In order to diagnose and monitor cardiovascular diseases (CVD), blood pressure must be measured in a convenient and uninterrupted manner. Since hypertension is a sudden health problem that does not exhibit symptoms until late stages of the disease, it is the main risk factor for CVD. Researchers tested whether deep neural networks can discriminate between hypertensive and healthy subjects based on photoplethysmography (PPG) recordings, rather than electrocardiograms (ECG). They also avoided manually extracting features from morphological features, as in previous studies. 50 patients’ simultaneous PPG and arterial blood pressure (ABP) recordings were analyzed. With GoogleLeNet, ResNet-18 and ResNet-50 prestrained convolutional neural networks (CNN), the scalogram of PPG segments obtained by continuous wavelet transformation (CWT) was used as input images for classification