แผลจากการรั่วซึมของยาออกนอกเส้นเลือดเป็นปัญหาที่ร้ายแรงของผู้ป่วย การตรวจหาแผลจากการรั่วซึม ของยาดังกล่าวจึงมีความสําคัญสําหรับการป้องกันอาการที่จะรุนแรงขึ้น ดังนั้นการวิเคราะห์แผลจากภาพดิจิทัลจะช่วยป้องกันการเกิดความรุนแรงเบื้องต้นได้ บทความนี้นําเสนอการประยุกต์โครงข่ายประสาทเทียมแบบคอนโวลูชัน(CNN)สําหรับระบุลักษณะของแผลจากการรั่วซึมของยาออกนอกเส้นเลือด เพื่อทําการคัดแยกลักษณะของแผล 2 ประเภท ได้แก่ ระดับปานกลาง (Moderate) และระดับมาก (Severe) ภาพสําหรับการสร้างโมเดลใช้ภาพขนาด 10x10 นิ้ว จํานวน 126 ภาพ และใช้ภาพสําหรับการทดสอบ จํานวน 61 ภาพ โดยแต่ละภาพมีผู้เชี่ยวชาญเป็นผู้ระบุ ลักษณะของภาพ โมเดล CNN ประกอบด้วย ReLU เลเยอร์ MAXPooling เลเยอร์ และ ADAM เลเยอร์ ผลการทดสอบ ของการใช้ประยุกต์ CNN สําหรับการระบุลักษณะของแผลมีความแม่นยําร้อยละ 98.36%
A wound from drug extravasation outside the vein is a serious problem for patients. Detecting the wound from the above problem is important for preventing further severe symptoms. Therefore, analyzing the wound from digital images can help to prevent its severity in the first place. This article presents the application of Convolutional Neural Network (CNN) for identifying a characteristic wound from drug extravasation outside the vein to classify the characteristic of two wound types: moderate and severe. The wound images, 10×10 inches, for training are 126 images, and for testing are 61 images. Each image is characterized by the experts. The CNN model includes the ReLU, MAXPooling, and ADAM layers. The experimental result of the CNN application for identifying the characteristic of wound has an accuracy of 98.36%.