About the MEDUSA project

Chronic arthritis is a heterogeneous group of diseases characterized by long-lasting inflammation of joints. They can influence the patients’ general condition and involve other organs besides joints. The estimated prevalence rate of chronic arthritis is up to 1.5% of population. Rheumatoid arthritis causes substantial functional disability in approximately 50% patients after 10 years.

Accurate measurement of the disease activity is crucial to provide an adequate treatment and care to the patients. Semiquantitative ultrasound with power Doppler is a reliable and widely used method of assessing synovitis. Synovitis is estimated by ultrasound examiner using the scoring system graded from 0 to 3. Activity score is estimated on the basis of the examiner’s experience or standardized ultrasound atlases. The method needs trained medical personnel and the result can be affected by a human error. A system that can automatically assess the activity of synovitis would eliminate human dependent discrepancies and reduce the cost of evaluation.

Thus, the goal of the proposed research project is the creation of algorithms and methods for automated assessment of synovitis activity and development of a prototype software system, that will be clinically validated by comparing its results with the assessment carried out by medical personnel.

Novel image processing techniques will be used to detect multiple types of localized features, which will provide a reference for measurements performed on ultrasound images. These measurements will be integrated into a function approximating the human assessment.Machine learning methods will train the feature detectors and the assessment function on ultrasound images of synovitis cases, annotated and scored by medical experts. An additional research result will be novel visualization methods for power Doppler images, intended to aid the examiners.

Expected project outcomes:

• Prototype of a software system, that will be useful for medical personnel and will help in diagnosis of rheumatic diseases.
• System that will be able to assess activity of synovitis in the joints of hand.
• Project output will be clinically validated by comparing its results with the assessment carried out by medical personnel.

Research challenges:

• Image Processing - the subjectivness of the examination can be to large extent alleviated by adopting a segmentation scheme able to eliminate the influence of image background non-homogeneity, by incorporating local thresholding schemes combined with morphological features of high-intensity of the power Doppler image regions.
• Novel segmentation methods that utilizing the generalized distance transforms, combined with the elements of the level-sets theory and application of the Local
• Binary Patterns technique will be used to effectively perform the segmentation exploiting the textural image features.
• An additional research result will be novel visualization methods for power Doppler images, intended to aid the examiners.Machine learning methods will be used for training the feature detectors and the assessment function on ultrasound images of synovitis cases, annotated and scored by medical experts.
• An adaptable function (estimator), that will be trained to assess a degree of synovitis activity from an ultrasound image of a joint using machine learning methods. The estimator will be trained using ultrasound images of joints along with their synovitis assessment activity scores provided by expert examiners, with the aim for the estimator to approximate the average expert score.
• The system will be compatible with DICOM format, which is widely used for archiving of ultrasound examination results.