Evaluation of Severe ultrasound and gene diagnosis in cardiac index and shock patient index of shock patients

Abstract

BACKGROUND:

The kidney is the most vulnerable organ in severe patients. In severe cases, the fatality rate of acute kidney damage is as high as 30% $\sim$ 60%. Severe ultrasound is a non-invasive method to evaluate renal blood flow. It can give a semi-quantitative score of renal blood flow and measure the Resistance Index (RI), which can reflect renal artery blood flow to a certain extent.

OBJECTIVE:

There is little literature on hemodynamic regulation in septic shock patients, but almost no research report on the relationship between hemodynamics and RI exists. Therefore, this paper proposed the analysis of severe ultrasound and gene diagnosis in cardiac index and peripheral vascular RI of shock patients.

METHODS:

This paper mainly expounded on detecting renal function parameters and RI in patients with viral shock to understand further the correlation between them and renal flow and RI.

RESULTS:

It could be seen from the experimental results that the P values before and after resuscitation in the two groups with and without elevated Cardiac Output (CO) were 0.41 and 0.12, respectively, which were more significant than 0.05.

CONCLUSION:

RI had no apparent relationship with CO, and RI could not be used as an evaluation index for patients with early septic shock.

Keywords

Severe ultrasound shock patients gene diagnosis genetic algorithm

1. Introduction

Shock is a disease with high mortality in the Intensive Care Unit (ICU). Its hemodynamic classification is an essential basis for clinical treatment, and it is also a milestone for people to recognize and treat critical illness. It promotes the treatment of shock and significantly changes the original clinical manifestations. In the rapid diagnosis of the causes of shock patients, there are many defects in the conventional clinical methods. Under the guidance of severe medicine, severe ultrasound is used to conduct a dynamic evaluation of multiple objectives integration for critically ill patients. Severe ultrasound is significant in evaluating cardiopulmonary function and hemodynamics and has more trials and benefits in evaluating difficult airways and intraoperative volumes. In recent years, Modified Critical Care Ultrasonic Examination (M-CCUE) technology has significant advantages in diagnosis and treatment, and applying the M-CCUE scheme in ICU can guide clinical decision-making. However, M-CCUE still has some reference value in the etiological diagnosis of shock. Therefore, this paper discusses and preliminarily analyzes the improved intensive care strategy and gene diagnosis expression.

In diagnosing and treating critically ill patients, adopting a flow-oriented ultrasound program is the guarantee for the rapid and efficient implementation of severe ultrasound. Alharty Abdulrahman prospectively registered and tracked 89 ICU patients diagnosed with COVID-19 and used a portable machine to use a phased array, convex, and linear transducers for ultrasonic examination of nursing points [1]. Zhang Jiao described the clinical experience of real-time lumbar puncture guided by ultrasound in patients with spinal muscular atrophy with severe scoliosis through experiments, including the technical points of this technology [2]. Lung Ultrasound (LUS) was successfully applied to diagnose neonatal pulmonary diseases. As a non-invasive and radiation-free tool, applying LUS in nursing points to diagnose pulmonary diseases has become a new global trend. Liu Jing introduced the latest application of LUS in guiding or assisting in treating neonatal pulmonary diseases [3]. Experts worldwide agreed that severe ultrasound could rapidly narrow the scope of differential shock diagnosis and was an essential method for early and continuous evaluation of hemodynamics.

This group of early clinical studies also showed the importance of ICU in evaluating patients’ heart and lung functions and lung tissues. A pulmonary ultrasound score is considered an appropriate semi-quantitative score to measure pulmonary ventilation loss. This score has been proven valuable in diagnosing and monitoring pulmonary pathology, but research has yet to show its relationship with the results. Yin Wanhong aimed to study the relationship between LUS scores and ICU shock patients’ prognosis [4]. Borges Rodrigo Cerqueira evaluated the relationship between the cross-sectional area of rectus femoris and the strength of bedside muscles of patients admitted to the ICU due to severe sepsis and septic shock [5]. However, they did not analyze the peripheral vascular RI or study their gene diagnosis scheme.

This paper focused on evaluating the application of improved M-CCUE in the etiological diagnosis of various shock patients. The clinical data of the Department of Critical Care Medicine were analyzed. The enrolled patients were evaluated preliminarily by M-CCUE within 30 minutes after entering the room. M-CCUE Score (MCS) was used to score them, and relevant data results were analyzed. For shock patients with unknown etiology admitted to the ICU, the M-CCUE scheme required a short time for initial diagnosis of the cause; with a high correct diagnosis rate, sensitivity, and specificity, its quantitative evaluation results could predict the severity of the patients.

2. Genetic algorithm

2.1 Characteristic gene selection technology of genetic algorithm

The interaction of external factors such as physical and chemical environment and genes causes diseases. In addition to body injury, human diseases are caused by direct or indirect expression of genes. Shock is a complex disease that seriously endangers human health, and its clinical manifestations are complex and diverse. It is easy to be found and has a high recurrence rate. The clinical diagnosis and subtype classification cannot meet the actual needs. Most scholars believe that shock is the main variation produced by normal tissues under the action of physical or chemical carcinogens, that is, the change of the base pair structure or relative arrangement order of normal genes, which makes it run counter to the original average distribution level. Therefore, studying the gene expression difference in different samples is necessary. At present, it is one of the leading research directions to identify and classify the central functional genes of shock.

