Poster Presentation 2014 International Biophysics Congress

Semi-automated analysis of myocardial fibrosis by fractal characterization of histological images (#421)

Michael Mayrhofer-Reinhartshuber 1 , Philipp Kainz 1 , Damián Sánchez-Quintana 2 , Yolanda Macías 2 , Martin Asslaber 3 , Ernst Hofer 1 , Helmut Ahammer 1
  1. Institute of Biophysics, Medical University of Graz, Graz, Austria
  2. Department of Anatomy and Cell Biology, Faculty of Medicine, University of Extremadura, Badajoz, Spain
  3. Institute of Pathology, Medical University of Graz, Graz, Austria

Structural and electrical remodeling occurs regularly in heart disease. An excessive deposition of collagen seems to have high influence in the enhancement of myocardial fibrosis. Cardiac collagen plays an important role in cardiac disease, because it may affect contractility, alter the activation sequence and lead to life-threatening arrhythmias. Besides the amount of collagen, also the type (pattern) of the occurring fibrosis is specific for the underlying disease 1,2. Thus, detection of collagen in patients with heart disease is of utmost importance.

Although non-invasive techniques, e.g. cardiovascular magnetic resonance imaging, are beneficial for patients, histological analysis of cardiac tissue is still the gold standard in the detection, quantification and characterization of fibrosis 3,4. Nevertheless, it heavily depends on the experience of the observer. To overcome this drawback, fully automated and observer-independent image analysis is a promising approach. Furthermore, non-invasive techniques can be improved by a comparison with high quality data sets obtained from automated histological image analysis

We used image analysis methods to investigate images of 20 human hearts (interventricular septum, left papillary
muscles - anterior/posterior), which suffered from ischemic heart disease. The amount of myocardial fibrosis was calculated for 40 different images (20 control, 20 fibrosis) stained with Masson’s trichrome. Statistically significant (p<0.001) differences in the amount of detected connective tissue were obtained (control: 4-6%, fibrosis: 41-47%). Images of sections stained with Picrosirius Red were used to discriminate between different grades of myocardial fibrosis.

Subsequently, fractal analysis was used to reveal a fractal dimension of (1.65 ± 0.10) for the investigated myocardial fibrosis in the segmented images. The fractal dimension was found to be a statistically significant (p<0.001) parameter to discriminate between control- and fibrosis-group, too. Actually, this value is comparable to results obtained from fractal characterization of liver fibrosis5  and may be used to differentiate between the types of myocardial fibrosis.

  1. de Jong, S.; van Veen, T. A.; van Rijen, H. V.; de Bakker, J. M.: Fibrosis and cardiac arrhythmias. J Cardiovasc Pharm, 57: 630-638, 2011.
  2. White, S. K.; Sado, D. M.; Flett, A. S.; Moon, J. C.: Characterising the myocardial interstitial space. Heart, 98: 773-779, 2012.
  3. Popescu, B. A., Roşca, M.: Imaging of myocardial fibrosis in hypertrophic cardiomyopathy: what is the gold standard?. Heart, heartjnl-2013-305359, 2014.
  4. de Jong, S.; Veen, T.; Bakker, J.; Rijen, H.: Monitoring cardiac fibrosis: a technical challenge. Netherlands Heart Journal, 20: 44-48, 2012.
  5. Grizzi, F.; Russo, C.; Franceschini, B. et al.: Sampling variability of computer-aided fractal-corrected measures of liver fibrosis in needle biopsy speci-mens. World J Gastroenterol, 12: 7660, 2006.