Monitoring of the Achilles tendon healing process: can artificial intelligence be helpful?

PURPOSE: The aim of this study was to verify improved, ensemble-based strategy for inferencing with use of our solution for quantitative assessment of tendons and ligaments healing process and to show possible applications of the method. METHODS: We chose the problem of the Achilles tendon rupture as an example representing a group of common sport traumas. We derived our dataset from 90 individuals and divided it into two subsets: healthy individuals and patients with complete Achilles tendon ruptures. We computed approx. 160 000 2D axial cross-sections from 3D MRI studies and preprocessed them to create a suitable input for artificial intelligence methods. Finally, we compared different training methods for chosen approaches for quantitative assessment of tendon tissue healing with the use of statistical analysis. RESULTS: We showed improvement in inferencing with use of the ensemble technique that results from achieving comparable accuracy of 99% for our previously published method trained on 500 000 samples and for the new ensemble technique trained on 160 000 samples. We also showed real-life applications of our approach that address several clinical problems: (1) automatic classification of healthy and injured tendons, (2) assessment of the healing process, (3) a pathologic tissue localization. CONCLUSIONS: The presented method enables acquiring comparable accuracy with less training samples. The applications of the method presented in the paper as case studies can facilitate evaluation of the healing process and comparing with previous examination of the same patient as well as with other patients. This approach might be probably transferred to other musculoskeletal tissues and joints.

Determining Geometrically Stable Domains in Molecular Conformation Sets.

Detecting significant conformational changes occurring in biomolecules is a challenging task, especially when considering tens to hundreds of thousands of conformations. Conformational variability can be described by dividing a biomolecule into dynamic domains, i.e., by finding compact fragments that move as coherent units. Typical approaches, based on calculating a dynamical cross-correlation matrix, are limited by their inability to reveal correlated rotations and anticorrelated motions. We propose a geometric approach for finding dynamic domains, where we compare traces of atomic movements in a pairwise manner, and search for their best superposition. A quaternion representation of rotation is used to simplify the complex calculations. The algorithm was implemented in a Java graphical program: Geometrically Stable Substructures (GeoStaS). The program processes PDB and DCD binary files with large structural sets for proteins, nucleic acids, and their complexes. We demonstrate its efficiency in analyzing (a) ensembles of structures generated by NMR experiments and (b) conformation sets from biomolecular simulations, such as molecular dynamics. The results provide a clear description of the molecular movements even for large biomolecules. Compared to a standard dynamic cross-correlation matrix, our algorithm detects the correlations in both translational and rotational motions.

On the impossibility of determination of stepwise binding constants for the 1 : 2 complex of (+)-camphor with alpha-cyclodextrin.

Knowledge of stepwise binding constants for complexes with higher than 1:1 stoichiometry would allow one to study the cooperativity of their formation. However, a detailed analysis of partitioning of the overall binding constant beta 12 determined by NMR titrations for the 1:2 complex of (+)-camphor with alpha-cyclodextrin into the stepwise ones K1 and K2 carried out analogously to published procedures revealed that the partitioning cannot be carried out unequivocally for K1 << K2. The programs for partitioning cannot be used as a black box and a satisfactory reproduction of the experimental dependence of relative shifts as a function of relative CD concentration should not be the only criterion of the reliability of the stepwise binding constants obtained using such programs.