With the continuous development of gene chip technology, people can compare the expression of microarray-based genes, understand their expression in different populations, and then understand their genetic mechanism so as to achieve the diagnosis and classification of shock. Traditional diagnostic methods are primarily based on the subjective judgment of pathology and medical staff or based on experience. With the help of a gene chip, although the cells in these tissues did not have noticeable physiological changes, they could get a correct diagnosis through gene analysis, which won precious time for prevention and treatment and extended the life of patients. Medical scoring systems use imaging investigations, clinical examination methods, laboratory testing, and medical gadgets to evaluate physiological factors. Based on predetermined criteria, numerical values are assigned to these variables, which include clinical signs and symptoms, laboratory data, and vital signs. Another important significance of this technology is that it can accurately distinguish different subtypes through comparative analysis of gene expression characteristics in normal and diseased populations to provide better treatment methods for specific types.

Relevant studies have proved that gene expression profiles contain a lot of noise and redundant genes, and machine learning of subtypes only needs to extract a small number of feature genes for subsequent classification. Efficient gene information extraction technology can eliminate the interference of noise genes and redundant genes and improve classification efficiency. With gene information extraction technology or bioinformatics, we can better comprehend genetics and use it in evolutionary biology, medicine, and agriculture. Using machine learning techniques, feature selection, dimensionality reduction, high-throughput data analysis, and multi-omics data integration improves accuracy in categorization. At the same time, it can significantly reduce the time and space complexity of machine learning. Faster, more memory-efficient, and scalable gene data analysis can be achieved by leveraging effective techniques such as feature engineering, dimensionality reduction, sparse learning, optimized model training, algorithmic optimizations, hardware acceleration, and distributed computing in gene information extraction technology.

There are mainly two kinds of feature gene screening techniques in gene expression profile data: One is based on a feature subset, and the other uses a single feature point for screening. However, the feature gene screening algorithm based on the feature subset considers the interaction between different genes, which makes the feature space too large and leads to the high cost of feature retrieval. It enables smooth interaction with downstream analytics, simplicity, efficiency, interpretability, adaptability, and scalability. Large datasets are handled efficiently, enabling the discovery of biomarkers and beneficial gene characteristics for genomic research. Large feature spaces in feature gene screening methods lead to higher costs because of scalability, computational complexity, memory utilization, search space exploration, and feature redundancy concerns. Practical techniques for feature selection, dimensionality reduction, and algorithmic optimization are essential to increase the efficiency of feature retrieval in genomic data analysis. Based on a single feature sequence, the recognition accuracy of this method is low because the interaction between genes is not considered. Because of their computational complexity, interpretability, dimension explosion, overfitting, and model parsimony, feature gene screening algorithms frequently neglect to consider gene interactions. Instead of comparing requiring exponential assessment, overfitting, or generalization performance, they evaluate single gene characteristics or straightforward combinations to find reliable biomarkers quickly. In addition, the feature selection method based on a single feature has a significant difference between candidate features that may appear in the same data set, which is the poor robustness of this algorithm. This paper proposes a new method based on characteristic genes to solve the above problems, as shown in Fig. 1.

Figure 1.

Method and technology route.

2.2 Characteristic gene selection method

In the gene expression profile, each sample contains thousands of different genes. Through feature screening of these genes, suitable genes can be screened from these genes. It is assumed that r features are selected from m genes, and the space in the feature subset to be searched is $E_{m}^{r}$ . If $m=$ 5000 and $r=$ 50, the computation is beyond the capacity of ordinary computers. Especially when the $r$ value is uncertain, this method needs a lot of computation. On this basis, a simple and reliable method is used to measure the difference of each gene in terms of how many classifications are used, and statistical methods are used to classify each gene. Finally, a specific number of genes are selected as characteristic genes, or a threshold is set. The characteristic genes are those that exceed this threshold. At present, the most commonly used method is single-feature classification. Figure 2 shows the overall process of single-feature sorting.

Figure 2.

Single-feature sorting process.

2.2.1 Signal-to-noise ratio

In a narrow sense, it is the ratio of signal voltage to noise voltage. The lower the signal-to-noise ratio of a device, the more noise it would emit. In a broad sense, it is the interference with practical information. This index is used to measure the contribution of each gene in the sample classification. The formula is as follows:

$\displaystyle t(h)=\left|{\frac{\varphi_{h}^{+}-\varphi_{h}^{-}}{\eta_{h}^{+}+% \eta_{h}^{-}}}\right|$ (1)

Among them, $t(h)$ is the score of gene $h$ . In different classifications, $\varphi_{h}^{+}$ and $\varphi_{h}^{-}$ are the mean values of gene h expression levels, respectively, while $\eta_{h}^{+}$ and $\eta_{h}^{-}$ are the standard deviations of gene h expression levels, respectively.

2.2.2 Sorting method

The sorting method is a machine learning algorithm for single attribute sorting based on weight. This method iterates the correlation between features many times. During each iteration, A is randomly selected from the data set. A is close to F in the same category, and A is close to N in other categories. The following formula calculates the difference between F and N to update the correlation between the characteristics.

