AI-Powered Blood Test Shows 90%+ Sensitivity for Early Cancers

GLASGOW, Scotland – April 02, 2026 – A Scottish technology firm has presented breakthrough data suggesting its AI-powered blood test can detect some of the deadliest cancers at their earliest stages with over 90% sensitivity. The findings, set to be detailed at the American Association for Cancer Research (AACR) Annual Meeting in San Diego, represent a significant step forward in the quest to diagnose cancers when they are most treatable.

Dxcover, a leader in spectroscopic liquid biopsy, announced it will showcase studies demonstrating its platform's ability to identify Stage I and II colorectal, pancreatic, and ovarian cancers—diseases that are notoriously difficult to screen for and often discovered too late. The technology analyzes subtle molecular changes in a patient's blood, offering a new window into the body's earliest response to malignancy.

“Cancer is detected too late because early signals are invisible to current clinical tools,” commented Matthew Baker, PhD, CEO & co-Founder of Dxcover, in a statement released by the company. “The Dxcover approach analyzes the body’s earliest reactions to malignancy – a perfect complement to existing high specificity diagnostics.”

The Science Behind the Signal

Unlike many liquid biopsy tests that hunt for fragments of circulating tumor DNA (ctDNA), Dxcover’s platform employs a different and potentially more comprehensive method. The technology, known as Multi-Omic Spectral Analysis (MOSA-Dx™), uses a combination of Attenuated Total Reflectance Fourier Transform Infrared (ATR-FTIR) spectroscopy and proprietary machine learning algorithms.

In simple terms, the platform shines infrared light through a small, prepared blood serum sample. The way molecules in the sample—including proteins, lipids, and metabolites—absorb or reflect that light creates a unique “molecular fingerprint.” Cancer can induce subtle but widespread changes in this fingerprint long before a tumor is large enough to be detected by imaging or shed significant amounts of DNA. The company's AI is trained to recognize the specific spectral patterns indicative of cancer, distinguishing them from healthy samples.

This “pan-omic” approach, which captures information from the proteome, lipidome, and metabolome simultaneously, is designed to detect the body's systemic response to the disease. The company has also developed its own hardware, the Dxcover® Infrared Platform, which uses an automated sampling system to dramatically increase the speed of analysis, delivering results in hours instead of weeks.

A New Front Against Hard-to-Detect Cancers

The implications of a highly sensitive, non-invasive test are most profound for cancers where effective early screening is either non-existent or has significant drawbacks. The three cancers highlighted in Dxcover's AACR presentations fall squarely into this category.

Pancreatic Cancer: With no recommended screening for the general population, pancreatic cancer is typically diagnosed at an advanced stage, contributing to its grim five-year survival rate. A blood test with high sensitivity for Stage I and II disease would be transformative, potentially creating the first-ever opportunity for widespread early detection in high-risk groups and, eventually, the broader population.

Ovarian Cancer: Often called a “silent killer,” ovarian cancer also lacks a reliable screening method. The commonly used CA125 blood test is notoriously unreliable for early-stage detection, as it is elevated in only about 50% of Stage I cases and can be raised by numerous benign conditions. The prospect of a test that can consistently detect early-stage ovarian cancer with over 90% sensitivity could fundamentally change clinical outcomes for women.

Colorectal Cancer (CRC): While effective screening methods like colonoscopy exist for CRC, compliance can be low due to the invasive nature and preparation required. A simple, accurate blood test could serve as a powerful frontline screening tool, encouraging more people to get tested and identifying those who require a follow-up colonoscopy. This could improve overall screening uptake and catch cancers earlier than current non-invasive stool-based tests.

The Path to the Clinic: Promise and Hurdles

Dxcover has already made significant strides toward commercialization. The company has secured patents for its technology in the US, EU, and China, and its platform has achieved CE-IVDR certification, clearing a major regulatory hurdle for use as an in-vitro diagnostic device in Europe. With headquarters in Scotland and a growing presence in the United States, the company is positioning itself for a global launch.

However, the road from promising research to a widely adopted clinical standard is fraught with challenges. A key consideration for any screening test is the balance between sensitivity (the ability to correctly identify those with the disease) and specificity (the ability to correctly identify those without it).

While the press release touts high sensitivity, previous research has shown this can come at the cost of specificity. For instance, one large-scale study indicated the test could be tuned to detect 99% of Stage I cancers, but with a specificity of only 59%. This would result in a substantial number of false positives—healthy individuals who are incorrectly told they may have cancer. Such results can cause immense patient anxiety and lead to a cascade of expensive and potentially invasive follow-up tests, placing a heavy burden on healthcare systems.

Furthermore, the real-world performance of the test must be validated in large, diverse, prospective clinical trials that compare it head-to-head against other emerging technologies and the current standard of care. Questions of cost-effectiveness, reimbursement from insurers, and the logistical scalability to process millions of samples annually remain critical hurdles to overcome before this technology could be implemented for population-wide screening.

Despite these challenges, the data presented by Dxcover marks a notable advancement in the multi-cancer early detection (MCED) space. By analyzing the body's holistic response to disease rather than just tumor-shed DNA, the spectroscopic approach offers a powerful new dimension to liquid biopsy. As researchers continue to refine and validate these powerful new tools, the prospect of fundamentally altering the timeline of a cancer diagnosis moves ever closer to reality.