Signal Quantification of Intravenous Contrast Agents Enhancement from Biphase Liver CT Scan Procedures

In multi-detector computed tomography (MDCT) abdominal and pelvic CT treatments with intravenous (IV) contrast media (CM), automated bolus monitoring with a fixed contrast enhancement delay was examined. Statistics assess all contrast enhancement variables, including patient data like body weight, cardiac output, and contrast injection settings. This study comprised 100 retrospective and 43 prospective patients. In the first group, the Hounsfield unit (HU) was measured before, 30 seconds, and 70 seconds after CM. The second group measured age, weight, heart rate, systolic and diastolic blood pressure, and creatinine. The radiographer computed CM time based on HU values around 120. The differences in HU levels across groups were used to create an equation for imaging time prediction utilizing auto-mated bolus monitoring. The Bolus Time Equation’s predictors included patient weight, heart rate, creatinine level, and systolic blood pressure, with 34.9% dependency and 59.1% influence on each variable. The equation is trustworthy since the ANOVA test indicated p = 0.002. The computation and study gave the same Bolus Time value with a p-value of (0.992 > 0.05). The first and second groups exhibited very different HU rates (p-value 0.00). The research found that fixed-time improved more than bolus monitoring, which performed better.


Introduction
Computed Tomography (CT) is a diagnostic imaging technique that produces comprehensive images.Tomography has become widely used in scientific medical research and industry for its nondestructive and high-resolution means of detecting internal structure.This includes a variety of applications in seismic data, digital rock physics, medical modalities such as computed tomography, magnetic resonance, and others [1][2][3][4][5][6][7].Clinicians can use CT to find and diagnose trauma injuries, identify the location and size of a lesion, identify tumor stages, evaluate diseases, identify the location of blood coagulates, detect pulmonary and heart diseases, and recognize ambiguous stomach disorders.
CT is a common modality that comes in a variety of sizes, devices, and slice numbers (number of channels).Philips, Siemens, Toshiba, and GE are the leading CT scan manufacturers.Each CT scanning brand offers a variety of CT scan ma-chines.All brands use computer technologies to calculate the Hounsfield Unit (HU), utilizing a compilation algorithm to build axial or horizontal images (slices) of the human body.CT scan cross-sectional images can be reconstructed in several planes to provide three-dimensional views [8][9][10][11][12].CT scanning is frequently the most effective diagnostic technique for various illnesses.Because it is rapid, painless, non-invasive, and extremely accurate, it aids clinicians in recognizing numerous diseases, saving valuable time for patients.It can identify injuries inside that are bleeding sufficiently rapidly in an emergency situation to help preserve human lives [13][14][15][16][17].
There are numerous CT protocols.One of the most popular liver and abdominal protocols.In this protocol, CT imaging uses intravenous (IV) contrast agent (CA) to visualize the intricate structure of the liver and other abdominal organs in or-der to improve both diagnosis and therapy [18][19][20][21][22].For all patients, the CT protocols employed in the imaging department are the same.To acquire optimal image quality at the ideal time to take an image with perfect organ enhancement, it must perform different protocols depending on the patient.CA reaches organs at various times in different people.CT has presented researchers with ongoing possibilities to improve image quality and clinical practice [23][24][25][26][27].
There are many factors that radiologic technologists should consider when using CA, such as the duration of injection, which is the most important injection-related factor influencing CT scan timing, and the amount of CA, which corresponds to the essential patient-specific variables influencing the amount of vascular and parenchymal contrast enhancement, the latter of which is represented by body weight [6,[28][29][30][31].
With the latest advances in helical computed tomography (CT) technology, multi-detector CT (MDCT) scanning of the whole liver in less than 20 seconds with two or more separate phases is now achievable.With this fast scanning, it is critical to reduce the time delay between the contrast agent injection and the diagnostic scan beginning, particularly for the arterial phase.Individual differences in body weight, heart rate, and cardiac blood pressure might affect the time frame and the required rate and amount of the contrast agent, making it challenging to achieve optimal contrast enhancement [32][33][34].
Tracking the bolus technique of the contrast agent could assist with individualizing the time delay, but it is a time-consuming procedure.To address these constraints, an automated bolus monitoring system that automatically launches diagnostic scans prompted by contrast enhancement has been recently developed.Using low-dose scans (approximately 50 mA), this method enables scans to begin, either manually or automatically, when the contrast enhancement in a region of interest (ROI) reaches a predetermined threshold.Some investigations found that automated bolus tracking might increase the degree of contrast enhancement and lesion-to-parenchyma complicity by better individualizing the time delay for the beginning of diagnostic scans of the liver and pancreas.However, this area still has considerable debate, and past research has yielded conflicting findings.Some studies propose bolus monitoring only if the patient is over 70 years old, has cardiovascular diseases, or has no appropriate antecubital vein for a contrast agent injection to decrease the excess radiation dosage [35].Furthermore, some studies have shown that using the bolus monitoring approach, 35% of patients may not attain a threshold of 50 HU above base-line within 60 seconds following injection beginning, necessitating a programmed delay [36].
Many medical imaging departments employ hepatic, abdominal, and pelvic CT procedures on a regular basis.The procedure is often carried out by injecting an iodine-containing contrast agent.However, for various individuals, contrast media enters the ROI at different times and with varying degrees of enhancement due to various parameters such as the patient's blood pressure, weight, injection rate, contrast agent amount, and other factors.