$\displaystyle S(i)=S(i)-\frac{\textit{diff}(i,A,F)}{m}+\frac{\textit{diff}(i,A% ,N)}{m}$ (2)

$S(i)$ represents the relationship between feature $i$ A and target, set to 0 at the beginning of iteration. $\textit{diff}(i,a,{a}^{\prime})$ represents the difference between samples $a$ and ${a}^{\prime}$ . If the feature is continuous, $\textit{diff}(i,a,{a}^{\prime})_{\textit{it}}$ can be normalized from 0 to 1.

$\displaystyle\textit{diff}(i,a,{a}^{\prime})=\frac{\left|{a_{i}-{a}^{\prime}_{% i}}\right|}{a_{i\max}-a_{i\min}}$ (3)

Among them, $a_{i},{a}^{\prime}_{i}$ respectively represent the values of feature $i$ in samples $a$ and ${a}^{\prime}$ , while $a_{i\max},a_{i\min}$ respectively represent the maximum and minimum values of feature $i$ in all samples.

2.2.3 Ranking method of asymmetric dependency coefficient

Asymmetric Dependency Coefficient (ADC) is used to measure specific B correlation characteristic i by using information gain to classify it. A measure used in information theory called the Asymmetric reliance Coefficient (ADC) measures the asymmetric reliance between genes. This measure helps discover genes that significantly impact the expression of other genes, which improves our knowledge of gene interactions in biological systems.

$\displaystyle\textit{ADC}(B,i)=\frac{\textit{NJ}(B,A_{i})}{F(B)}$ (4)

Among them, $F(B)$ represents the entropy value of B, and $\textit{NJ}(B,A_{i})$ represents the mutual information between B and feature $i$ . The formulas are as follows:

$\displaystyle F(B)=-\sum\limits_{b}{q(B=b)\log q(B=b)}$ (5) $\displaystyle F(A_{i})=-\sum\limits_{a}{q(A_{i}=a)\log q(A_{i}=a)}$ (6) $\displaystyle\textit{NJ}(B,A_{i})=F(B)+F(A_{i})-F(B,A_{i})$ (7)

2.2.4 Absolute weight vector machine ranking method

Support vector machine is a widely used classification method not sensitive to data dimension, so it is particularly suitable for high-dimensional data. The absolute weight vector machine is developed based on the support vector machine. Because of concerns about privacy, legal compliance, and bias, the absolute weight vector machine (AWVM) ranking algorithm does not include dimensions for sensitive data. This preserves secrecy, privacy, compliance, equity, and efficacy in the decision-making processes. It can evaluate each attribute well and classify it according to its weight.

$\displaystyle\sum\limits_{i=1}^{q}{s_{i}a_{i}+y_{0}}=0$ (8)

Among them, $q$ is the general characteristic. $s_{i}$ refers to the weight of feature $i$ , which represents the importance of feature $i$ .

2.2.5 Rank sum check

Rank sum check does not strictly require the assumption of overall allocation. According to this feature, it is used to describe the differences between different types of features. Utilizing ranked data, the suggested approach uses the Mann-Whitney U test, commonly called the “rank-sum check,” to determine the statistical significance of group differences. In particular, when dealing with small sample sizes or data that is not regularly distributed, this non-parametric test facilitates reliable statistical comparisons, hypothesis testing, and data interpretation. It employs U-statistics based on ranks of observations, is less sensitive to sample size, works well with ordinal or ranked data, and evaluates population distribution hypotheses. Unlike parametric tests, which concentrate on means or variances, it offers trustworthy inference in various research contexts. The formula is as follows:

$\displaystyle R=(M-1)\frac{12}{M(M+1)}\sum\limits_{j=1}^{h}{m_{j}(\bar{K}_{j}-% \bar{K})^{2}}$ (9)

Among them, $M$ is the total number of samples, and $h$ is the number of different species. The number of samples for category $j$ is represented by $m_{j}$ . $\bar{K}_{j}$ represents the average rank sum of all samples in category j, and $\bar{K}$ represents the average value of samples in each category.

Therefore, the feature gene screening algorithm based on single attribute sorting has the advantages of a small amount of calculation and fast operation speed. This method does not rely on specific classification algorithms but is based on the distribution characteristics of the data itself. However, the threshold selection is too subjective, and the interaction between multiple genes is not considered, so it isn’t easy to find the best classifier.

2.2.6 Hybrid gene selection in genetic algorithm

Fewer characteristic genes with robust classification can effectively improve the classification accuracy of diseases. Use ensemble learning, cross-validation, regularization, integration of multi-omics data, and the selection of informative genes to increase the accuracy of illness categorization. These approaches ensure the reliable categorization of illnesses with fewer distinctive genes by optimizing generalization and predictive power. However, in actual research, Dapu’s gene expression is characterized by high dimensions and a small sample size. However, due to the shortcomings of significant noise and too much information, the classification of subtypes has always been a complex problem. The primary task of disease classification is to use the minor information gene to classify efficiently to identify specific types and achieve a higher classification rate. The selection of information genes includes the following aspects:

The less information genes are selected, the more noise can be removed;

Selecting fewer information genes can greatly reduce the cost of diagnosis. Strategies like feature selection, biological prioritization, machine learning models, cross-validation, and clinical validation decrease the number of informative genes while lowering diagnostic costs. This maximizes resource utilization and improves diagnostic accuracy without sacrificing accuracy.

