Can Cancer Be Detected Earlier and More Accurately Through Artificial Intelligence and Blood Tests?

Every year, cancer causes nearly ten million deaths worldwide and remains the leading cause of mortality in Japan. However, when detected at an early stage, the chances of long-term survival increase significantly. Currently, screening programs cover only a few types of cancer, such as breast, colon, and lung cancers. This leaves more than half of all cases without early detection solutions, leading to late diagnoses and often poor prognoses.

A new approach is emerging to change this situation: analyzing blood for tiny traces left by tumors. These tests, known as multi-cancer screenings, look for specific markers in the blood, such as fragments of tumor DNA or abnormal proteins. So far, their effectiveness remains limited, especially for cancers at a very early stage or for monitoring potential recurrence after treatment.

A team of Japanese researchers has launched an ambitious study to improve these tests. Their idea is to use multiple types of biological analyses simultaneously. They examine DNA, proteins, metabolism, and even the microbiome present in blood, saliva, or stool. All this data is then combined using artificial intelligence. This process creates more sensitive tools to detect early-stage cancer or relapse after treatment.

The study relies on two large Japanese databases. The first includes patients with various types of cancer, while the second involves healthy individuals. By comparing the results, scientists hope to identify weak but revealing signals of cancer. Artificial intelligence helps distinguish these signals and determine which organ is affected, enabling rapid referral for appropriate examinations.

The advantage of this method is its simplicity: a simple blood sample is sufficient. It could also detect cancers for which there is currently no systematic screening, such as pancreatic cancer or certain blood cancers. By identifying the disease earlier, treatments could be less aggressive and more effective.

To ensure the reliability of the results, all analyses follow strict and standardized protocols. Samples are processed in the same way, avoiding errors related to handling differences. Initial tests show that this multi-analysis approach could offer far greater accuracy than current methods.

In the long term, this advancement could transform cancer care. Broader and earlier screening would not only save lives but also reduce the costs associated with treating advanced-stage diseases. The challenge now is to validate these tools in real-world conditions and make them accessible to everyone.

Sources and Credits

Source Study

DOI: https://doi.org/10.1007/s10147-026-03001-6

Title: Clinical development of molecular residual disease (MRD) and multi-cancer early detection (MCED) using liquid biopsy multiomics with artificial intelligence (AI)

Journal: International Journal of Clinical Oncology

Publisher: Springer Science and Business Media LLC

Authors: Taro Shibuki; Riu Yamashita; Tadayoshi Hashimoto; Takao Fujisawa; Mitsuho Imai; Junichiro Yuda; Takeshi Kuwata; Toshihiro Misumi; Yoshiaki Nakamura; Hideaki Bando; Kaname Kojima; Sayuri Tokioka; Ippei Chiba; Naoki Nakaya; Atsushi Hozawa; Seizo Koshiba; Nobuo Fuse; Sakae Saito; Ritsuko Shimizu; Woong-Yang Park; Kengo Kinoshita; Takayuki Yoshino