SADIAX, an innovative computer-vision-based solution for monitoring patients in hospital rooms

Abstract

According to data from the National Registry of Health Establishments (CNES), Brazil has almost 41,000 Intensive Care Unit (ICU) beds. Half of them were designed to serve about 204 million Brazilians through the SUS, while the other half is reserved for private health, to serve up to 50 million people. In São Paulo, the number of beds offered corresponds to 30% of the total number of beds in Brazil. In the United States, more than 5.7 million adults are admitted to the ICU each year, costing the healthcare industry more than $67 billion a year. When it comes to hospital beds in general, this number jumps to 430,000 in Brazil, 64% of which are private. The quality of hospitalization and the length of stay of patients are essential factors for both the patient and the health institution to assess the service experience and financial sustainability. Monitoring patients in beds is a critical activity to improve these factors. When not well monitored, the patient may have a prolonged stay and an increased risk of mortality. Monitoring systems with sensors or not are used to ensure control of patients' vital signs. Critical behaviors such as agitation that can lead to falls or prolonged sedation in the same position leading to a risk of pressure ulcer generation, for example, are evaluated manually and sometimes subjectively. The autonomous detection of these behaviors emerges as a major challenge to be overcome, both for the protection and safety of patients and for the financial and legal protection of hospitals. Controlling agitation, risk of falling and sedation is vital for patient safety. From 20 to 30% of cases of falls result in an injury. And as for pressure ulcers, the prevalence of these wounds in hospitals in the United States ranges from 3% to 14%, while in cases where patients are in absolute rest this number rises to 15% to 25%, where the cost The estimated cost of a treatment is US$ 2,000 to US$ 30,000/patient, reaching US$ 8.5 billion in total. Therefore, detecting the position of the patient's body, to signal the probability of a pressure wound or agitation to avoid falls is fundamental. In order to minimize or extinguish the error in the detection of these behaviors, several investments in technologies and research have already been carried out using body sensors, sensors in mattresses and, more recently, cameras. However, the lack of precision has prevented these technologies from being commercially exploited on a large scale. During PHASE 2, SADIAX not only became the first fully autonomous patient monitoring system for predicting falls and pressure injuries, but began to add privacy monitoring functionality (health professionals), contributing to better auditing. scale and workload analysis. SADIAX reached the end of PHASE 2 with its prototype sold under a pre-sales contract for deployment in more than 100 beds. In PHASE 3, the finalization of the prototype will be sought, focusing on aspects that will allow the solution to be integrated into the most diverse hospital systems, scale to commercialization, increase precision so that it can run directly on hospital platforms, research and refine screen design for better interaction between professionals. The goal of the final solution is to increase the health professional's control over the patient's condition, increase the comfort level and quality of hospitalization of patients, decrease the number of falls in beds and Ulcer Pressure in a scalable and commercial way. (AU)