Research on Data Optimization of CT System for Patients with Severe Pneumonia Based on Filtered Back Projection and Internet of Things

CT technology can clearly and intuitively display the internal structure, composition, material, and defect status of the detected object under the condition of no damage to the detected object. This paper is based on the filtering back projection and the Internet of Things technology to carry out the data optimization of the CT system of patients with severe pneumonia. First of all, the Internet of Things technology is used in the structural design and operation process of the CT system for patients with severe pneumonia, which improves the advancement and convenience of the original system. Secondly, the system data is optimized through the filtering back projection algorithm, which improves the data accuracy and transmission efficiency. Finally, simulation tests have verified the effectiveness of the established CT system data optimization method for patients with severe pneumonia, which is helpful for medical comparison, analysis, and diagnosis.


Introduction
Computed tomography (CT) is a medical diagnosis method that combines electronic computer technology and X-ray imaging technology. Due to the rapid development of electronic computer technology in recent years, this method has become one of the most common medical clinical diagnostic methods [1,2]. Severe pneumonia is a common respiratory disease. The various organs of the elderly are gradually failing, the immune system is weakened, and they are more likely to suffer from severe pneumonia [3]. The highresolution CT image can accurately detect the subtle differences in the radiation attenuation characteristics between the tissues on the cross-section of the human body to be measured and display them in the form of digital images and can finely distinguish the different densities [4]. Therefore, CT technology is widely used in cross-sectional tissues for medical purposes [5]. The disease has the characteristics of insidious onset, critical illness, and rapid progress. The initial symptoms are local inflammation gradually worsening to systemic inflammation, causing severe infectious compli-cations such as septic shock, sepsis, and multiple organ dysfunction, which poses a great threat to the life safety of patients and requires effective treatment intervention as soon as possible [6,7].
Although the above-mentioned medical imaging modes have a wide range of important applications, due to factors such as the performance of their own equipment and external conditions, medical images still have many unavoidable defects [8,9]. CT technology can clearly and intuitively display the internal structure, composition, material, and defect status of the inspected object without damaging the inspected object [10,11], for example, the image is blurred or the image boundary is not obvious. In addition, in the process of forming and transmitting medical images, its image quality will also be affected to a certain extent [12]. All these have caused certain difficulties and obstacles to doctors' image interpretation and accurate diagnosis. In order to enhance the quality of medical images, enable doctors to observe and analyze the lesions and anatomical structures more clearly, and improve the accuracy of diagnosis, postprocessing of medical images becomes necessary [13,14]. This paper is based on the filtering back projection and the Internet of Things technology to carry out the data optimization of the CT system of patients with severe pneumonia. The established CT system data optimization method for patients with severe pneumonia is helpful for medical comparison, analysis, and diagnosis.

CT Tomography Technology
CT is computed tomography. CT imaging uses the attenuation characteristics of X-rays as a diagnostic basis to scan through a certain thickness of a certain part of the human body [14,15]. The high-sensitivity receiver receives the Xrays that pass through this layer and converts them into visible light, which is converted into electrical signals by photoelectric conversion, and then converted into digital signals by an analog-digital converter. Each number in the digital matrix is converted into small squares from black to white (i.e., pixels) by a digital-to-analog converter and arranged into a matrix to obtain CT images [16,17]. After entering the computer, the selected layer is divided into multiple voxels, and the scanned information is calculated to obtain the X-ray attenuation coefficient or absorption coefficient of each voxel and arranged in a digital matrix.
X-rays have varying degrees of penetrating ability to various substances. When a beam of X-rays passes through the human body, it will be attenuated due to absorption and scattering [18,19]. However, due to the different density and atomic composition of various tissues of the human body, the degree of absorption and scattering is also different, that is, different tissues of the human body have different attenuation coefficients for X-rays [20,21]. In general, when X-rays with the same energy are used for incidence, the higher the density of human tissues, the greater the attenuation of X-rays, and the lower the density of human tissues, the lower the attenuation of Xrays. CT images express the absorption degree of X-rays by organs and tissues in different gray scales, which is consistent with X-ray imaging. The black areas indicate low absorption areas, i.e., low-density areas, such as lungs; white areas indicate high absorption areas, i.e., high-density areas [22,23].
However, compared with X-ray images, CT is completely dependent on the projection data of the attenuation value of the object in the inspected layer. It has nothing to do with the structure of the inspected layer. It is a complete tomographic image with more accurate layer thickness and clearer image. The density resolution is higher. In addition, the tomographic images obtained by CT scanning can also be processed by computer software to obtain multidirectional cross-sectional images, such as coronal images and sagittal images, for easy observation [24]. The above characteristics provide an important guarantee for the research to obtain clear, high-resolution, and economical geometric information of cerebral aneurysms. The reconstructed image can be displayed on the monitor in different Di levels or stored in the computer hard disk in digital form.