Information genes with high classification accuracy may be closely related to the occurrence and development of diseases.

There are a lot of genes in each sample. Therefore, the subset space needed to be searched by feature set-based feature screening method needs to be bigger, resulting in a waste of patients’ diagnosis time and possibly delaying patients’ conditions. In this paper, multi-standard fusion technology was used for preliminary screening, which could not only avoid the computational complexity of the subsequent feature selection algorithm based on feature sets but also consider the irrelevance between multiple features from different perspectives to better discover potential essential genes and screen candidate genes with poor evaluation in each standard. Multi-standard fusion technology integrates data from several modalities, reducing computing complexity in feature selection methods. This method maintains classification performance while increasing computational tractability and improving resilience, efficient feature space exploration, and discriminative power.

2.3 Cardiac index and peripheral vascular RI

Septicemia is a common and fatal disease. From the perspective of hemodynamics, multiple causes may affect the function of blood circulation and tissue ischemia. If not handled in time, it would cause multiple organ failure and even death. These factors generally change over time. Although vascular tension and low blood volume are essential at the initial stage, endothelial and microcirculation dysfunction, abnormal distribution, vascular paralysis, capillary exudation, and various degrees of myocardial dysfunction are all related to the progress of septic shock. It is difficult to determine the main reasons for the early evaluation of ICU hospitalization; however, these different mechanisms of action would eventually lead to a common clinical feature, namely hypotension and hypoperfusion.

Continuous Blood Purification (CBP) is also known as continuous renal replacement therapy. Continuous blood purification, or CBP, includes taking a patient’s blood, passing it through a specialized circuit, and then returning the cleansed blood to the patient. It preserves the equilibrium of solutes and fluids while eliminating waste, pollutants, and extra fluid. Continuous blood purification is essential for metabolic regulation, renal, fluid, toxin elimination, hemodynamic support, multi-organ support, and therapeutic intervention in critically sick patients. In the early stage, it was mainly used to replace the renal function of critically ill patients. However, with the deepening of research, there is evidence that CBP can improve the hemodynamics of shock patients by clearing inflammatory mediators. Hemodynamic instability in shock patients during cardiopulmonary bypass (CPB) during cardiac surgery can be caused by an inflammatory reaction. For better hemodynamics, inflammatory mediators must be removed. Techniques for CRRT, specific adsorption materials, and ultrafiltration or modified ultrafiltration are examples of strategies.

Infectious shock is a hotspot and is challenging to research in severe medicine. Its pathogenesis is an uncontrolled inflammatory reaction released by a series of “waterfall” inflammatory mediators. It is characterized by high heart volume and low peripheral vascular resistance, causing tissue ischemia. The primary strategy is to monitor and analyze hemodynamics and provide corresponding treatment guidelines. Pulse waveform CO monitoring continuously detects CO, cardiac volume, peripheral vascular resistance, and other parameters to guide the rate of drug administration and fluid infusion. CBP can eliminate inflammatory mediators and endotoxin non-selectively through various ways, such as convection and adsorption, to end the cascade of cytokines and prevent the secondary damage of inflammatory mediators and endotoxin to organs. Secondary damage from inflammatory mediators can result in tissue damage, immunological dysregulation, brain problems, organ malfunction, microvascular dysfunction, systemic inflammation, and abnormal coagulation. Improved patient outcomes in various inflammatory disorders depend on efficient control and regulation.

The routine monitoring method in ICU is to place central venous and arterial catheters. Without advanced monitors or ultrasonic electrocardiograms, one of the biggest problems is whether traditional clinical monitoring methods can detect blood flow manifestations. With known signals, doctors can determine early resuscitation intervention. The drawbacks of conventional monitoring techniques in the context of early resuscitation measures include lack of integration, invasiveness, sporadic readings, restricted parameters, subjective evaluation, and detection delays. The following is the physiological background of the main components of the clinical hemodynamic assessment.

2.3.1 Pulse pressure as a substitute for stroke output

Since the early 1900s, many scholars have been trying to understand the relationship between arterial pressure and the output of each pulse. Pulse pressure is an exciting variable because it is easy to obtain. The heart’s function and interaction with the vascular system can be easily observed. According to the 3-ventricle arterial elastic chamber model, the characteristics of the arterial system are determined by the peripheral resistance, the compliance of the common artery, and the characteristic impedance of the aorta. The 3-ventricle arterial elastic chamber model provides insights into arterial hemodynamics, pressure-volume relationships, and clinical relevance for cardiovascular physiology and pathology. It comprises interconnected chambers that symbolize the left ventricle, systemic arterial tree, and peripheral tissues. The mathematical derivation of this model lays a foundation for establishing a CO monitoring device. In addition, many studies have shown that pulse pressure can fully reflect the output of each stroke, whether in a simulated state or different clinical conditions.

