Risk Prediction and Risk Factor Analysis in Patients with Chronic Obstructive Pulmonary Disease Complicated with Heart Failure under Lightweight Deep Learning Model and Echocardiography

Abstract This work was developed to explore the risk prediction and risk factor analysis of chronic obstructive pulmonary disease (COPD) patients with heart failure (HF) based on a lightweight deep learning (LWDL) model and echocardiography. A total of 127 patients with COPD in Affiliated Hospital of Weifang Medical University were recruited. According to whether they were complicated with HF, they were assigned to the group with HF (n = 35) and the group without HF (n = 92). A dual-channel lightweight model with multifeature fusion was constructed for ultrasound imaging. The blood index, arterial blood gas index, and pulmonary function index of the two groups were analyzed to explore the risk factors for COPD complicated with HF. The results showed that the recognition rate of different methods was above 85%, which indicated that the network had good performance. PaCO2 and PaO2 in the HF group were drastically superior to those in the non-HF group (P<0.05). The proportion of GOLD1 and GOLD2 in the HF group was markedly inferior to that in the non-HF group, and the proportion of GOLD3 and GOLD4 in the HF group was higher than that in the non-HF group (P<0.05). The vital capacity (VC), forced VC (FVC), and forced expiratory volume in 1 s (FEV1) in the HF group were lower than those in the non-HF group (P<0.05). The AO, left atrial dimension (LA), left ventricular end diastolic diameter (LVEDD), left ventricular posterior wall thickness (LVPW), and interventricular septal thickness (IVS) showed no great differences, but the pulmonary artery (PA) and right ventricular end-diastolic diameter (RVEDD) were dramatically superior to those in the non-HF group (P<0.05), and PaP was notably higher in the HF group than in the non-HF group (P<0.05). Moreover, hypercapnia was associated with COPD combined with HF. In summary, the LWDL model can clearly present ultrasound images. Hypercapnia had a great impact on COPD combined with HF, which should be controlled as early as possible in clinical practice to prevent disease deterioration.