Study Design
The present study evaluated two patient groups retrospectively and prospectively.The first group includes all liver CT scan patients who have completed the abdominal and pelvic protocols.The data came following the CT scan.A pelvic and abdomen automated bolus-tracking CT scan was performed on the second group.All the different coefficients affecting on time bolus were combined from first group to create mathematical equations that can be applied to the second group in determining the bolus timing.With correct scheduling, prospective data were obtained.Medical hospitals were contacted and given permission to collect data.This research comprised male and female participants who received abdominal and pelvic liver CT exams with IV contrast and had suitable creatinine level.

Data Collection and Interpretation Procedure
The patient was supine with the feet-first protocol.The gantry is cantered on the abdomen, and the arms are hoisted above the patient's head.A single breath-hold NECT scan was done first from the diaphragm to the symphysis pubis.Power injectors delivered water-soluble, non-ionic IV contrast at varied volumes and flow rates.Postcontrast arterial, venous, and delay measurements were taken at different intervals using bolus tracking or fixed time delay.Kilovoltage peak (kVp) parameters were manually adjusted to 120 to 140 kVp, 313 to 438 mAs dependents on patient size, 0.75 rotation time, and 0.938 pitch.The CA volume ranges from 90 to 120 ml, according to patient size, with flow rates from 3 to 3.3.
Three radiologists with experience more than 5 years estimated patients' HU in the liver's middle portal vein.A circular region of interest (ROI) of roughly 2 mm2 was created on a commercial DICOM to insert patients' images.The data was encoded and entered into SPSS 20.Descriptive statistics were calculated for each variable.as well as Bolus time predicting forecasts.The mean and standard deviation were determined for each domain.

Results
The research included 143 individuals who had liver CT scan treatments at three different institutions.The average age of the participants was 47.64, with 89 (62.2%) men and 54 (37.8%) females.
Information coefficients and descriptive statistics may summarize datasets.Standard deviation, mean, minimum, and maximum measured variability.Table 1 provides the variables necessary to develop a bolus time predictor equation.To predict Bolus time, the degree to which each variable was investigated related to this aspect.Given that age, patient weight, patient height, heart rate, creatinine level, CM quantity, flow rate, and systolic and diastolic blood pressure were studied.
The Anova P-value determined the association between each variable and Bolus time.Patient weight, heart rate, creatinine, and systolic blood pressure were be-low 0.05.This implies that each of these elements affects Bolus's time (see table 2).
To determine the relationship between the four variables and bolus time, R and R Square were calculated for the patient weight (0.386 and 0.149), heart rate (0.372 and 0.139), and creatinine level (0.330 and 0.109).Systolic blood pressure rang-es from 0.389 to 0.151.Each variable's dependence and effect on bolus time were low, as indicated by R square, which was less than 0.2 for all variables.Thus, the researcher will calculate coefficient factors for each of the four variables to create the bolus time equation.

