Ebook: Computer Methods in Medicine and Health Care

Computer technology has brought about incredible changes in medicine and healthcare, greatly improving the efficiency and accuracy of medical treatment. Since December 2019, in the face of the global effects of COVID-19, the significance of computer technology – big data in particular – the collaborative network and unmanned technology have been recognized by all medical staff. We have to admit that modern medical science cannot evolve without the involvement of computer science.

This book presents the proceedings of the 2021 Workshop on Computer Methods in Medicine & Health Care (CMMHC 2021), the autumn edition of the TDI conferences. The webinar of CMMHC 2021 is scheduled for September 26^th, 2021. Researchers from renowned universities, laboratories and hospitals in China, Italy and Japan have contributed to the workshop. These include Fondazione Politecnico di Milano, CAS Institute of Healthcare Technologies, and University of Chinese Academy of Sciences. Findings from both basic and clinical medicine are included in this collection. Big data technology appeared in 20% of all papers as the most popular topic, with one paper covering big data optimization and two describing its application. It is also inspiring to see researchers from a university specializing in traditional Chinese medicine starting to focus on big data technology.

The organizers and editors of this book are happy to have the opportunity to share practical experiences and enlightening ideas with researchers and practitioners in modern medicine.

The crude incidence of liver cancer ranks top five among all cancers in China, and the death rate ranks the top two. Identifying critical risk factors of liver cancer helps people adjust their lifestyles to reduce cancer risk. Launched in 2012, Early Diagnosis and Treatment of Urban Cancer project has been carried out in major cities of China, which collected a broad range of epidemiological risk factors including definite, probable and possible causes of cancer. We retrieved data from 2014 to the present and obtained 184 liver cancer cases among 55 thousand people. We explored 84 risk factors and implemented liver cancer prediction model with machine learning algorithms, where deep neural network achieved the best performance using non-clinical information (mean AUC=0.73). We analyzed model parameters to investigate critical risk factors that contribute the most to prediction. Using 50% top-ranking risk factors to train a model, the performance showed no significant difference from that using all risk factors. Using top 10% risk factors induced a sensitivity drop and a lower false positive rate. These phenomena prove that the identified risk factors are critical in liver cancer prediction. This work is a reference in public health research, and provides a scientific lifestyle guideline for individuals to prevent liver cancer based on machine learning technology.

With the development of deep convolutional neural network, recent research on single image super-resolution (SISR) has achieved great achievements. In particular, the networks, which fully utilize features, achieve a better performance. In this paper, we propose an image super-resolution dual features extraction network (SRDFN). Our method uses the dual features extraction blocks (DFBs) to extract and combine low-resolution features, with less noise but less detail, and high-resolution features, with more detail but more noise. The output of DFB contains the advantages of low- and high-resolution features, with more detail and less noise. Moreover, due to that the number of DFB and channels can be set by weighting accuracy against size of model, SRDFN can be designed according to actual situation. The experimental results demonstrate that the proposed SRDFN performs well in comparison with the state-of-the-art methods.

The possibility to access healthcare fairly and equally among all the patients can be enhanced with the development of collaborative networks. To achieve their goals and exchange relevant information, they must be combined with a proper digital support. Several works dealing with this aspect can be found in literature; however, works defining a general methodological approach to design a digital solution for a collaborative network were not found. In addition to this, to assess the impact of a pathology network and its digital support, and ensure quality improvement as well as proper clinical outcomes, a suitable panel of key performance indicators (KPIs) should be designed. This paper describes a methodology to design a digital support of a collaborative pathology network, together with a set of KPIs to assess the impact of the pathology network and its digital solution. This approach was specifically applied for the Italian Rare Cancer Network in the context of the project “Italian Rare Cancer Network: Process monitoring and System Impact Assessment”.

In recent years, studies have found that the hierarchical neural network with LSTM network has higher accuracy than another feature engineering. Therefore, this paper first tries to build a multi-stage blood pressure estimation model through VGG19 and LSTM network. Based on the time node of the R wave peak in the QRS waveform in ECG, VGG19 is used to extract various higher-dimensional and rich life characteristics in the PPG signal segment by heartbeat as the unit and focus on processing the dynamics of SBP and DBP Correlation, finally use the LSTM model to extract the time dependence of the vital signs. Results: Experiments show that compared with similar multi-stage models, this model has higher accuracy. The performance of this method meets the Advancement of Medical Instrumentation (AAMI) standard and reaches the A level of the British Hypertension Society (BHS) standard. The average error and standard deviation of the estimated value of SBP were 1.7350 4.9606 mmHg, and the average error and standard deviation of the estimated value of DBP were 0.7839 2.7700 mmHg, respectively.