2.3.2 Arterial waveform analysis is a qualitative hemodynamic fine-tuning tool

Arterial waveform analysis can help people better understand the classical components of cardiovascular signals. For example, arterial waveforms show small, slow waves when aortic stenosis occurs. In addition, in the case of circulatory insufficiency, an essential analysis of arterial shape can provide helpful information for clinical application. Arterial waveform analysis is a qualitative method for evaluating cardiovascular health at the bedside, which offers current information on cardiac performance, vascular resistance, and volume status. It guides tailored measures to optimize hemodynamic support and enhance outcomes in critically sick patients, assisting in diagnosing and managing cardiovascular diseases.

The arterial waveform has apparent stenosis and diastolic pressure in patients with hypovolemic shock. When the left heart function is incomplete, its systolic period increases slightly. In distributive shock, such as septic shock, the arterial curve becomes more expansive, and diastolic pressure decreases, consistent with paralysis. In addition, the shape and position of the severe line can also provide patients with relevant clinical data in pulse curve tracing. For example, patients with low output and high peripheral resistance would produce heavy beats; vasodilation and low vascular resistance are related to a longer delay and a smaller reentry notch.

2.3.3 Central venous pressure

Central Venous Pressure (CVP) is an excellent hemodynamic index because it studies dynamic predictors of fluid reactivity, but its significance and interpretation need to be addressed. Although CVP cannot predict the response of body fluid well, it is a handy parameter for the role of heart and blood circulation. CVP is caused by venous return and cardiac function. Therefore, in the most straightforward estimation, high CVP can make the right ventricle only have a venous return so that it can work under high filling pressure.

In the same case, CVP is the upstream pressure of venous return. These changes can safely limit the fluid load under high pressure or before and after liquid excitation. In addition, the system’s perfusion pressure can be obtained from the difference between Mean Arterial Pressure (MAP) and CVP. From a physiological point of view, pre- and post-capillary pressure is a more meaningful indicator. At the same time, the motion track of CVP also provides essential information for the pathological status of patients.

3. Experimental evaluation of the role of severe ultrasound in gene diagnosis of shock patients

3.1 Experimental design

3.1.1 Research object

This paper treated various types of shock patients over 14 years old. Patients in critical illness or shock may find it difficult to adjust their posture while receiving treatment because of a variety of compromised physiological conditions. These conditions can make the issue worse and include hemodynamic instability, respiratory distress, neurological impairment, musculoskeletal weakness, and logistical difficulties.

Inclusion conditions: The patient entered ICU Ward 5 and met the diagnostic criteria for shock: 1) There was a cause of shock; 2) consciousness was abnormal; pulse fineness was more significant than 100 times or could not be touched; 3) the peripheral circulation was ischemia: The hands and feet were wet and cold; finger pressure of sternum skin was positive (refilling $>$ 2 seconds after pressing); the skin had patterns; mucosa was pale and cyanotic, etc; 4) urine volume was less than 30 ml/h or cannot urinate; 5) systolic blood pressure was less than 80 mm Hg; 6) pulse pressure was less than 20 mm Hg; 7) the systolic blood pressure of patients with hypertension was reduced by 30%. Any patient meeting the first item above and one of items 2, 3, 4, 5, 6, and 7 could be diagnosed as a shock; the age was 14 years and above; the informed consent of the patient or his authorized client was obtained.

Exclusion conditions: Some patients did not easily change their posture; patients with chest malformation, subcutaneous emphysema, dressing coverage, and inability to perform cardiac ultrasound examination; patients who received active treatment in the hospital.

3.1.2 M-CCUE scheme

All enrolled patients received systematic M-CCUE treatment within 30 minutes after admission. The program included ultrasound evaluations of severe lungs and severe hearts. This paper was conducted by a clinician who completed severe ultrasound training, obtained a certificate, and had over half a year of experience in severe ultrasound operation [6, 7]. To execute efficiently, analyze data correctly, and guarantee patient safety, ultrasound procedures call for qualified medical professionals. When it comes to severe medical diseases, skilled physicians may improve outcomes and maintain continuity of care by identifying irregularities, putting safety precautions in place, planning treatments, and prioritizing interventions. If two doctors disagree, the patient will be excluded from the experiment.

3.1.3 Experimental analysis

In this paper, 95 shock patients met the inclusion and exclusion conditions. Three patients received emergency treatment immediately after entering the room, and two patients were eliminated due to the inconsistent diagnoses of two doctors.

As shown in Table 1, 90 patients were 55–60 years old, including 60 men and 30 women. When entering the ICU, their heart rate was (112 $\pm$ 15.7) times/minute. Their body temperature was (37 $\pm$ 1.87)^∘C, and Acute Physiology and Chronic Health Evaluation (APACHE) was (25.45 $\pm$ 12.58).

Table 1
Statistical description of baseline data

Index	Result
Gender
Male	60
Female	30
Age	55–60
Heart rate	112 $\pm$ 15.7
Temperature	37 $\pm$ 1.87
MAP	37.9 $\pm$ 3.4
MCS	12.68 $\pm$ 5.31
APACHE	25.45 $\pm$ 12.58

Under the evaluation of the M-CCUE scheme, the MCS score of all patients was (12.68 $\pm$ 5.31).

Figure 3.

AUC analysis of various shock types.

