Adolescent suicidal ideation is an escalating concern, underscoring the urgent need to identify and understand the factors contributing to this issue. Our study utilized a hybrid Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) model to analyse the intricate relationships between suicidal ideation and various risk factors, such as depression, anxiety, and social support. By leveraging a substantial dataset of mental health surveys from 3,075 participants, we aimed to uncover patterns and key risk indicators associated with suicidal thoughts among teenagers. The hybrid model, designed to capture both spatial and temporal patterns, analyzed 24 parameters, achieving a remarkable F1-score of 97.8%. This high accuracy highlights the model's potential to identify adolescents at risk of suicidal ideation with precision. The findings emphasize significant correlations between mental health factors and suicidal thoughts, offering deeper insights into the complex interplay of these variables. These results serve as a crucial step toward developing targeted preventive interventions. By identifying adolescents at risk, stakeholders can design more effective support systems and mental health strategies to address the root causes of suicidal ideation.