Weight coefficient
The coefficient of the patient's weight that affects bolus duration might be linear, logarithmic, inverse, quadratic, cubic, compound, power, growth, exponential, or logistic.R and R-squared were assessed to determine the best connection from the above.Instead of correlation (R), R-squared illustrates how much of a data set's variability (patient weight) can be attributable to a single independent variable (bolus time).The value of R square may be any percentage between 0 and 1, whereas R can be -1 to +1.R square is a proportion between 0% and 100%.When the proportion approaches 100%, the independent variable utilized to determine the dependent variable is optimum.Investing stocks and funds with an R-squared rating between 85% and 100% track the index closely.A fund with an R-squared below 70% does not closely follow the index.A higher R-squared indicates a better beta value.Adjusted R-squared provides a more accurate connection by including all independent variables (patient weight) on the regression function.This makes it easy to pinpoint the link's causes.Knowing which aspects are most important helps.A cubic relationship may be used to compute the coefficient of the patient weight, which is the most significant variable in the value of bolus time.Figure 1 shows the equation between bolus time and patient weight.This equation ( 1), which can be expressed as follows: ω=183-(5.84*weight)+(0.07*weight^2) -(0.000265*weight^3) (1) Where ω is the weight coefficient, this symbol was selected in role of first letter of weight -just for easy memorized-.R-square 21.1% which indicates the highest percentage among all relationships which means that patient weight at-tributes 21.1% on the bolus time, the value R equals to 45.9% which indicates the strength of a link between patient weight and bolus time.Adjusted R-square have the height value in cubic relationship comparing with other relationships.The value was 15% which means the patient weight can gives additional contributes on the bolus time by 15%.

Heart Rate Coefficient
Another cubic relationship may be used to compute the coefficient of the heart rate, which is the second most significant variable in the value of bolus time.

Creatinine Coefficient
Another cubic relationship may be used to compute the coefficient of creatinine.Equation (3) shows the equation between bolus time and patient creatinine level.This equation ( 3), which can be expressed as follows: ∁=-59.16+(325*creatinine) -(396*〖creatinine〗^2) +(163*〖creatinine〗^3) (3) Where ∁ is the creatinine level coefficient, this symbol was selected in role of first letter of creatinine -just for easy memorized-.R-square 12.9 % which indicates the highest percentage among all relationships which means that patient weight attributes 12.9 % on the bolus time, the value R equals to 36 % which indicates the strength of a link between patient weight and bolus time.Adjusted R-square have the height value in cubic relationship comparing with other relationships.The value was 6.2 % which means the creatinine level can gives additional contributes on the bolus time by 6.2 %.

