Early diagnosis of lung cancer is possible with new artificial intelligence-supported method

The TÜBİTAK 1001-supported project, led by Assoc. Prof. Dr. Handan Tanyıldızı Kökkülünk from Altınbaş University, has achieved significant scientific success in the early diagnosis of lung cancer . The developed artificial intelligence-based method combines PET/CT images, sarcopenia markers and blood biomarkers, providing early diagnosis with high accuracy rates.

In the project titled “Predicting Lung Cancer Diagnosis with Machine Learning via Sarcopenia, New Generation Inflammatory Markers and PET/CT Anatomical-Metabolic Biomarkers”, not only imaging technologies but also data on the patient’s general physical condition and immune system were included in lung cancer diagnosis. PET/CT imaging data, physical performance measures related to sarcopenia and inflammatory indicators such as CRP, WBC, and NEU were analyzed with the Random Forest algorithm. The artificial intelligence model developed with this multimodal data approach successfully classified benign, malignant and non-cancerous cases with 97 percent accuracy and 99 percent AUC (area under the curve) during the testing phase.

Holistic Approach Could Start a New Era in Healthcare

Project leader Assoc. Prof. Dr. Handan Tanyıldızı Kökkülünk stated that the developed method does not only rely on imaging data, but also provides a comprehensive analysis by taking into account physical performance measures such as waist circumference muscle area, walking speed, and immune system indicators. Thanks to this holistic approach, lung cancer is easier to diagnose at an early stage and patients can access faster treatment.

Quality of Life Increases, Treatment Costs Decrease

The new diagnostic method not only provides early diagnosis, but also provides economic contribution to the healthcare system. The high accuracy rates obtained by combining different data sources prevent unnecessary tests and reduce costs for both the patient and the healthcare system. On the other hand, thanks to early diagnosis, patients' quality of life increases and treatment processes can be managed more effectively.

This innovative study, led by Assoc. Prof. Dr. Kökkülünk, concretely demonstrates how integrating artificial intelligence and multiple biomarker use into health technologies can yield groundbreaking results, especially in fatal diseases such as cancer.

Source: UAV