This paper studies the conversion of photoplethysmography (PPG) signals into electrocardiography (ECG) signals, aiming to harness the complementary strengths of both techniques in cardiovascular monitoring. PPG, known for its non-invasive nature and ease of use, holds promise for various healthcare applications, including emergency care and wearable devices. However, its conversion to ECG entails several preprocessing steps to enhance signal quality and align key features for synchroniza- tion with ECG signals. To facilitate this endeavor, a hybrid dataset combining publicly available data with a meticulously curated custom dataset is employed, ensuring a robust assessment of the conversion model’s performance. The methodology encompasses various steps such as filtering, alignment, normalization, and segmentation of the dataset, all designed to optimize the efficiency of the conversion process. A pivotal aspect of the approach involves the utilization of a bidirectional long short-term memory (BiLSTM) model, for handling sequential data inherent in physiological signals like PPG and ECG. By integrating regularization techniques into the model architecture, over fitting is mitigated, and generalization performance is enhanced, ensuring reliable predictions. The evaluation framework employs rigorous metrics such as Pearson’s correlation coefficient (r) and root mean square error to quantitatively assess the accuracy and fidelity of the reconstructed ECG signals. The maximum accuracy obtained is \(r=0.951\) . This thorough evaluation not only validates the effectiveness of the conversion process but also provides valuable insights for refining future signal processing methodologies in healthcare. Ultimately, this research contributes to advancing the integration of PPG and ECG technologies, paving the way for enhanced cardiovascular monitoring and diagnosis in clinical settings.