Figure 4.

Correlation between MCS score, ICU fee, and APACHE score.

As shown in Fig. 3, Fig. 3(a) showed the sensitivity, specificity, and Area Under Curve (AUC) analysis of each shock type, and Fig. 3(b) showed the proportion of people in each shock type and the mean value of Confidence Interval (CI). In clinical trials, epidemiological research, public health studies, quality improvement programs, and market research, confidence intervals (CIs) are essential for precise population parameter estimates. Narrower intervals indicate more confidence and accuracy, and decision-makers can successfully evaluate CIs by evaluating the mean value and precision of the interval. Among them, 55 cases were distributed shock (A), accounting for 61.1%; 24 cases were hypovolemic shock (B), accounting for 26.8%; there were 4 cases of cardiogenic shock (C) and 4 cases of obstructive shock (D), respectively, accounting for 4.4%; 3 cases were mixed shock (E), accounting for 3.3%. The sensitivity of M-CCUE to A, B, C, D, and E were 0.966, 0.957, 0.843, 0.843, and 0.798, respectively. AUC values were 0.938, 0.962, 0.933, 0.933, and 0.788, respectively.

As shown in Fig. 4, Fig. 4(a) shows the correlation between the APACHE score and the total score of MCS, and Fig. 4(b) shows the correlation between ICU costs and the APACHE score. It can be seen from Fig. 4 that there is a positive correlation between the MCS score and the APACHE score. Still, there is no significant correlation between ICU cost and the APACHE score. The APACHE score is essential for clinical management because it directs prognosis prediction, severity rating, and risk assessment. It also helps doctors identify patients who are at high risk, prioritize interventions, and provide care on time.

The condition of shock patients in the ICU was complex. It changed rapidly and had multiple organs and high mortality. Hemodynamic classification of shock provided an essential basis for the clinical treatment of shock and was a milestone in people’s recognition and treatment of critical illness. The patients with shock could quickly diagnose the cause. The traditional clinical methods mainly included X-ray, CT, pulmonary artery CT, angiography, other imaging examinations, and invasive hemodynamics (Swan-Gans catheter, pulse indicating continuous CO). Conventional clinical procedures such as CT scans and X-rays include dangers, including cancer, radiation exposure, overuse, damage to the environment, and depletion of resources. Technology improvements combined with steps to reduce pointless testing and optimize imaging processes can reduce these hazards and enhance patient safety. Using a Swan-Ganz catheter, continuous cardiac output monitoring provides physicians with real-time information on a patient’s hemodynamic status. This information helps them identify changes, optimize fluid management, titrate drugs, and assess treatment efficacy. However, it had the disadvantages of time-consuming, high cost, inability to obtain relevant information at the bedside, and high transfer risk. It was impossible to find the cause of shock in time and accurately in clinical practice, which brought significant obstacles to future treatment. Therefore, it was a simple, accurate, and feasible method to find the cause of shock. Critical ultrasound had the advantages of being dynamic, real-time, convenient, and repeatable. Many studies believed that critical ultrasound was a very suitable means for evaluating essential patients, and it could provide some valuable information for clinical diagnosis and treatment of COVID-19 [8, 9]. Due to its quick, non-invasive, and real-time nature, critical ultrasonography is an essential tool for assessing shock patients. It allows for a multimodal evaluation of heart function, volume status, and organ perfusion, providing safety and direction for interventions.

3.2 Gene diagnosis evaluation of shock patients

The patients hospitalized in the ICU from October 2021 to January 2022 were selected for statistical analysis of their age, gender, APACHE, vasoactive drug use, and ventilator parameters.

Inclusion criteria: The diagnosis of septic shock was met, that is, severe systemic infection accompanied by hypotension (systolic blood pressure was lower than 90 mmHg, or systolic blood pressure was reduced by $>$ 40 mmHg. 1 mmHg $=$ 0.133 kPa). Although vasoactive drugs or favorable muscle strength treatment could maintain normal blood pressure, there was still abnormal tissue perfusion.

Exclusion conditions: under the age of 18; pregnant; severe chronic renal failure or renal failure caused by postrenal causes with definite diagnosis; renal artery stenosis; arrhythmia; transthoracic echocardiography and renal Doppler examination could not be performed; patients without carotid or subclavian arteries.

Methods: The Dutch bedside monitor detected the patient’s blood pressure, heart rate, and cardiac arterial pressure. The patient’s heart rate, blood pressure, CVP, left ventricular outflow tract velocity time integral, renal blood flow score, renal vascular RI, and urine volume before and after volume resuscitation were recorded in the ICU for 1 hour and 6 hours. Most studies on volumetric reaction set the critical value between 12% and 15% increase. In this paper, the increase of CO and MAP by at least 15% was selected as the indicator of a significant increase of CO and MAP, and the increase of CVP to at least 2 mmHg was chosen as the indicator of CVP increase.