Overview of IoT Technology
The Internet of Things has entered all aspects of people's lives, and as a technical support for economic development, it has shown great potential as a promoter of world economic growth in the 21st century [25,26]. The Internet of Things is a systematic fusion of multiple technologies. To understand the Internet of Things, we should start from its system architecture and fully understand it to grasp the overall situation from the top level and lay a solid foundation for specific application design [27,28]. A sensor conduction network is composed of probes, remote sensing components, and various terminal devices [29,30]. The sensing layer is composed of different forms of sensing components, QR codes, RFID tags, and barcode readers distributed in various building groups. A schematic diagram of IoT technology architecture is shown in Figure 1.
The access network is a fusion of the communication network and the Internet composed of various wired and wireless nodes and fixed and mobile gateways and can be accessed through the Internet, the broadcasting network, and the communication network. Different from classic information systems, IoT requires large-scale data access and transmission capabilities [31][32][33]. The role of the middleware layer is to shield heterogeneity, achieve two-way interaction, and data preprocessing. At the same time, various system platforms and technical resources with different advantages are integrated and shared to form a multifunctional integrated system with diverse resource allocation and complete functions. The middleware layer can provide a variety of services to the upper layer, integrate multiple applications on the basic system, and provide differentiated services for different user needs.

Optimization of CT System Data for
Patients with Severe Pneumonia 4.1. Filtering Back Projection Algorithm. Interpolation here refers to a model-based method used to estimate continuous data from discrete data within a known coordinate range. Classical linear interpolation techniques include nearest neighbor interpolation, bilinear interpolation, and bicubic interpolation [34,35]. In the back-projection process of the FDK algorithm, due to the discreteness of the data, the projection address of the pixel will appear "misaligned," which usually requires interpolation, especially for the edges of the image. The nearest neighbor interpolation will produce blackness and jaggedness, while bilinear interpolation and bicubic interpolation will blur the edges of the image. Therefore, researchers have introduced many nonlinear interpolation methods, such as local adaptive interpolation methods.
In an ideal situation, the projection is the integral of the specific parameters of the object along a straight line passing through the object, and the measured object is expressed as a two-dimensional function, that is, the angle between the ray normal and the light.

Wireless Communications and Mobile Computing
According to the conversion relationship between polar coordinates and coordinates, β = arctan ðy/xÞ and r = x 2 + y 2 , the formula can be rewritten as A schematic diagram of the filtered projection geometric ray is shown in Figure 2. According to the geometric properties of the parallel projection, after the ray scan, the complete image of the original image should be obtained, and a series of projection data are measured [36]. The projections of all rays passing through a certain pixel are accumulated in the projection domain. This pixel can be reconstructed: The geometric relationship between these pixels is shown in the figure. According to the geometric relationship, the angle between the ray and the pixel is A schematic diagram of the filtered projection polar coordinate geometry is shown in Figure 3. It can be seen from the above formula that the projection length is no longer dependent on this angle, so the area to be solved for the image is transformed from the plane with XY as the horizontal and vertical coordinates to the polar coordinate plane with the length as the rotation angle [37]. In the process of CT image reconstruction, it is first converted from the XY coordinate domain to the polar coordinate domain and then reconstructed [38].

