Radiomics Analysis

In the field of medicine, radiomics is a method that extracts large amount of features from radiographic medical images using data-characterization algorithms. These features, termed radiomic features, have the potential to uncover disease characteristics that fail to be appreciated by the naked eye. The hypothesis of radiomics is that the distinctive imaging features between disease forms may be useful for predicting prognosis and therapeutic response for various conditions, thus providing valuable information for personalized therapy. Radiomics emerged from the medical field of oncology and is the most advanced in applications within that field. However, the technique can be applied to any medical study where a disease or a condition can be imaged.

Based on the theory and methods of radiomics, we developed imaging-aided diagnosis software and constructed a big data resource platform. In the image-aided diagnosis software, we independently developed the integrated image-aided diagnosis software for lung cancer. Based on the large data of nearly 10,000 cases of lung cancer, it integrates the radiomics prediction model based on deep machine learning, and integrates 1,000 items of Quantitative three-dimensional image features and doctor's experience features, which can realize automatic tumor segmentation, precise registration and three-dimensional visualization, and can predict and evaluate tumor benign and malignant, TNM staging, survival and other clinical key indicators, and provide clinicians with second diagnostic opinions.

Model Components

Medical Imaging ToolKit
Medical Imaging ToolKit (MITK) is a C++ class library with easy-to-understand interface specifications, efficient algorithm framework, and rigorous data structure. Its main purpose is to provide a comprehensive medical image processing such as reconstruction, segmentation, registration, and visualization.

Figure 1 Medical Imaging ToolKit (MITK)

3D Medical Image Data Integrated Processing Platform
The 3D Medical Image Data Integrated Processing Platform (3DMed) is a system integration platform for medical image data. It seamlessly provides an interactive interface for all the algorithms in the MITK toolkit. The functions cover data acquisition, data format conversion, surface layout, entity resolution, image segmentation, image registration, three-dimensional image processing, three-dimensional virtual cutting and three-dimensional measurement, etc., and defines a core module interface, allowing users to customize plug-ins.

Figure 2 3D Medical Image Data Integrated Processing Platform (3DMed)

Radiomics Individual Predication Models
The most important thing in imaging analysis is to propose suitable, identifiable, reliable, and repeatable features that have potential application value in clinical diagnosis. Radiomics Individual Predication Models (RIPM) achieves accurate diagnosis of hepatitis B fibrosis equivalent to pathology, accurate evaluation of neoadjuvant treatment effects for locally advanced rectal cancer, accurate preoperative prediction of colorectal cancer lymph node metastasis, and accurate progression-free survival of advanced nasopharyngeal carcinoma Prediction and precise prediction of TKI treatment resistance time for advanced EGFR mutant non-small cell lung cancer.

Figure 3 Radiomics Individual Predication Models (RIPM)

Advantages

Innovative
Through high-throughput extraction of a large number of imaging features describing tumor characteristics, fusion of genetic information and imaging multi-modal information, radiomics based on artificial intelligence provides a new way to achieve accurate diagnosis.

Reliable
Our radiomics model is based on nearly 10,000 patient data with complete information, and combines 1,000 quantitative three-dimensional image features to form a sufficiently reliable and credible prediction model.

Convenient
Our Radiomics model provides a friendly user interaction interface and defines a set of core module interfaces so that users can customize plug-ins.

CD ComputaBio has formed a team of experts excellent in imaging science and clinical domain knowledge, providing AI-driven solutions for radiomics analysis according to your detailed requirements.