Abstract Background: Methylation-based analysis of cell-free DNA (cfDNA) has emerged as a key technology for MCED. However, existing approaches rely on traditional machine learning algorithms, which inherently limit detection performance. With the rapid advancement of artificial intelligence (AI), we have developed Genie-ADLA, a deep learning algorithm designed specifically for MCED. By integrating state-of-the-art deep neural network architectures with the intrinsic patterns inherent in methylation data, Genie-ADLA significantly enhanced MCED performance. Methods: Genie-ADLA was trained and evaluated on a dataset of 4,781 participants aged 40-75 years, including 2,702 pathologically confirmed cancer cases across 16 cancer types and 2,079 non-cancer controls (NCT06217900). The training set comprised 3,217 samples (1,756 cancer cases and 1,461 non-cancer controls), and the model’s performance was evaluated on an independent test set of 1,564 samples (618 non-cancer controls and 946 cancer cases). To address challenges inherent to methylation data—high dimensionality, sparsity, and noise—we applied feature dimensionality reduction and embedding strategies, reducing computational burden, mitigating overfitting, and improving learning efficiency. An ensemble learning approach further strengthened robustness and generalization. Results: Across all stages of 16 cancer types, Genie-ADLA achieved an overall sensitivity of 63.43% (600/946, 95% CI: 60.26%, 66.50%) at 99.3% (612/618, 95% CI: 97.90%, 99.64%) specificity in the test cohort. Compared with the XGBoost model trained on the same dataset, Genie-ADLA demonstrated improved overall sensitivity in 11 of the 16 cancer types, with an average increase of 4.86%.For stage I-III cancer patients, the sensitivities at 99.3% specificity showed notable gains over XGBoost: colorectal cancer achieved 76.98% (97/126, 95% CI: 68.65%, 84.01%), an improvement of 9.52% from 67.46%; esophageal cancer reached 80.95% (51/63, 95% CI: 69.09%, 89.75%), up 6.35% from 74.60%; breast cancer reached 37.14% (26/70, 95% CI: 25.89%, 49.52%), improving by 5.71% from 31.43%. Lung cancer was subdivided into adenocarcinoma and non-adenocarcinoma, with stage I-III sensitivities of 40.90% (27/66, 95% CI: 28.95%, 53.71%) in adenocarcinoma, an increase of 10.6%, and 84.44% (38/45, 95% CI: 70.54%, 93.51%) in non-adenocarcinoma, improving by 2.22%. Conclusions: Genie-ADLA, leveraging advanced deep neural network architectures and data processing strategies, substantially elevates the performance ceiling of methylation-based early cancer detection, offering a new paradigm for AI-driven cancer screening. Citation Format: Kezhong Chen, Ziyu Li, Xiaojian Wu, Jian Huang, Guoyue Lv, Weiping Wen, Dahong Zhang, Xiangyu Zhao, Danbo Wang, Zhihua Liu, Lixin Sun, Shu Wang, Xiangnan Li, Zhigang Li, Jiandong Tai, Jiayin Yang, Zhentong Wei, Ming Cai, Qiang Zhang, Songbing He, Shuhua Yi, Shenhong Qu, Wenhui Zhao, Xianjun Yu, Ruixia Guo, Jianhong Lian, Desong Yang, Huaiwu Lu, Xi Guo, Yan Zhang, Zhuowei Liu, Yingjiang Ye, Chang Lin, Jie Gao, Xuanhui Liu, Yushu Guo, Suying Ding, Guoqiang Zhao, Yanzhan Yang, Jiangyu Li, Shiqing Chen, Hui Yu, Fang Liu, Yang Wang, Min Li, Baoliang Zhu, Yonghui Li, Xiaohui Wu, Fan Yang, Jun Wang. Genie-ADLA: A deep learning algorithm for methylation-based multiple cancer early detection (MCED) abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5471.
Chen et al. (Fri,) studied this question.