Colon cancer remains one of the most common and deadly cancers worldwide, with diagnosis often relying on time-consuming procedures such as colonoscopy and histopathological analysis. Delays or inaccuracies in detecting polyps and distinguishing benign from malignant tissue can significantly affect patient outcomes. At the same time, growing clinical workloads have increased the need for faster and more reliable diagnostic support. A new study now shows that artificial intelligence (AI) has meaningfully improved the speed, accuracy, and reliability of colon cancer diagnosis and prognosis.

In the study led by investigators from the University of Sharjah (Sharjah, UAE), in collaboration with academic institutions across Europe, the Middle East, and North Africa, the team conducted a comprehensive meta-analysis examining how AI, including machine learning and deep learning, has been applied to colon cancer diagnosis over the past five years. The study focused on how AI tools are being integrated into clinical workflows rather than on a single diagnostic product.

The analysis covered 80 peer-reviewed studies published between 2020 and 2024. These studies examined four main AI applications: classification of cancerous tissue, detection of polyps during colonoscopy, segmentation of glands and tumors in pathology slides, and prediction of disease outcomes. Particular attention was given to explainable AI methods, which help clinicians understand how algorithms reach their conclusions and build trust in real-world medical settings.

Across the reviewed studies, AI-based systems consistently improved diagnostic accuracy compared with traditional approaches. Deep learning models showed strong performance in detecting polyps during colonoscopy and differentiating benign from malignant tissue in histopathology. The findings, published in the International Journal of Medical Informatics, also showed that AI enhanced cancer grading and gland segmentation, supporting more precise staging and treatment planning.

The results suggest that AI is already transforming colon cancer care by enabling earlier detection, reducing invasiveness, and streamlining clinical workflows. However, the study also highlights key challenges, including limited data diversity, lack of external validation, high computational demands, and incomplete integration into hospital information systems. The authors emphasize that addressing these gaps will be essential before AI tools can be widely adopted in routine clinical practice.

“Explainable AI is not just a feature. It is essential for building clinician confidence and closing the gap between technology and medical practice. The promise of AI in medicine lies not just in speed or accuracy, but in creating transparent systems that doctors can rely on,” said Professor Saad Harous, co-author of the study.

http://m.allmycbs.com/en/index.asp