Innovative Frontiers in Multi-Cancer Early Detection and Screening Technologies
DOI:
https://doi.org/10.64764/ijietas.v2.i2.04Keywords:
Diabetic retinopathy; Deep learning; Vision transformers; Cross-attention; Medical image classificationAbstract
The early detection of cancer can mean the difference between a long and a short life and effective treatment. But, let's be honest, most cancer screening tests check for a single type of cancer, and each screening has its own drawbacks. You have low-dose CT for lung cancer, mammograms for breast cancer and PSA tests for prostate cancer all of which can cause undue anxiety, prompt unnecessary follow-up testing or identify slow-growing cancers that would never pose a threat to life. But then for cancers such as pancreatic or ovarian? It is often impossible to screen for any standard. Here, the paper looks into how a multi-cancer early detection (MCED) strategy could help fill these gaps. Rather than relying on a single method, researchers are combining their concepts: blood tests examining cfDNA methylation and fragmentation patterns; sniffers that detect VOCs through an electronic nose; deep-learning systems that interpret ultrasound or EB-OCT images; and explainable AI systems that categorize patients by symptom. The results are quite remarkable: one study reported achieving 99% accuracy based on symptom information, while another obtained 98.41% accuracy for breast ultrasounds using ResNet50, and a bio-AI volatolomics platform achieved 94% accuracy with seven cancer types. The DELFI-Pro fragmentomics test achieved a 72% sensitivity in stage I cancers with an over 99% specificity. As mentioned, these are data from various research groups, so it is crucial to remember each group's specific settings. Not in the least, the paper doesn't overlook the human element — not all have been given the same opportunities to research. It examines the issue of fair inclusion and considers younger people who develop cancer before age 50 (EAOC), as their cases may be more difficult to identify using the REPRESENT framework. Ultimately, the paper suggests, synthesis of all the clues, as they might be found in blood, breath, scan, or symptoms, could advance improved, earlier cancer detection. But all signs point to the fact that, before it becomes the norm, the tests must be more thoroughly validated. The terms ―Multi-Cancer Early Detection (MCED),‖ ―Liquid Biopsy,‖ and ―Cell-Free DNA (cfDNA) Methylation‖ are frequently used.Three terms are often used: Multi-Cancer Early Detection (MCED), Liquid Biopsy, and Cell-Free DNA (cfDNA) Methylation. Volatolomics, Explainable AI (XAI), Early-Age Onset Cancer (EAOC), Multi-Modal Fusion, Population Health Equity, Electronic Nose (eNose).
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Author(s)

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.