Most kidney data were obtained from the right kidney, which was on the same side before and after resuscitation. The selection of renal RI was as follows: interlobar artery; the RI $=$ diastolic peak velocity/diastolic peak velocity/systolic peak velocity. By referring to the previous renal blood flow score and clinical experience, the following integration methods were used: 1 point $=$ only a tiny amount of blood flow signals were found at the renal portal; 2 points $=$ renal artery had apparent blood flow and a small amount of blood flow signal; 3 $=$ when the blood flowed to the arcuate artery, the blood flow of the whole kidney could be seen. In the case of more blood flow in the renal portal, blood flow could be seen in the intersegmental artery, but the blood flow that did not reach the renal portal artery was 1.5 minutes. In the renal portal artery, there was a lot of blood flow. Only a few arcuate arteries could see blood flow, but not all kidneys had blood flow of 2.5 minutes.

Figure 5.

Hemodynamic parameters before and after resuscitation.

As shown in Fig. 5, Fig. 5(a) showed the parameter changes among HR, MAP, and CVP before and after resuscitation, and Fig. 5(b) showed the parameter changes among CO, blood flow score, and RI before and after resuscitation. With relationships with renal artery blood flow integral and renal artery resistance index, mean arterial and central venous pressure provide insights into systemic hemodynamics and renal perfusion, impacting renal function in critically sick patients. MAP and CVP were higher after resuscitation than before. The P values were 0.02 ( $P<$ 0.05) and 0.005 ( $P<$ 0.01) respectively. There was no significant difference in HR, CO, blood flow score, and RI after resuscitation compared with before resuscitation.

Figure 6.

RI changes before and after recovery in each group.

As shown in Fig. 6, Fig. 6(a) showed the changes in CO indicators before and after recovery, and Fig. 6(b) showed the changes in MAP indicators before and after recovery. According to the standard that CO was more significant than or equal to 15%, it could be divided into two types: CO increase and CO not increase. To assess the efficacy of therapies in critically sick patients, patients are divided into “CO increase” and “CO does not increase” groups according to the kind of intervention, cardiac output measuring technique, response threshold, clinical context, and outcome measures. The P values of the two groups before and after resuscitation were 0.41 and 0.12, respectively ( $P>$ 0.05). There was no significant difference in RI between the two groups before and after resuscitation; according to the criterion that MAP was more significant than or equal to 15%, they were divided into two groups: MAP increased group and MAP not increased group. The P values of the two groups before and after resuscitation were 0.33 and 0.23, respectively ( $P>$ 0.05). There was no significant difference in RI between the two groups before and after resuscitation.

In shock patients, the evaluation of renal function has always been a vital link, and severe ultrasound has become a better method to evaluate renal function because of its non-invasive, reproducible, and other advantages [10, 11]. Evaluation of renal function is essential for managing shock patients because it helps with metabolic clearance, prognostication, fluid responsiveness prediction, diagnosis of acute kidney damage, and fluid balance maintenance. Monitoring facilitates drug changes, resuscitation, and the detection of problems. The research on renal microbubble contrast and ultrasound technology was increasing in various countries, and more and more scholars accepted its role in evaluating renal microcirculation and renal flow. Renal microbubble contrast and ultrasound technologies are hampered by allergic reactions, scarcity, expense, legal restrictions, and substitute imaging techniques. To ensure patient safety, healthcare practitioners must weigh benefits against risks. However, some problems still existed, such as allergy [12, 13]. Therefore, this paper will continue to explore the potential of current non-invasive treatment and distinguish it from previous studies in two aspects: First, CO, MAP, and CVP were combined to investigate the correlation between hemodynamics and renal artery blood flow integral and renal artery RI. A precise renal blood flow score was used to evaluate renal perfusion; in addition, in combination with cardiac and renal ultrasound, the advantages of non-invasive testing were used for the qualitative evaluation of critical patients [14, 15].

4. Conclusions

According to this study, the association between RI and CO, MAP, and CVP was not found to be significant at the early stage of ICU entrance. The following were this paper’s primary explanations for the phenomena above: Renal blood flow rate (RI) was measured as a ratio instead of blood flow rate. Apart from the actual renal blood flow, the following factors were associated with it: age, MAP, vascular elasticity, renal interstitial pressure, and intraperitoneal pressure. Furthermore, the Doppler method of measuring renal blood flow could have been more reliable, and even a small variation in the measurement angle would have resulted in significant variations in blood flow velocity. As CO and MAP levels rose in ICU patients experiencing septic shock, so did the accuracy of their renal blood flow indicators. No discernible relationship was needed between aAP and CVP and RI and CO. Therefore, in patients experiencing early septic shock, RI could not be considered a significant predictor of renal perfusion. The study’s weaknesses are the sample size, possible bias, confounding variables, temporal correlation between gene diagnosis and ultrasound, and emphasis on surrogate outcomes. Quality assurance and standardization are essential. Future investigations to evaluate the predictive significance of severe ultrasonography and gene diagnosis in patients with cardiac dysfunction and shock should use multi-omics, machine learning, artificial intelligence, and independent validation studies.

Data availability statement

No datasets were generated or analyzed during the current study.

Authors’ contributions

Weihua Wu, Jie Chen and Peng An are responsible for designing the framework, analyzing the performance, validating the results, and writing the article. Yangmei Feng, Chen Li, Meiqi Zhang and Zhenfei Yu collected the information required for the framework, provided software, performed a critical review, and administered the process.

Funding

The authors did not receive any funding.

Footnotes

Conflict of interest

The authors do not have any conflicts of interest to report.

