657–659. boembolism: a comparison of three different kV values, Med. Sci. Monit. 19 (2013)
[27] R.R. Haghighi, S. Chatterjee, A. Vyas, P. Kumar, S. Thulkar, X-ray attenuation 908–915.
coefficient of mixtures: inputs for dual-energy CT, Med. Phys. 38 (10) (2011) [30] K. Zarkovic, P. Larroque-Cardoso, M. Pucelle, R. Salvayre, G. Waeg, A. Negre-
5270–5279. Salvayre, N. Zarkovic, Elastin aging and lipid oxidation products in human aorta,
[28] D. Norman, D. Price, D. Boyd, R. Fishman, T.H. Newton, Quantitative aspects of Redox Biol. 4 (2015) 109–117.
computed tomography of the blood and cerebrospinal fluid, Radiology 123 (2) [31] A. Saha, S. Roy, Papillary muscles of right ventricle-morphological variations and
(1977) 335–338. its clinical relevance, Cardiovasc. Pathol. 34 (2018) 22–27.
[29] O. Yilmaz, E.D. Ustun, M. Kayan, F. Kayan, A.R. Aktas, E.N. Unlu, B. Degirmenci,
241