Advancements in AI Enhance Precision and Speed in Lung Tumor Monitoring

The proliferation of thoracic CT scans worldwide has significantly improved the early detection and monitoring of lung diseases such as bronchial carcinoma and metastases. In Germany, the number of lung CT examinations increased from 800,000 in 2009 to approximately 1.3 million in 2020, reflecting their importance in clinical diagnostics.

These scans are instrumental not only in identifying even minor treatment effects and side effects but also in optimizing patient care. However, comparing serial scans remains a complex, time-intensive process prone to errors, especially given the pressures faced by radiologists.

To address this challenge, innovative software utilizing artificial intelligence is being developed. The SPIRABENE project, a collaborative effort between Fraunhofer MEVIS, jung diagnostics GmbH, and the University Medical Center Mainz (UM), focuses on AI-driven solutions to improve diagnostic accuracy and streamline clinical workflows.

Jan Moltz, a lead researcher at Fraunhofer MEVIS, explains that their deep learning-based software can precisely locate and measure lung lesions rapidly, as well as detect new lesions when they appear. The software automates the registration process—comparing previous and current CT scans to establish anatomical correspondence—which is complicated by natural variations like breathing or weight loss.

Enhanced by deep learning, the technology offers faster and more accurate scan comparison than traditional methods. Trials show that AI can automatically detect 11% more tumors in follow-up scans and do so in under a second, using considerably less computational resources. This efficiency not only expedites diagnoses but also reduces energy consumption.

Designed for seamless integration into existing clinical systems, the software has undergone real-world testing and shows promising potential for widespread clinical application. Future developments aim for AI-supported monitoring to encompass the entire body, potentially revolutionizing follow-up care.

The impact of this technology extends beyond improved diagnostic precision. It enables healthcare professionals to identify ineffective treatments promptly, minimize unnecessary side effects and costs, and ultimately enhance patient outcomes. As Moltz states, working on tools that are actively used in clinics motivates ongoing innovation in medical imaging.

This advancement is a significant step towards smarter, faster lung cancer monitoring using artificial intelligence, promising benefits for both clinicians and patients.

Source: https://medicalxpress.com/news/2025-06-ai-lung-tumors-greater-precision.html