AI helps Helicobacter pylori detection, smartphone assisted system improves diagnostic efficiency

2022-07-28

Recently, a study published in PMC showed that researchers have developed a smartphone assisted Helicobacter pylori detection system based on deep learning. The system uses Faster R-CNN and ResNet 50 algorithms to analyze 270 gastric biopsy specimens, with a detection accuracy of 89.23%. By connecting microscopes, smartphones, and AI algorithms through 5G networks, real-time detection prompts can be pushed to doctors, effectively reducing missed diagnoses caused by pathologists' fatigue (the traditional manual detection missed diagnosis rate is about 14%). This system can be used in conjunction with traditional microscopes without adding additional operational steps, and is expected to become an efficient auxiliary tool for Helicobacter pylori screening, improving diagnostic consistency and accuracy.

Previous: