Artificial intelligence applied to the diagnosis of rare diseases associated with collagen VI
Researchers at the Institut de Recerca Sant Joan de Déu (IRSJD), in collaboration with researchers from the Institut de Robòtica i Informàtica Industrial, have developed an artificial intelligence system to help diagnose rare diseases related to deficienc
The system makes the diagnosis from images obtained with a confocal microscope located in the Hospital Sant Joan de Déu Barcelona and is based on artificial learning techniques, using cases previously diagnosed by IRSJD's specialists to generate a fully automatic diagnostic system with a reliability greater than 95%. This valuable tool will allow an objective evaluation of the effectiveness of any new therapy that can be developed to treat these diseases.
Deficiencies in the collagen VI' structure are a common cause of neuromuscular diseases with manifestations ranging from Bethlem myopathy to severe congenital muscular dystrophy Ullrich. Symptoms of these diseases include proximal and axial muscle weakness, distal hyperlaxity, joint contractures and critical respiratory failure that requires assisted ventilation, dramatically reducing life expectancy.
"Although we know that structural defects of collagen VI are related to mutations of the COL6A1, COL6A2 and COL6A3 genes. Diagnosis remains difficult, despite current genetic sequencing technologies." Comment Cecília Jiménez, PhD, research coordinator of the Neuromuscular Diseases Unit.
This difficulty generally occurs in diseases caused by dominant mutations, where there is no complete absence of a major protein, and when the effect of a genetic variant on the protein structure may not be evident. Currently, the diagnosis of dystrophies related to collagen VI is made from the analysis of the images of fibroblast cultures by specialists.
For this analysis, professionals take into account different aspects of the images: the coherence in the orientation of the collagen fibers, the distribution of the collagen network and the disposition of the cells in said network to identify potential patients. However, this evaluation is only qualitative, and regulatory agencies will not approve any treatment (such as genetic editing using CRISPR / Cas9 technology) without an objective methodology to assess its effectiveness.
For this reason, the system proposed in the work published in Applied Soft Computing will be a precise methodology to quantitatively monitor the effects of any new therapy. This system solves, on the one hand, the problem of the lack of data for typical learning in rare diseases; on the other, it indicates the possibly problematic areas in the query images. And it also provides a general quantitative evaluation of the condition of the patients.
Currently, the Neuromuscular Diseases Unit (Institut de Recerca Sant Joan de Déu · Hospital Sant Joan de Déu) is a national and international reference in research and diagnosis of neuromuscular diseases due to collagen deficiency VI.
In the study, researchers from the Neuromuscular Diseases Unit, the Paediatric Neurology Department and the Confocal Microscopy Unit (Daniel Bravo Center for Diagnosis and Research in Minority Diseases and Pediatric Institute of Rare Diseases) collaborated.
The Institute of Robotics and Industrial Informatics is a Joint Research Center of the Spanish Council for Scientific Research (CSIC) and the Technical University of Catalonia (UPC).
Image description: diagnosis display of a fibroblast culture image. The different areas of the image are evaluated independently which allows to quickly identify areas with defects in collagen VI. The system also provides a general diagnosis to track patients.