References

Alharthy

Faqihi

Abuhamdah

Noor

Naseem

Balhamar

Al Saud

Brindley

Memish

Karakitsos

Blaivas

. Prospective longitudinal evaluation of point-of-care lung ultrasound in critically ill patients with severe COVID-19 pneumonia. Journal of Ultrasound in Medicine.2021 Mar; 40(3): 443-56.

Zhang

Cui

Chen

Dai

Huang

Zhang

. Ultrasound-guided nusinersen administration for spinal muscular atrophy patients with severe scoliosis: an observational study. Orphanet Journal of Rare Diseases.2021 Dec; 16: 1-8.

Liu

Xia

Ren

. The new application of point-of-care lung ultrasound in guiding or assisting neonatal severe lung disease treatment based on a case series. The Journal of Maternal-Fetal & Neonatal Medicine.2020 Dec 1; 33(23): 3907-15.

Yin

Zou

Qin

Yang

Zeng

Kang

, Chinese Critical Ultrasound Study Group (CCUSG). Poor lung ultrasound score in shock patients admitted to the ICU is associated with worse outcome. BMC Pulmonary Medicine.2019 Dec; 19: 1-8.

Borges

Soriano

. Association between muscle wasting and muscle strength in patients who developed severe sepsis and septic shock. Shock.2019 Mar 1; 51(3): 312-20.

Perrone

Soldati

Padovini

Fiengo

Lettieri

Sabatini

Gori

Lepore

Garolfi

Palumbo

Inchingolo

. A new lung ultrasound protocol able to predict worsening in patients affected by severe acute respiratory syndrome coronavirus 2 pneumonia. Journal of Ultrasound in Medicine.2021 Aug; 40(8): 1627-35.

Ilvemark

Wilkens

Thielsen

Dige

Boysen

Brynskov

Bjerrum

Seidelin

. Early intestinal ultrasound predicts intravenous corticosteroid response in hospitalised patients with severe ulcerative colitis. Journal of Crohn’s and Colitis.2022 Nov 1; 16(11): 1725-34.

Baciarello

Bonetti

Vetrugno

Saturno

Nouvenne

Bellini

Meschi

Bignami

. Is lung ultrasound score a useful tool to monitoring and handling moderate and severe COVID-19 patients in the general ward? An observational pilot study. Journal of Clinical Monitoring and Computing.2022 Jun; 36(3): 785-93.

Baciarello

Bonetti

Vetrugno

Saturno

Nouvenne

Bellini

Meschi

Bignami

10.

Blanco

Ferreyra

Badie

Carabante

. Severe massive pulmonary thromboembolism: a case reinforcing the crucial role of point-of-care ultrasound in emergency settings. Journal of ultrasound.2021 Sep; 24(3): 323-9.

11.

Golemati

Yanni

Tsiaparas

Lechareas

Vlachos

Cokkinos

Krokidis

Nikita

Perrea

Chatziioannou

. CurveletTransform-Based Texture Analysis of Carotid B-mode Ultrasound Images in Asymptomatic Men With Moderate and Severe Stenoses: A Preliminary Clinical Study. Ultrasound in Medicine & Biology.2022 Jan 1; 48(1): 78-90.

12.

Wang

Zhang

Yuan

Zhang

Wang

Xie

. Bedside ultrasound in assessment of 510 severe and critical patients with COVID-19 pneumonia in Wuhan, China. Advanced Ultrasound in Diagnosis and Therapy.2020 Apr 17; 4(2): 60-6.

13.

Burkitt

Kang

Jyoti

Mohamed

Chaudhari

. Comparison of cranial ultrasound and MRI for detecting BRAIN injury in extremely preterm infants and correlation with neurological outcomes at 1 and 3 years. European Journal of Pediatrics.2019 Jul 1; 178: 1053-61.

14.

Shivapour

Javed

Brinza

Hornacek

Conic

Gornik

Kim

. Changes in carotid duplex ultrasound velocities after aortic valve replacement for severe aortic stenosis. Journal of Ultrasound in Medicine.2020 Jan; 39(1): 139-45.

15.

Qiong

Tao

Ma-Li

Wen-Dan

Ming

. The Monitoring and Application of Transcranial Doppler Ultrasound in Patients with Severe Neurovascular Diseases. Journal of Sichuan University (Medical Sciences).2020 Jul 20; 51(4): 578-81.

Evaluation of Severe ultrasound and gene diagnosis in cardiac index and shock patient index of shock patients

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Genetic algorithm

2.1 Characteristic gene selection technology of genetic algorithm

2.3 Cardiac index and peripheral vascular RI

2.3.1 Pulse pressure as a substitute for stroke output

2.3.2 Arterial waveform analysis is a qualitative hemodynamic fine-tuning tool

2.3.3 Central venous pressure

3. Experimental evaluation of the role of severe ultrasound in gene diagnosis of shock patients

3.1 Experimental design

3.1.1 Research object

3.1.2 M-CCUE scheme

3.1.3 Experimental analysis

Table 1 Statistical description of baseline data

Data availability statement

Authors’ contributions

Funding

Footnotes

Conflict of interest

References

Table 1
Statistical description of baseline data