Lung cancer is the most common cause of death worldwide. So, early detection is crucial which may be helpful to reduce mortality rate. Low dose computed tomography (LDCT) screening may reduce the lung cancer associated mortality by 20%. But the potential harm due to radiation exposure, false-positive results, and costs associated with LDCT, most organizations only recommend annual screening that targets high-risk individuals. However, due to the potential harm associated with false-positive results, the cost-effectiveness of implementing annual LDCT screening remains controversial. In order to overcome this problem, multiple research groups have attempted to develop risk prediction models to classify patients who might benefit from LDCT screening. In this scenario, the use of artificial intelligence (AI) has resulted in good performance of predicting image related tasks, specifically the use of convolutional neural networks (CNNs). In lung cancer research, CNNs have been applied to LDCT and chest radiographic images to facilitate detection. Also, in predicting lung cancer risk, the electronic medical records (EMR) is suited to the task of identifying high-risk individuals. The goal of the present study was to generate a model that facilitated the prospective identification of individuals at higher risk for developing lung cancer; these individuals might then undergo further follow-up examinations, including LDCT. The use of a predictive model to identify individuals at high risk could serve to limit unnecessary radiation exposure and reduce costs associated with LDCT and related interventions. Methodology The patients EMRs were initially sampled from the Taiwan National Health Insurance Research Database (NHIRD). These EMRs included the demographic information, diagnoses, and procedure codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) and prescriptions from both outpatient clinical declaration files and in-hospital declaration files. This study included participants between the ages of 20 and 90 years who had at least 4 years of medical records on file. Participants with missing data were excluded. The index date for patients with lung cancer was defined as the date of first diagnosis. The inputs included age, gender, and an image representing the patient’s 3-year history of diagnosis and medication. The image was input into Xception, a 126-layer neural network, in which feature extraction was performed. The final layer of the Xception network was connected to an average pooling layer and then connected to a fully connected layer with the patient’s age and gender. A three subgroup analyses was performed to investigate the performance of the model in different populations. According to the age criteria used in previous trials focused on lung cancer screening, patients above and below 55 years of age were included among the subgroups. Also, patients both with and without previous lung disease, including subgroups of patients diagnosed with asbestosis, bronchiectasis, chronic bronchitis, chronic obstructive pulmonary disease (COPD), emphysema, fibrosis, pneumonia, sarcoidosis, silicosis, and tuberculosis were examined. Finally, to focus on the discriminative power of the diagnosis and medication without the confounding effects of age, a subgroup of age- and gender-matched controls was identified. All patient data were split into training, validation, and testing sets based on their respective index dates. To understand the model prediction, occlusion sensitivity analysis was performed by iteratively masking information from a single diagnosis or medication followed by evaluating any changes in the model prediction. In addition, a dimensional reduction technique, t-distributed stochastic neighbor embedding (t-SNE), was performed on the fully connected hidden layer output of the final testing data. The model construction, data preprocessing, model training, and statistical processing were performed using the Python programming language, version 3.6. Results A total of 11,617 lung cancer patients and 1,423,154 control patients were identified in our data set. The mean age of the lung cancer group was 66.62 years; the overall data set included 856,558 (59.7%) men and 578,213 (40.3%) women. For all patients, the model revealed an AUC of 0.821 when the input image-like array included sequential diagnostic information only. By contrast, the AUC was 0.894 when the input features included sequential medication information only; when the sequential diagnostic and medication information was simplified to binary variables, the model performance decreased (AUC=0.827). When both sequential diagnostic and medication information were integrated, the model reached an AUC of 0.902 on prospective testing, with a sensitivity of 0.804 and specificity of 0.837. The model performance at different age cut offs was then investigated. Screening using an age cut off of 55 years revealed a superior AUC of 0.871 compared to those obtained when cut offs of 50 or 60 years were used (0.866 and 0.863, respectively). Analyses of the subgroups included one that was both age and-gender-matched, those at ages above and below 55 years, and those with or without lung disease were performed. This model revealed an AUC of 0.818 with a sensitivity of 0.647 and a specificity of 0.873. For patients above 55 years of age, the model revealed an AUC of 0.869 with a sensitivity of 0.784 and a specificity of 0.785. The discriminatory powers of these models were both excellent among patients with and without a history of lung disease; the AUCs for these subgroups were 0.914 and 0.887, respectively. Among all the subgroups, the model had the weakest performance in patients below 55 years of age who had no history of lung disease. By contrast, the model provided the strongest performance for individuals above the age of 55 years with a history of lung disease, which revealed the highest PPV of 14.3%. The model exhibited the lowest PPV of 1.0% for individuals less than 55 years of age with no history of lung disease. Conclusion CNN model exhibited robust performance with respect to the 1-year prospective prediction of the risk of developing lung cancer. This model may be deployed as a means to analyze medical databases, thus paving the way for efficient population-based screening and digital precision medicine. A future randomized controlled trial will be required to explore the clinical benefit of this model in diverse populations. Impact of research 1. The model could be readily deployed as a means to evaluate centralized health care databases. 2. To perform efficient population-based screening of risk prediction of lung cancer. 3. The similar models may be incorporated for detection of other types of cancer as well.