Artificial Intelligence Applications in Gastrointestinal Oncology

Artificial intelligence (AI) has emerged as a transformative tool for the detection, characterisation and management of tumours within the gastrointestinal tract. By harnessing advances in deep learning, convolutional neural networks and other machine-learning architectures, researchers have developed systems that can automatically identify mucosal lesions, predict invasion depth and delineate tumour margins in endoscopic images and videos. Real-time instance segmentation and object-detection frameworks enable computers to highlight suspicious regions during endoscopy, thereby improving sensitivity and specificity for early neoplastic change in the oesophagus, stomach and colon. Explainable AI models, which integrate domain knowledge into their decision pathways, are increasing clinical trust by offering interpretable rationale for each prediction. Beyond image analysis, AI has been applied to spectral data, multimodal fusion of white-light and narrow-band imaging, and to the computational staging of lymph node metastasis from radiological datasets. These innovations have the potential to standardise diagnostic performance across centres, reduce operator dependency and guide therapeutic decisions, such as suitability for endoscopic resection versus surgical intervention. Globally, AI-assisted screening programmes promise to alleviate disparities by equipping non-expert endoscopists with decision support and by increasing throughput in resource-limited settings. As algorithms mature, multimodal platforms that combine imaging, histopathological and clinical inputs may offer personalised risk stratification, treatment planning and follow-up surveillance strategies for patients with gastrointestinal malignancies.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Artificial Intelligence Applications in Gastrointestinal Oncology publication trend

The graph below shows the total number of articles in artificial intelligence applications in gastrointestinal oncology across all publications each year (not limited to Nature Index journals).

Technical terms

Convolutional neural network (CNN): A class of deep-learning model that learns hierarchical image features through stacked convolutional filters.

Computer-aided detection (CADe): Software designed to automatically identify and highlight regions of interest, such as potential neoplastic lesions, in medical images.

Instance segmentation: An image-processing task that simultaneously classifies and delineates each object instance within a scene, often via bounding boxes and pixel-level masks.

Explainable AI (XAI): AI frameworks that provide interpretable justifications for their outputs, enhancing transparency and user trust in clinical settings.

Narrow-band imaging (NBI): An endoscopic modality that uses specific light wavelengths to enhance mucosal and vascular patterns, aiding lesion characterisation.

References

1. Fusion of colour contrasted images for early detection of oesophageal squamous cell dysplasia from endoscopic videos in real time. Information Fusion (2023).
2. A deep learning system for detection of early Barrett's neoplasia: a model development and validation study. The Lancet Digital Health (2023).
3. Explainable artificial intelligence incorporated with domain knowledge diagnosing early gastric neoplasms under white light endoscopy. npj Digital Medicine (2023).