Gateway for:

Member Countries

Inter-Modality Lung Cancer Diagnosis

#B-1926


Integrated CT, MRI and X-ray Image Analysis for Advancing Early Lung Cancer Diagnosis

Tech Area / Field

  • BIO-RAD/Radiobiology/Biotechnology
  • INF-SIG/Sensors and Signal Processing/Information and Communications
  • MED-DID/Diagnostics & Devices/Medicine

Status
3 Approved without Funding

Registration date
14.03.2011

Leading Institute
National Academy of Sciences of the Republic of Belarus / Institute of Informatics Problems, Belarus, Minsk

Supporting institutes

  • N.N. Alexandrov National Cancer Center of Belarus, Belarus, Minsk

Collaborators

  • University Medical Center / Image Sciences Institute, The Netherlands, Utrecht\nRadboud University Nijmegen Medical Centre, The Netherlands, Nijmegen

Project summary

Project goal.

The goal of the project is to develop a methodological support and computer-aided detection procedure for the computer-aided early differential diagnosis and prognosis of solitary pulmonary nodules in lungs based on integrated analysis of radiological (chest X-rays, Computed Tomography (CT) and Magnetic Resonance (MR)) images.

Introduction and Overview.

The problem. According to the World Health Organization (WHO) the Lung cancer (LC) is the leading cause of cancer-related death with the current fatality rate exceeding that of the next three most common cancers combined. One of the key issues is the fact that lung cancer is rarely diagnosed in the early potentially treatable stage which is relatively asymptomatic and that is why it is disproportionately deadly because it lacks any clear early detection, i.e. most cases are not diagnosed until they have reached advanced stages. The expected 5-year survival rate for all patients in whom lung cancer is diagnosed is 15%, compared with 61% for colon cancer, 86% for breast cancer, and 96% for prostate cancer. The principal problems are as follows: (a) incidence of cancer increases worldwide; (b) lung cancer is the first in incidence and mortality among cancers; (c) survival rate for patients with lung cancer is minor. Therefore lung cancer still needs novel solutions in estimation, investigation, automation and efficient early diagnosis.

What is pulmonary nodules. The most common manifestation of an early lung cancer is a solitary pulmonary nodule. A solitary pulmonary nodule (SPN) is defined as a single discrete pulmonary opacity up to 30 mm in diameter. Nodules are extremely common in clinical practice that may represent primary lung cancer or other malignant or benign lesions, but challenging to manage, especially small, subcentimetre nodules. An estimated 150,000 such nodules are identified each year. The prevalence of malignancy varied by size (0 to 1% for nodules < 5 mm, 6 to 28% for nodules 5 to 10 mm, and 64 to 82% for nodules > 20 mm). Almost 100% of SPNs larger than 30 mm are malignant.

Past works on quantitative analysis. For recent decades we have seen a remarkable improvement in the diagnostic managements of oncology patients. Several imaging modalities were acknowledged as qualified for diagnostics of lung cancer, from which X-ray chest imaging is historically the earliest and still is the most common type of radiological procedure. Further, the Computed Tomography (CT) is considered to be the best imaging modality for the detection of small pulmonary nodules, particularly since the introduction of the helical technology. Also it was shown that the dynamic Magnetic Resonance (MR) imaging has a potential role in the accurate staging of non-small cell lung cancer and in conjunction with dynamic helical CT – for differentiating between malignant and benign SPNs. However, every imaging technique alone has limitations for localizations and differential diagnosis of SPNs.

The demand. Project results are required for medical institutions, biomedical research laboratories. Both practical and scientific institutions as well as educational centers may use the production (software modules) of the project. Necessary requirement for utilization of the software modules to be implemented is the capability to perform inter-modality observation. The primary consumer is the N.N. Alexandrov National Cancer Center of Belarus and other collaborators.

Influence to the progress. The project will influence to the progress by providing the community for a novel computerized inter-modality image analysis and diagnosis tools and covering the existing problem in the early diagnosis of lung cancer.

Project participants and their roles.

The project team consists of a harmonized combination of specialists carefully selected from both information technology and oncology camps. They came from the main Belarusian state research centers, namely the United Institute of Informatics Problems of Belarus National Academy of Sciences (UIIP) and N.N. Alexandrov National Cancer Center of Belarus (NCCB) respectively. The UIIP plays a leading role in the information processing, image analysis, and pattern recognition research and developments conducted in Belarus. The NCCB is the leading clinical and research oncology institution in Belarus. It is the largest comprehensive oncology center in the country also serving as a teaching hospital.

The NCCB will contribute to creation of electronic database of X-rays, CT and MR images of patients with peripheral lung lesions. The main activity will be is setting up the research hypotheses, patients selection, collection of medical radiological images and development (in collaboration with United Institute of Informatics Problems) of new algorithms and techniques for lung tumours classification. Other activities such as testing research hypotheses and creating prognostic models are performed jointly.

Expected Results and their Application.

Specific results. The execution of the five project tasks is resulted in the following specific outcomes.