Allergic rhinitis (AR) has now become one of the major diseases affecting people’s lives, and Traditional Chinese medicine (TCM) always has good efficacy in clinical treatment. In the present study, we analyzed the most frequently used drug pair of Astragalus-Saposhnikoviae Radix (SR) in prescriptions for the treatment of allergic rhinitis by network pharmacology to reveal the modern pharmacological mechanisms of drug prevention and treatment of the disease. Firstly, the 38 active ingredients with good ADME properties from the Astragalus-SR drug pair were collected from the database, and the collated drug targets of Astragalus and SR and the targets of allergic rhinitis were mapped against each other by the network visualization software Cytoscape, followed by the establishment of a “drug active ingredient-target-disease” network diagram and the construction of a high-confidence protein-protein interaction network. Then, the common targets obtained from the disease and drug active ingredients were imported by R language for GO enrichment analysis and KEGG pathway enrichment analysis. The KEGG pathways associated with the targets of Astragalus and SR for the treatment of allergic rhinitis obtained from R enrichment analysis were imported into Cytoscape, and the CytoNCA plug-in was loaded to construct a “target-pathway” network map, and the core target wogonin (FN1) was screened. These evidences suggest that the drug pair of Astragalus-SR works in a multi-component, multi-target and integrated modulation manner for the treatment of allergic rhinitis, which provides an important basis for the treatment of allergic rhinitis.

Preterm delivery is currently a global concern of maternal and child health, which directly affects infants’ early morbidity, and even death in several severe cases. Therefore, it is particularly important to effectively monitor the uterine contraction of perinatal pregnant women, and to make effective prediction and timely treatment for the possibility of preterm delivery. Electromyography (EHG) signal, an important measurement to predict preterm delivery in clinical practice, shows obvious consistency and correlation with the frequency and intensity of uterine contraction. This paper proposed a deep convolution neural network (DCNN) model based on transfer learning. Specifically, it is based on the VGGNet model, combined with recurrence plot (RP) analysis and transfer learning techniques such as “Fine-tune”, marked as VGGNet19-I3. Optimized with the clinical measured term-preterm EHG database, it showed good auxiliary prediction performances in 78 training and test samples, and achieved a high accuracy of 97.00% in 100 validation samples.

“Paradigm” theory is an important ideological and practical tool for scientific research. The research means and methods of Geographic Information Science follow the laws of four paradigms. Automatic cartographic generalization is not only the key link of map making, but also a recognized difficult and hot issue. Based on large-scale map data and deep learning technology, an automatic cartographic generalization problem-solving model is proposed in this paper. According to the key and difficult problems faced by residential area selection and simplification, residential area selection models and simplification models based on big data and deep learning are constructed respectively, which provides new ideas and schemes to solve the key and difficult problems of residential area selection and simplification.

In the fight against New Coronary Pneumonia Epidemic, Chinese Ministry of Health put forward the inevitable requirements of precise policy implementation and scientific epidemic prevention. Accordingly, the big data technology has been applied in the analysis of epidemic dynamic, information inquiry, disease prevention and treatment, and prediction of epidemic trend. And, great success has been achieved in the fight, where the big data technology has played a vital role. This article outlines the main applications of big data technology in the prevention and control of New Coronary Pneumonia Epidemic, and proposes suggestions based on the problems in the application of big data during the epidemic prevention and control period. In the later stage, the integration of big data technology in various fields should be accelerated, information should be further shared and the utility value of data should be maximized.

Continuous blood pressure monitoring is of great significance for the prevention and early diagnosis of cardiovascular diseases. However, the existing continuous blood pressure monitoring methods, especially the sleeveless blood pressure monitoring methods, are complex and computationally heavy. In this paper, we propose a method, using plethysmography (PPG) signals alone, to estimate continuous blood pressure by extracting multiple PPG features related to intravascular blood flow characteristics. The performance of our method was evaluated using ten minutes synchronized PPG signals and continuous blood pressure from 21 healthy volunteers and 19 patients with hypertension and diabetes. The test results have shown that the absolute mean errors and standard deviation errors between the estimated and referenced blood pressure are 3.22±0.66 mmHg for systolic blood pressure and 2.11±1.11 mmHg for diastolic blood pressure, which meet AAMI (the association for the advancement of medical instrumentation) error acceptance.

