Artificial intelligence technology helps detect and diagnose lung cancer
Artificial intelligence (AI) is revolutionizing the medical field, especially in the diagnosis of cancer. One of the prominent applications of AI is the ability to detect lung cancer early and accurately.
AI is being widely applied in the medical field to improve diagnostic efficiency. A clear demonstration is the emergence of a new digital pathology platform, leveraging AI's image analysis capabilities to support accurate and rapid lung cancer diagnosis. This is an important step forward, opening up new prospects for effective treatment of this dangerous disease.
In a recent study, scientists at the Faculty of Medicine and University Hospital of Cologne (Germany), led by Dr. Yuri Tolkach and Professor Dr. Reinhard Büttner, have created an extremely effective tool to support lung cancer diagnosis.
.jpg)
The team has successfully developed a digital pathology platform that uses AI to automatically analyze lung tissue samples. This tool not only increases the accuracy of the diagnostic process but also saves doctors time and effort.
Lung cancer is one of the most dangerous silent killers, claiming the lives of millions of people worldwide every year. In particular, non-small cell lung cancer (NSCLC), accounting for more than 80% of all lung cancer cases, is the leading cause of death from epithelial cancer, posing a serious threat to public health.
Accurate diagnosis is essential for effective cancer treatment. Pathology, which analyzes tissue samples, plays a key role in this process. Thanks to the development of digital technology and artificial intelligence, pathologists can now analyze samples more quickly and accurately, helping doctors develop appropriate and personalized treatment regimens for each patient.
“The application of new AI tools will not only help us diagnose diseases more accurately, but also open a new era in cancer treatment,” said Dr. Tolkach. “By providing detailed information about each patient’s condition, we can make more intelligent and personalized treatment decisions, increasing the likelihood of successful treatment.”
Scientists trained the AI on a massive, high-quality dataset that was able to quickly and accurately analyze biopsy samples, clearly distinguishing 11 types of tumors and benign tissue at the pixel level.
In a study published in the prestigious journal Cell Reports Medicine, the team demonstrated that their AI tool can accurately classify non-small cell lung cancer, confirming this by comparing the model’s results with real-world data from multiple hospitals.
Next, the team identified four biomarkers that can be measured in tissue samples, which can more accurately predict disease progression and patient survival. Additionally, to advance global lung cancer research, the team publicly shared three large datasets, facilitating the development of new algorithms.
The team believes that the platform will become a useful tool for accurate disease diagnosis, predicting disease progression, and even supporting future treatment decisions. To confirm the feasibility of the platform, the team is collaborating with five leading pathology research institutes in Germany, Austria, and Japan to conduct large-scale validation studies.
In addition to automating healthcare tasks and analyzing medical images, AI is also gradually entering the lives of patients. AI tools like ChatGPT are increasingly used by people to search for health information and even self-diagnose.
Studies have shown that ChatGPT has the potential to provide useful health information, support mental health assessments, counseling, medication management, and patient education. However, the effectiveness of ChatGPT remains limited, especially in diagnosing health conditions in children, with some studies showing an accuracy of only about 17%.
Although large language models such as Llama-2-chat, Vicuna, Medllama2, Bard/Gemini, Claude, ChatGPT-3.5, and ChatGPT-4 have made significant progress, their application to genetic disease diagnosis remains limited. The complexity of genetic diseases requires more specialized diagnostic tools.