Task 1: Retrospective radiological databases containing chest X-ray, high resolution CT and MR images of patients in Belarus with morphologically verified diagnosis of lung cancer and benign tumors.
Task 2: Putting forward the efficient image features to characterize nodules for each image modality including integrated inter-modality features for CT/MRI images.
Task 3: Pattern recognition and image analysis procedures for localization of nodule candidates for every image modality the project dealing with.
Task 4: A collection of multivariate statistical analysis procedures and results of the study of pulmonary nodules differentiation.
Task 5. Software for early computer-aided diagnosis (CAD) of lung cancer which would contain modules implemented in Tasks 2-4.

General results. The project will be resulted in new methods, algorithms and software improve the performance of early computer-aided cancer diagnosis. The authors intend to develop multivariate statistical and pattern recognition methods to assess the effectiveness and perspective of integrated image-based inter-modality CAD diagnosis system.

Project consequences. In scientific area the project will contribute towards the advanced radiological image analysis methods and new knowledge on lungs cancer early detection and follow-up. In practical domain it will provide certain improvements on diagnosis, prognosis and treatment of lung cancer patients. In the field of commerce the project will introduce new products on the market of medical image analysis software for computer-aided diagnosis.

Application areas. They include general radiological image analysis problem and early lungs cancer diagnosis in clinics and medical education.

Dissemination. Since the participating institutions represent leading Belarusian centers in their professional areas, this secures a high level of conducted research and legal, duly organized dissemination of the results in Belarus and worldwide.

Meeting ISTC Goals and Objectives

Execution of the project will allow:

  • to offer weapon-oriented scientists and specialists participating in the project, who participated earlier in the development of biological weapon and rocket technologies, an opportunity to redirect their skills to peaceful activities in the field of a diagnostics of lung cancer;
  • to be integrated with international scientific community and to cooperate with foreign scientists who will support applied research and new integrated IT development for making a diagnosis of diseases;
  • to contribute to solve a problem dealing with the automation of diagnosis of peripheral lung cancer at local and international level.

Scope of Activities.

Project scope. The main project activities are represented by five specific tasks including creating integrated inter-modality radiological image database and clinical database containing diagnosis reports on selected patients, developing specialized software and performing analysis of radiological images and joint statistical analysis of image and clinical data. These tasks roughly reflect natural sequence of biomedical data acquisition, processing, statistical analysis and reporting.

Project efforts. The total number of project participants is 26 with the total project effort of 7770 person*days. It includes 15 engineering participants with the total effort of 5390 person*days and 11 medical with the efforts of 2380 person*days.

Project specificity. The characteristic point of this project is the quite large image acquisition of different modalities and sophisticated inter-modality image analysis. In particular, it is supposed that there will be about 300 images of CT and 100 MR modalities obtained on the up-to-date scanners and about 500 digital X-ray chest images. The research hypothesis is that inter-modality analysis may decrease shortcomings of each particular modality for lung cancer early diagnosis thus increasing the accuracy of cancer detection and potentially 5-year survival rate.

Role of Foreign Collaborators.

Technical competence of collaborators has confirmed an actuality of the subject while preparing the application material to support the project. For the project execution the scope of cooperation with the foreign collaborators will include:

  • to provide high level expertise on developing advanced image analysis methods and interpreting radiological data;
  • to validate the intermediate project results and to provide advices on further steps;
  • to participate in preparation of joint publications resulted from the project and dissemination of the results;
  • to provide technical help and occasional, non-systematic use of lab equipment in cases of high mutual interests.

Technical Approach and Methodology.

Employees from United Institute of Informatics Problems of National Academy of Sciences of Belarus (UIIP) have wide experience in development of "weapon" technologies and software for image processing, object analysis and recognition. Besides there is an experience in development of algorithms in the field of medical images processing which will be advanced here and will be used to solve problems of the project. During many years, participants from N.N. Alexandrov National Cancer Center of Ministry of Health of Belarus (NCCB) are engaged to work with oncological diseases. They use all required equipment professionally and are excellent experts in the field.

Standard methods will be used to obtain necessary chest X-ray, CT and MR materials. Algorithms and methods for image processing and visualization, information coding and transferring will be used for development of automated system for early diagnosis and prognosis of lungs cancer based on inter-modality analysis of chest X-ray, CT and MR images.

The quantitative relationships will be established with the help of (a) statistical univariate methods; (b) statistical multivariate regression models and (c) the most recent algorithms that make use the ability of Support Vector Machines (SVM) and some other clustering methods to quantify regression-like relationships (including multivariate ones) between the groups of measurements.

The above methodology has been introduced very recently and it originates from fruitful ideas of generalized co-occurrence matrices suggested by project participants (see Section 12.2) and applied to 2D and 3D shape analysis, 2D and 3D gray-scale texture characterization.


Back

The International Science and Technology Center (ISTC) is an intergovernmental organization connecting scientists from Kazakhstan, Armenia, Tajikistan, Kyrgyzstan, and Georgia with their peers and research organizations in the EU, Japan, Republic of Korea, Norway and the United States.

 

ISTC facilitates international science projects and assists the global scientific and business community to source and engage with CIS and Georgian institutes that develop or possess an excellence of scientific know-how.

Promotional Material

Значимы проект

See ISTC's new Promotional video view