Treffer: Monitoring and data acquisition of automated non-invasive blood pressure reading with ESP32-CAM and YOLO algorithm.

Blood pressure monitoring is essential for treatment of patient with hypertension and hypotension symptom. However, the displayed systolic, diastolic, and the pulse rate are often written manually by the patient at home or nurse and doctor during their visit to the patient at the hospital. This paper discusses the development of Non-Invasive Blood Pressure (NIBP) device reading system with the use of ESP32-CAM and YOLOv5 (You Only Look Once) algorithm for NIBP digit recognition. For the development of the model, datasets consist of 431 images of NIBP device reading were split and used for training (392 images) and validation (39 images). The label of each digit from digit class 0 until digit class 9 were annotated with the use of Roboflow website and trained by using Python programming language in Google Colab platform. The result shows that the YOLOv5 model could recognize all the displayed digits in the NIBP device testdataset with mean average precision of 99,5%, precision 99,5%, and recall 99,6% from 39 validation images in which consists of total 245 labels for all digits from 0 to 9. Comparison results with the traditional OCR method shows that the trained YOLOv5 model has obtained 100% detection accuracy, while the OCR method obtained 73.06% for the same NIBP digit test dataset. The result also shows that the proposed system can be integrated with Internet of Things (IoT) technology to conduct automatic data acquisition and cloud monitoring of blood pressure data with ThingsBoard IoT platform. Automated alarm system with IoT system can also be created to notify the hospital by email, in order to requestambulance delivery immediately to the patient's house if the blood pressure value of the patient is critical. [ABSTRACT FROM AUTHOR]

Copyright of AIP Conference Proceedings is the property of American Institute of Physics and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)