Optimization of CT System Data for Patients with Severe
Pneumonia Based on Filtered Back Projection. What is obtained after reconstructing the image in the polar coordi-nate domain is not the final result. It is necessary to use the coordinate conversion relationshipx = r cos α and y = r sin α and map it to the XY coordinate system to truly reflect the original image. MATLAB software is used to run the program to get the image shown in Figure 4.
In this paper, after processing the absorption matrix with MATLAB, the absorption rate image of the object can be obtained, as shown in Figure 1. The figure shows that the object is uniform. When X-rays penetrate an object, the rays  3 Wireless Communications and Mobile Computing will be absorbed and scattered by the material, which will weaken the intensity of the X-rays. It can be known from Beer's theorem that when a single light is emitted, the attenuation of the X-ray of the object in a single uniform material will change, and the attenuation of its intensity satisfies formula (1).
Assuming that the test object is segmented and uniform, the linear attenuation coefficients of each segment are as shown in formula (2): Take an ideal ray source, which emits extremely thin pen beam rays, and place a detector on the opposite side. Measure the ray receiving signal g and the intensity of the rays reaching the detector after being attenuated by the object, thus obtaining a set of data. Then, rotate at an angle to obtain another set of data under the new angle. Repeat this until 180 sets of data are obtained after 180 rotations. The fitting diagram of the filtered projection received information data is shown in Figure 5.
According to Radon transformation, suppose the absorption function of a beam of parallel light is shown in formula (3): Among them, Ω is the medium, and the attenuation rate of the intensity is dI/I = −f ðx, yÞdu. Formula (4) is solved: The relationship between the coordinate system X-Y established with the pallet as the reference system and the coordinate system s-u established with the detector as the reference system is shown in From the data fitting diagram of the received information of the template, it can be seen that the curve in the figure is approximately fitted to a sine function. According to the symmetry of the sine function, the abscissa of the origin can be calculated as 62, and the abscissa of the lowest point of the sine function is 144. Corresponding to the distance between the abscissas of the two points, the radian corresponding to the entire curve can be calculated, that is, the entire CT system rotates counterclockwise around the center of rotation, and a total of 180 sets of received information are obtained. It can be calculated that the CT system receives information every time it passes.
Calculate the expression of the sine function, and then obtain ϕ = 3π/4, that is, the initial angle is 3π/4, so the 180 directions of the X-ray used by the CT system are135 ∘ + 1:0980N ∘ , where N represents the number of rotations.

Optimization of CT System Data for
Patients with Severe Pneumonia 5.1. Data Source and Test Environment. In order to verify the effectiveness of this method on CT images, a simulated severe pneumonia CT image was used as a test sample, the similarity  Wireless Communications and Mobile Computing window radius was selected as 3, the search window radius was selected as 7, and the smoothing coefficient was selected as 0.2, and the algorithm comparison experiment was performed.
On the basis of the description of the filtering back projection algorithm, the Radon inverse transformation is performed on the projection image generated by the data again, and the filtering back projection algorithm is established. The sample image is added with Gaussian noise inter-ference with a mean value of 0 and a variance of 0.01 before proceeding. In terms of noise processing, we use the filtered back projection algorithm, the DCT self-similar algorithm and the DCT threshold de-coefficient self-similar algorithm. The result of comparing the denoising effect is shown in Figure 6, which compares the peak signal-to-noise ratio and the signal-to-noise ratio. In addition, the experimental results of the three result images are shown in Figure 7. After acquiring a digital image, the first thing to do is to preprocess the

Wireless Communications and Mobile Computing
image to improve the image quality or reduce the amount of calculation. The specific content of image preprocessing depends on the specific system requirements. Edge extraction first detects the discontinuities of the local characteristics of the image and then connects these discontinuous edge pixels into a complete boundary. In the image, the areas where the gray level changes drastically, that is, the intensity discontinuity, correspond to the edges. Edge extraction is to detect the intensity discontinuities and to determine their precise position in the image. It can be seen that the peak signal-tonoise ratio and signal-to-noise ratio of the simulated CT image obtained by the DCT self-similar algorithm and the DCT threshold decoefficient self-similar algorithm are larger than those of the original algorithm. Therefore, through experimental simulation comparison, we believe that the improved filtered back projection algorithm has better results in CT image processing.

Conclusion
CT technology can clearly and intuitively display the internal structure, composition, material, and defect status of the inspected object without damaging the inspected object. In order to improve the quality of medical images, enable doctors to observe and analyze lesions and anatomical structures more clearly, and improve the accuracy of diagnosis, postprocessing of medical images becomes necessary. Based on the filtering back projection and the Internet of Things technology, this paper optimizes the data of the CT system of patients with severe pneumonia. The established data optimization method of the CT system for severe pneumonia is helpful for medical comparison, analysis, and diagnosis. Although the above-mentioned medical imaging modes have extensive and important applications, medical images still have many unavoidable defects due to factors such as the performance of their own equipment and external conditions. For example, the image is blurred or the image boundary is not obvious. In addition, in the process of forming and transmitting medical images, the image quality will also be affected to a certain extent. We will further carry out research in related fields to improve the data optimization level of the CT system.

Data Availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethical Approval
This article does not contain any studies with human participants or animals performed by any of the authors.

Conflicts of Interest
The authors declare that they have no conflicts of interest.