Targeted at the current issues of communication delay, data congestion, and data redundancy in cloud computing for medical big data, a fog computing optimization model is designed, namely an intelligent front-end architecture of fog computing. It uses the network structure characteristics of fog computing and “decentralized and local” mind-sets to tackle the current medical IoT network’s narrow bandwidth, information congestion, heavy computing burden on cloud services, insufficient storage space, and poor data security and confidentiality. The model is composed of fog computing, deep learning, and big data technology. By full use of the advantages of WiFi and user mobile devices in the medical area, it can optimize the internal technology of the model, with the help of classification methods based on big data mining and deep learning algorithms based on artificial intelligence, and automatically process case diagnosis, multi-source heterogeneous data mining, and medical records. It will also improve the accuracy of medical diagnosis and the efficiency of multi-source heterogeneous data processing while reducing network delay and power consumption, ensuring patient data privacy and safety, reducing data redundancy, and reducing cloud overload. The response speed and network bandwidth of the system have been greatly optimized in the process, which improves the quality of medical information service.

Alzheimer’s disease (AD) has become a major issue around world, including China. The two major challenges for AD are the difficulty in early detection and poor treatment outcomes. Over the past decades, artificial intelligence (AI) was more and more widely used in the prevention, diagnosis and treatment of AD, which might be helpful to deal with the aging of population in China. Here, after a systematic literature searching on three English databases (MEDLINE, EMBASE, the Cochrane library), we briefly reviewed recent progress on the utilization of AI in the susceptibility analysis, diagnosis and management of AD. However, it is still in its infancy. More researches should be performed to improve the prognosis of patients with AD in the future.

The results of previous studies showed that ECG could detect CHD in children with a detection rate of 76.43%. Although this result is better than the traditional CHD screening method, the sensitivity still needs to be improved if it is to be popularized clinically. Based on the previous ECG recording data, this study selects the more representative cardiac cycle segments to identify CHD, in order to achieve better screening effect. Firstly, better cardiac cycle segment data were extracted from ECG records of each patient. The final data set contains 72626 patients and each patient has a 9-lead ECG segment with duration of about one second. Then we trained a RoR network to identify the underlying patients with CHD using 62626 samples in the dataset. When tested on an independent set of 10000 patients, the network model yielded values for the sensitivity, specificity, and accuracy of 0.93, 86.3%, 85.7%, and 85.7% respectively. It can be seen that extracting more effective cardiac cycle fragments can significantly improve the sensitivity of CHD screening on the basis of ensuring better specificity, so as to find more potential patients with congenital heart disease.

Alzheimer’s disease (AD) is a degenerative disease of central nervous system, which seriously threatens the life and health of the elderly people. It has been for long time that Traditional Chinese medicine (TCM) treatment for AD is effective. This study analyzed the potential target and molecular mechanism of the most often used drug pair of Astragalus membranaceus and Acorus tatarinowii to treat AD by network pharmacological method. Firstly, the method was performed to screen and sort out the active ingredients with good ADME properties and drug targets of Astragalus membranaceus and Acorus tatarinowii. Then, we searched for the disease targets related to AD, followed by the construction of the “active ingredients-target-disease” network. We implemented GO enrichment analysis and KEGG pathway enrichment analysis of related overlapped target proteins, and then constructed the “target-pathway” network diagram. Finally, the above overlapped target proteins are mapped to build a PPI high-position protein interoperability network, and we conducted the network topology analysis to screen out the core targets of Astragalus membranaceus-Acorus tatarinowii drug pair in the treatment of AD. According to network pharmacology, this research predicted the potential targets of the drug pair of Astragalus membranaceus and Acorus tatarinowii in the treatment of AD, and explored that Astragalus membranaceus-Acorus tatarinowii drug pair in the treatment of AD was the overall systematic regulating action of “multiple-ingredients, multiple-target and multiple-pathway”. It affords the reference for understanding the pathogenesis of AD and exploring new therapeutic methods and drug development in the future.

Map is not only the result of geospatial environment cognition, but also a tool for geospatial environment cognition. The new concept advocated by Constructivist cognitive theory is highly consistent with the concept of map service in the era of Internet plus space-time big data. This paper analyzes the geographic information transmission process from the perspective of constructivism, and constructs the geographic information transmission process model. Based on the traditional map cognitive process model, a map cognitive process model based on constructivism is constructed. According to the four elements of “situation, cooperation, communication and meaning construction” advocated by Constructivist cognitive theory, a map service function model based on constructivism is constructed.