Systolic Blood Pressure (SBP) coefficient
Another cubic relationship was used to compute the coefficient of the systolic blood pressure.Equation ( 4) shows the equation between bolus time and patient creatinine level.This equation, which can be expressed as follows: τ=1250-(32.96*SBP) +(0.29*〖SBP〗^2) +(0.000828*〖SBP〗^3) (4) Where τ is the systolic blood pressure coefficient.R-square 17.9 % which indicates the highest percentage among all relationships which means that patient weight attributes 17.9 % on the bolus time, the value R equals to 38.9 % which indicates the strength of a link between patient weight and bolus time.Adjusted R-square have the height value in cubic relationship comparing with other relationships.The value was 10.9 % which means the patient SBP can gives additional contributes on the bolus time by 10.9 %.
The coefficient of determination, or R-Square, indicates how much of the variability in the dependent variable (bolus time) can be accounted for by changes in the independent variables (weight, heart rate, creatinine level, and systolic blood pressure).Table 3 implies that the factors weight, heart rate, creatinine level, and systolic blood pressure may predict 34.9% of the variation in bolus time.This is a global measure of the strength of association and does not indicate the strength of the link between any two independent variables.R-Square is a popular measure of statistical significance.The strength link between four variables and bolus time was 59.1%.
As more predictors are added to the model, part of the variation in the dependent variable (bolus time) will be explained by chance alone, and the R-square statistic will need to be adjusted accordingly.As more predictors are included, the model's ability to describe the dependent variable improves; however, some of this improvement in R-square is attributable to random variation in the sample.The purpose of the adjusted R-square is to provide a more accurate assessment of the population's R-squared.Adjusted R-square was 28%, whereas R-square was 34.9% in total.4 displays the results of the ANOVA analysis and whether there is a statistically significant difference between regression and residual.The F-value is calculated by dividing the Mean Square Regression (234.798)divided by the Mean Square Residual (46.112), which gives us F=5.092.The related p-value for this F-value is very low (0.002).When answering the question "Do the independent variables reliably predict the dependent variable?"these values are consulted.Your alpha level (usually 0.05) is compared to the p-value.The dependent variable (bolus time) may be accurately predicted using independent factors (body weight, heart rate, creatinine, and systolic blood pressure).The capacity of individual independent variables to predict the dependent variable is not an issue here; rather, the test analyzes whether the collection of independent variables can consistently predict the dependent variable when used collectively.The table 5 below lists each independent variable and discusses the capacity of each variable to predict the dependent variable.Using these estimations, the various factors affect the outcome of interest can be assessed.A one-unit increase in the predictor is assumed to increase in bolus time of approximately these values.In analyzing the coefficients, keep in mind that for the insignificant independent variables, there is no statistically significant difference between them and 0. (For information on whether or not the coefficients are statistically significant, refer to the rows labeled "tvalue" and "p-value").Patient Weight Coefficient (ω)-The coefficient (parameter estimate) is .634.So, for every unit (i.e., point, since this is the metric in which the tests are measured) increase in math, a .389unit increase in science is predicted, holding all other variables constant.(It does not matter at what value you hold the other variables constant, because it is a linear model.)Or, for every increase of one point on the ω, the bolus time is predicted to be higher by .634points.This is significantly different from 0. Heart Rate Coefficient (∂)-For every unit increase in ∂, there is a 1.178 unit increase in the predicted bolus time, holding all other variables constant.The variable ∂ is technically not statistically significantly different from 0, because the p-value is greater than 0.05 which was 0.144.Creatinine Coefficient (∁)-The coefficient for ∁ is .615.This means that for a 1-unit increase in the ∁, we expect an approximately 0.615-point increase in the bolus time.This is not statistically significant with p-value 0.144; in other words, .615 is not different from 0. Systolic Blood Pressure Coefficient (τ)-The coefficient for τ is .362.Hence, for every unit increase in τ we expect a 0.362-point increase in the bolus time.This is not statistically significant with p-value 0.116.Standard Deviation (Std.Error): These are the coefficients' standard deviations.By dividing the parameter estimate by the standard error, a t-value can be calculated (see the column with t-values and p-values) that may be used to evaluate whether or not the parameter is substantially different from 0. The final two columns of this table provide a confidence interval calculated from the standard errors and applied to the parameter.The outcomes of the predictors' use of coefficients with Std.Error 1 are estimates that are close to the true values.
Standardized coefficients are denoted by the symbol "beta."After normalizing the dependent (bolus time) and all of the independent variables (ω, ∂, ∁, and τ), these are the coefficients that would be obtained by running the regression.Before conducting the regression, standardizing the variables ensures that they are all on the same scale, allowing you to compare the coefficients' magnitudes to determine which variable is more influential.Also, bigger t-values (1.7) are correlated with larger betas (0.254).However, because all coefficients are scaled to the same value, standardized coefficients cannot be included into the equation.Unstandardized Coefficients illustrates the range of coefficient magnitudes.
Sig. (p-value) and t-value.The t-value and two-tailed p-value for testing the 0-significance level of the coefficient or parameter in these columns.With a 2-tailed test, you would check each p-value against the significance level you've established (0.05).P-values for coefficients that are smaller than alpha indicate statistical significance.With a 2-tailed test and alpha of 0.05, you should not reject the null hypothesis that the coefficient for (ω, ∂, ∁, and τ) is equal to 0, because p-value = 0.097, 0.144, 0.144, and 0.116 > 0.05, respectively.All coefficient p-values are larger than 0.05, indicating that there is no statistical significance at the 0.05 level.All of the coefficients' impacts on the bolus time are represented by the same indicators throughout a rather narrow range.Because of this, the researcher adjusted the coefficients for each patient's weight, heart rate, creatinine level, and systolic blood pressure as shown in equation (5).While the coefficient had no effect on the bolus time, the early findings without coefficients (patient's weight, heart rate, creatinine level, and systolic blood pressure) did.
Bolus Time= -47.711+(0.634*ω) +(1.178*∂) +(0.615*∁) +(0.362*τ) (5) 3.5 Differences of bolus time between tracker controlled in real measurement and equation measurement Table 6 and Table 7 display the mean and standard deviation of the research samples when applying the Bolus Time equation and the actual values obtained from the patients.The values for real bolus time (31.32± 8) as well as the values derived using the equation (31.33±4.7).There is no difference between the two measurements since the Sig.(2-tailed) value of p-value (0.992> 0.05), indicating that Bolus Time may be approximated using the equation.

