Radiomics is a field of study that focuses on imaging biomarkers, utilizing algorithms to extract a large number of features from medical images to describe tumor characteristics. However, radiomics analysis requires manipulating three-dimensional volumetric data to perform slice-by-slice tumor segmentation, a labor-intensive process prone to observer variability. 

Streamlined from Data Input to Tumor Analysis 

To address these challenges, a team led by Dr. Yu-Chun Lin, Deputy Director of the Department of Diagnostic Imaging at Linkou Chang Gung Memorial Hospital, has developed a fully automated MRI radiomics analysis platform. This one-stop solution automates the entire process from data input to tumor radiomics analysis. The platform's capabilities include image type recognition, fully automated tumor segmentation, and radiomics feature extraction. By simply inputting a batch of patient MRI images, the platform can automatically segment potential tumor regions in each image and extract 105 radiomics features of the tumor, including shape, gray-level co-occurrence matrix, gray-level run length matrix, gray-level size zone matrix, and neighboring gray-level difference matrix. 

Significance of the Platform's Development 

• Enhanced Diagnostic Accuracy: The AI model can quickly segment tumor regions and combine radiomics data to assist physicians in making more accurate diagnoses, thereby providing more suitable treatment plans. 
• Increased Medical Efficiency: The AI model rapidly analyzes MRI scan data, identifying potential tumor locations and their radiomics features, significantly reducing the time physicians spend on image analysis and allowing them to focus more on patient care and treatment. 
• Early Detection for Better Treatment Outcomes: The AI model can swiftly detect potential tumors from a large volume of medical images, enhancing early tumor detection and improving patient treatment outcomes. 
• Personalized Treatment Planning: Radiomics features provide deeper biological insights into tumors, facilitating personalized treatment plans that offer the most suitable therapy for each patient. 
• Market Potential: With the growing application of AI in healthcare, the global cancer diagnosis and treatment market is expanding. This technology can be applied in various medical scenarios requiring precise diagnosis, presenting significant commercial value and market potential. 

Dr. Yu-Chun Lin stated, "We hope this technology platform can help physicians diagnose and treat gynecologic cancer patients more effectively, leading to better prognoses for patients." 

Resource (mandarin): 

新創焦點 - AI賦能影像組學！ 

一站式婦癌MRI圖像分割及特徵提取平台