1. Introduction {#sec1} =============== Sedation is a common medical procedure used by anesthesiologists, which helps patients to tolerate uncomfortable situations such as pain, fear, anxiety, and/or nausea or vomiting \[[@B1]\]. Sedation is used by anesthesiologists to provide safe and satisfactory anesthesia for patients. Besides, it can also reduce the discomfort of endoscopic procedures for patients. There are several types of sedative drugs for medical purpose. Fentanyl, propofol, and midazolam are most commonly used in sedation \[[@B2]\]. In general, sedative drugs are administered intravenously by anesthesiologists, and the patient experiences moderate sedation. Many sedatives may cause serious adverse reactions, especially on central nervous system. Besides, they can also result in other side effects. Previous studies \[[@B3], [@B4]\] have shown that the risk of respiratory depression induced by anesthetic agents is higher than the sedative drugs. In the operating room, anesthesiologists in combination with medical personnel such as nurses play an important role. They are responsible for monitoring the vital signs of patients as well as the clinical information of the procedures. However, as important tasks, monitoring the vital signs and collecting patient information could be time-consuming. It can divert the concentration of anesthesiologists, and it can also make them neglect the clinical information of the procedures. Due to this, monitoring and collection of information by anesthesiologists have become difficult tasks. Therefore, a tool that can provide timely and good quality information related to sedation of patients is required to improve the efficiency of the process. With the fast development of computer technology, patient monitoring system has been developed in recent years. At present, with the emergence of artificial intelligence (AI), we are more likely to take advantage of various kinds of intelligent techniques to assist medical personnel to provide better treatment. Therefore, it is essential to adopt advanced intelligent techniques to the medical domain. For this purpose, the clinical information of sedation can be automatically collected by a software tool via Internet and processed. As a result, the information will be provided by this software tool in a simple and convenient way. For example, it can provide the vital signs of patients before and after the drug administration. In the future, such clinical information will be used by healthcare practitioners in the operation room to estimate the clinical conditions of patients and make clinical decisions more accurately and timely. In general, AI techniques have been used in clinical domain to help healthcare practitioners to achieve better diagnoses, to monitor the vital signs of patients, and to enhance patient safety. Nowadays, researchers \[[@B5], [@B6]\] have proposed a new research paradigm named as patient-centered care to promote the interprofessional collaborative teamwork. AI technologies were then widely used for patient monitoring by healthcare practitioners. For example, Zhang et al. \[[@B7]\] developed a hybrid intelligent system for patient monitoring. The system has been designed for monitoring various body signs related to diseases. In their work, a neural network model is used to construct a predictive model of patient\'s vital signs and clinical conditions. Such a model can help healthcare practitioners to understand the condition of patients and provide better treatment. By using the technology, healthcare practitioners can also monitor vital signs of patients in real time. Thus, better patient monitoring is achieved. Another example is the development of a patient monitoring tool based on intelligent techniques \[[@B8], [@B9]\]. This tool has been designed to record clinical information, such as vital signs and ECG records, to assist health practitioners to understand the conditions of patients. Such a tool can help the health practitioners to monitor patients, even if they are physically distant from the patients. In order to improve the efficiency of sedation monitoring and collection of information by anesthesiologists, we use intelligent techniques to design a tool. By using this tool, anesthesiologists can obtain the clinical information of patients during the procedures. Moreover, it can also assist the anesthesiologists to understand the patient\'s conditions by monitoring the vital signs. In this paper, an intelligent sedation monitoring tool has been proposed for automatic monitoring of the vital signs of patients during sedation. To implement the monitoring task, a cloud server was used to access patient information and vital signs. Furthermore, a data mining approach was used to build a predictive model for identifying the normal vital sign status of patients. The main contributions of this paper can be summarized as follows:To construct a predictive model for identifying the normal vital sign status of patients, we used the data mining approach. This approach has the ability to extract the knowledge automatically from data, as well as to handle various types of missing data.The tool was designed to collect patient information related to procedures and vital signs, for monitoring sedation procedures. With regard to patient information, various types of clinical information such as patient ID, patient\'s age, sex, operation and surgical location, drug name, drug dose, and so forth can be automatically collected from our tool via Internet.With respect to vital signs, the systolic blood pressure, diastolic blood pressure, heart rate, respiratory rate, and oxygen saturation were monitored. In this way, anesthesiologists can provide better and safe treatment for patients.The algorithm for model construction was validated by the clinical data, and it can also be used to estimate the clinical conditions of patients. The remainder of this paper is organized as follows. [Section 2](#sec2){ref-type="sec"} briefly reviews related works on intelligent technologies used in medical domain. The proposed intelligent sedation monitoring tool has been described in detail in [Section 3](#sec3){ref-type="sec"}. [Section 4](#sec4){ref-type="sec"} discusses the experiments. [Section 5](#sec5){ref-type="sec"} introduces the conclusions and future work of this paper. 2. Related Work {#sec2} =============== With the advancement of deep learning, AI techniques have been widely used in many domains. The application of AI in healthcare was started in 1997 \[[@B10]\]. Nowadays, it has become one of the important techniques in the health industry \[[@B11]\]. AI has also been used to help healthcare practitioners to monitor patients during the procedures. For example, AI has been adopted to monitor medical procedures and provide better treatments. Such a tool could assist physicians to monitor the vital signs of patients and patients\' condition during the procedure \[[@B12]\]. Other related work related to our approach could be found in \[[@B13]--[@B15]\]. In this section, we are going to briefly review two kinds of widely adopted AI techniques: knowledge discovery in database (KDD) and machine learning (ML). In the KDD technique, the knowledge discovery process has been adopted to discover hidden patterns from databases and knowledge bases. This technique has the ability to extract the knowledge automatically from data and to handle various types of missing data \[[@B16]\]. Such techniques have been widely used in medical domain. For example, the AI technique was used for managing the patients in the waiting list \[[@B17], [@B18]\]. Besides, ML is another technique widely used for predicting complex human behavior patterns and predicting patients\' vital signs. This technique has been widely applied in medical domain. For example, the heart disease, hypertension, and hyperlipidemia have been predicted from the vital signs of patients \[[@B19]\]. Moreover, a system using ML technique has also been proposed for monitoring patients with diabetes \[[@B20]\]. 3. Sedation Monitoring Tool {#sec3} =========================== This section presents the details of the sedation monitoring tool. We describe how to monitor patients using the sedation monitoring tool, which has the ability to collect patient information as well as monitor the vital signs of patients. 3.1. Patient Information Collection {#sec3.1} ----------------------------------- The patient information collected by the sedation monitoring tool contains patient information related to procedures and vital signs. This section mainly discusses how to automatically collect vital sign data from patients. The collected information includes a variety of vital signs of patients. For example, the information of patient ID, operation, and surgical location can be used to identify the patient who requires sedation. Patient information such as vital sign records, surgery records, and medication order information can be collected from the tool. These patient information can be used to estimate the clinical conditions of patients. The details of this information collection can be found in [Table 1](#tab1){ref-type="table"}. 3.2. Vital Sign Monitoring {#sec3.2} -------------------------- This subsection is divided into two subsections for describing how to monitor the vital signs of patients before and after the drug administration. During the sedation process, patients need to be monitored because their vital signs might be affected by the medication. For example, some of the drug, such as propofol, can affect the cardiovascular system of patients. Some of the drug, such as opioids, may cause respiratory depression \[[@B21], [@B22]\]. Therefore, it is important to monitor and record the vital signs of patients in real time. In this way, anesthesiologists can recognize the signs of respiratory depression and notify healthcare practitioners or family members of the need for appropriate treatment. 3.3. Patient Monitoring {#sec3.3} ----------------------- This subsection is divided into two subsections to describe how to monitor patient vital signs after the drug administration and how to notify medical practitioners if any abnormal vital sign is detected. In the sedation process, the vital sign measurements of patients are periodically recorded in real time. We describe how to monitor patient vital signs after the drug administration. The details of this step can be found in [Table 2](#tab2){ref-