Discussion
To avoid CM side effects, hospitals prepare patients a certain way.Referring doctors and radiologists should consider the risk-to-benefit profile of CM IV-enhanced CT exams and other options that may provide the same or better diagnostic information, as well as a compelling clinical indication [37].
Radiologists and referring clinicians should examine renewal alternatives that give higher-level diagnostic data, possibly decreasing risk-to-benefit.Some institutions utilize fixed-time contrast alone, whereas others combine bolus monitoring with CM for abdominal reasons.Injection length is one of the most critical injection-related characteristics governing phase timing since it directly affects organ and artery peak enhancement over time [38].Injection time, CM volume, flow rate, patient weight, height, heart rate, blood pressure, and CM concentration were ignored across all institutions, resulting in the disparity in enhancement.A regression equation predicts patient age, weight, height, heart rate, creatinine level, contrast agent quantity, flow rate, and systolic and diastolic blood pressure based on one or more independent factors.Data shows that the patient's weight, heart rate, creatinine level, and systolic blood pressure predict bolus time.Table 3 (R = 0.591) shows the association between actual and expected values.The independent variable explains 0.349 of the variation in the dependent variable, according to R square.Std.indicates in-sample predictive power, whereas R-Square does.The estimate error (6.79057) accounts for all other factors that impact the dependent variable.Regression and residual regression are compared to see whether two groups differ.The regression represents the patient's weight, heart rate, creatinine level, and systolic blood pressure, while the residual regression represents the remaining variables.The regression's sum of squares (=939.190) is more than the residual regression's (=1752.251).The significance value (=0.02) is less than 0.05, showing the greatest group differences.The beta coefficient is in Table 5.The coefficient test summarizes the studies: Bolus Time= -47.711+(0.634*ω) +(1.178*∂) +(0.615*∁) +(0.362*τ) (5) Make sure the resulting equation is applied to the variables in the provided equation.Outstanding results, with a mean estimated bolus time using the equation of 31.3361 and a standard deviation of 4.72976.Meanwhile, the genuine measurements had an arithmetic mean of 31.3256 and a standard deviation of B8.00512.For the equation's application, the Paired Samples Test was used to compare the two readings, which assessed the p-value at 0.992, indicating that the two equations are near.
An independent-samples t-test assessed ROI after CM HU for equal variances predicted and not assumed.The mean differences (P=0.00).The difference between bolus monitoring and fixing time, where the difference between the equal average of 22.4 is larger, indicating that we must use bolus time.
This study found that the CA's ROI arrival time depends on the patient's weight, heart rate, creatinine level, and systolic blood pressure.The results contradict the previous study [38], which found no significant association between the patient's weight and the bolus tracking technique, which is better than the fix time technique for improving abdomen CT with IV.The study [39] found that the total subjective quality of the image and diagnostic confidence were comparable.

Conclusion
Certain conditions need computed tomography (CT) with IV CA to be detect-ed.The CT scan with an IV contrast agent should not be generalized or considered routine.The employment of the described CA technique will provide superior results; on the one hand, the timing may be readily modified based on well-known and defined parameters; on the other hand, a considerably higher improvement in the diagnostic value will be obtained.The fixed-time usage of time resulted in a broad spectrum of diverse enhancements.When bolus tracking was applied, it generated superior enhancement outcomes than fix time.
Fig 1 shows the strongest patient weight-bolus time association.

Figure 1 .
Figure 1.mathematical relationship between bolus time and patient weight.
Figure 2 shows the equation between bolus time and patient heart rate.This equation (2), which can be expressed as follows: ∂=411-(15.79*HR)+(0.21*〖HR〗^2) -(0.000914*〖HR〗^3) (2) Where ∂ is the heart rate coefficient R-square 15.5 % which indicates the highest percentage among all relationships which means that patient weight attributes 15.5 % on the bolus time, the value R equals to 38.7 % which indicates the strength of a link between patient weight and bolus time.Adjusted R-square have the height value in cubic relationship comparing with other relationships.The value was 10.7 % which means the patient heart rate can gives additional contributes on the bolus time by 10.7 %.

Figure 2 .
Figure 2. Mathematical relationship between bolus time and patient Heart Rate.

Table . 1
Variable using the Bolus-tracking method for all prospective patients.

Table 2 .
Summary of the variables affecting bolus time.

Table 3 .
R, R Square and adjusted R square between Bolus time and independent variables.

Table 4 .
ANOVA test between regression independent variables and bolus time.

Table 5 .
Equation coefficient between independent variables and bolus time.

Table 6 .
Paired Sample Statistics Between Real Bolus Time and Equation Measurement.

Table 7 .
P-value between real bolus time and measuring time using the equation.