Using Machine Learning to Detect Risky Behavior in Psychiatric Clinics

Overview

The aim of this study is to ensure the safety of patients in a psychiatric clinic and to detect risky behaviors by using machine learning method. Risky behaviors are defined as behaviors that are personally, socially and developmentally undesirable and endanger life and health.Patient safety and maintaining a safe environment are among the primary duties of healthcare professionals. Suicide is the most important evidence-based risk factor, especially among individuals with psychiatric illnesses, and is one of the most important factors that threaten patient safety. At the end of this study, it is aimed to detect risky behaviors of patients before they harm themselves and to enable healthcare professionals to make early intervention for these behaviors, thus supporting a safe treatment environment, with the computer system that has been trained with the machine learning model installed in the clinics.

The aim of this study is to detect risky behaviors of patients in a psychiatric clinic using machine learning method. Risky behavior; It is defined as behaviors that are personally, socially and developmentally undesirable and endanger life and health. Patient safety and ensuring a safe environment are among the primary duties of healthcare professionals. Suicide is the most important evidence-based risk factor, especially in individuals with psychiatric illnesses, and is one of the most important factors that threaten patient safety. Suicide attempt is a crisis situation frequently encountered in clinics. It is known that the rate of suicide attempts increases 5-10 times in hospitalized patients. In clinics with inpatients, suicide attempts as well as other risky behaviors are frequently encountered. Preventing risky behavior and providing a safe environment in psychiatric clinics is an important issue in our country and in the world. In order to detect risky behaviors and ensure patient/employee safety, there are measures to monitor patients with cameras in psychiatric clinics within the scope of quality standards in health. However, these measures are not sufficient to completely solve the problem. In psychiatric clinics, patient monitoring is provided by a nurse who constantly monitors the camera images placed in the rooms on the computer screen. The low number of nurses, especially on night shifts, makes camera monitoring difficult during night shifts and poses a problem in terms of patient safety. Constant monitoring of monitors by the nurse reduces the time spent with the patient and increases the workload. Additionally, when screen monitoring is not done, risky behaviors cannot be detected. Therefore, new methods need to be developed to ensure a safe environment in psychiatric clinics. In this sense, the machine learning method, which is increasingly used in artificial intelligence and data analysis, is a specialized sub-branch of artificial intelligence algorithms that tries to derive meaningful results/predictions from existing data. Machine learning method is frequently used in the field of health, and psychiatry is one of these fields. The main purpose of this study is to detect risky and high-risk behaviors of patients treated in a psychiatric clinic using machine learning method and to ensure that patients receive treatment in a safer environment. The behaviors that are desired to be detected are risky and high-risk behaviors. The high-risk behavior that is targeted to be detected is an act of suicide through hanging. Risky behavior; Behaviors that include acts of violence such as slapping, pushing, dropping to the ground, pulling hair, pushing against the wall, choking from behind, kicking, putting a pillow on one's face, and struggling. Our primary aim in our study is to detect risky and high-risk behaviors of patients treated in a psychiatric clinic by using machine learning method and to ensure that patients receive treatment in a safer environment. The aim is to send an alert to healthcare workers' phones and to the computer screen in the clinic where the system will be installed. A red alarm with the room number for hanging, which is a high-risk action, and an orange alarm for violent behavior will be sent to both the healthcare worker's phone and the clinic computer screen. Body movements and limb movements will be used in the training of artificial intelligence.Two-dimensional images of behaviors will be created with the Openpose application. Then, a Long Short Term Memory (LSTM) based deep learning model will be created. In the final stage, the success of the model will be evaluated with the F1 score.