Application of Drug Coated Balloon Using Lightweight Deep Learning Model on Coronary Bifurcation Lesions in Patients with Coronary Artery Disease: A Medical Engineering Technology

We aimed to evaluate the drug coated balloon (DCB) based on lightweight deep learning model on coronary bifurcation lesions in patients with coronary artery disease (CAD). A lightweight deep learning model was applied to coronary angiography in patients. According to the different use of balloon after pre-treatment of branch vessel cutting balloon, the patients were assigned to observation ( n = 45) and control ( n = 45) groups. In terms of demographic data, no significant difference was observed between the groups ( P >0·05). Similarly, no intergroup difference of statistical significance was found in Medina classification, reference diameter and minimum diameter of main branches ( P >0·05). Following a 9-month period of follow-up, a significantly lower rate of adverse reaction (4.44%) was observed in the observation group, as opposed to the control group (17.78%) ( P <0·05). While there was a postoperative decrease in the extent of vascular stenosis in both groups, the extent to which vascular stenosis was reduced in the observation group was markedly higher ( P <0·05). In the results of branch re-examination diameter confirmed an intergroup difference of statistical significance ( P <0·05). The duration of exposure to X-rat, operation time, and average dosage of contrast agent was substantially lower in the observation group ( P <0·05). DCB active protection technology can reduce the extent of vascular stenosis in CAD.