Matteo Zago1, Chiarella Sforza2, Alessia Bona3, Veronica Cimolin4, Pier Francesco Costici5, Claudia Condoluci6, Manuela Galli7. 1. Department of Electronics, Information and Bioengineering (DEIB), Politecnico di Milano, Piazza Leonardo da Vinci 32, 20133 Milano, Italy; Department of Biomedical Sciences for Health, Università degli Studi di Milano, via Mangiagalli 31, 20133 Milano, Italy. Electronic address: matteo2.zago@polimi.it. 2. Department of Biomedical Sciences for Health, Università degli Studi di Milano, via Mangiagalli 31, 20133 Milano, Italy; Institute of Molecular Bioimaging and Physiology, National Research Council, Segrate, Italy. Electronic address: chiarella.sforza@unimi.it. 3. Department of Electronics, Information and Bioengineering (DEIB), Politecnico di Milano, Piazza Leonardo da Vinci 32, 20133 Milano, Italy. Electronic address: alessia.bona@mail.polimi.it. 4. Department of Electronics, Information and Bioengineering (DEIB), Politecnico di Milano, Piazza Leonardo da Vinci 32, 20133 Milano, Italy. Electronic address: veronica.cimolin@polimi.it. 5. IRCCS Pediatric Hospital Bambino Gesù, Piazza Sant'Onofrio 4, 00165 Roma, Italy. Electronic address: costici@libero.it. 6. IRCCS San Raffaele Pisana, Tosinvest Sanità, Via della Pisana 235, 00163 Roma, Italy. Electronic address: claudia.condoluci@sanraffaele.it. 7. Department of Electronics, Information and Bioengineering (DEIB), Politecnico di Milano, Piazza Leonardo da Vinci 32, 20133 Milano, Italy; IRCCS San Raffaele Pisana, Tosinvest Sanità, Via della Pisana 235, 00163 Roma, Italy. Electronic address: manuela.galli@polimi.it.
Abstract
BACKGROUND: The relationship between gait features and coordination in children with Cerebral Palsy is not sufficiently analyzed yet. Principal Component Analysis can help in understanding motion patterns decomposing movement into its fundamental components (Principal Movements). This study aims at quantitatively characterizing the functional connections between multi-joint gait patterns in Cerebral Palsy. METHODS: 65 children with spastic diplegia aged 10.6 (SD 3.7) years participated in standardized gait analysis trials; 31 typically developing adolescents aged 13.6 (4.4) years were also tested. To determine if posture affects gait patterns, patients were split into Crouch and knee Hyperextension group according to knee flexion angle at standing. 3D coordinates of hips, knees, ankles, metatarsal joints, pelvis and shoulders were submitted to Principal Component Analysis. FINDINGS: Four Principal Movements accounted for 99% of global variance; components 1-3 explained major sagittal patterns, components 4-5 referred to movements on frontal plane and component 6 to additional movement refinements. Dimensionality was higher in patients than in controls (p<0.01), and the Crouch group significantly differed from controls in the application of components 1 and 4-6 (p<0.05), while the knee Hyperextension group in components 1-2 and 5 (p<0.05). INTERPRETATION: Compensatory strategies of children with Cerebral Palsy (interactions between main and secondary movement patterns), were objectively determined. Principal Movements can reduce the effort in interpreting gait reports, providing an immediate and quantitative picture of the connections between movement components.
BACKGROUND: The relationship between gait features and coordination in children with Cerebral Palsy is not sufficiently analyzed yet. Principal Component Analysis can help in understanding motion patterns decomposing movement into its fundamental components (Principal Movements). This study aims at quantitatively characterizing the functional connections between multi-joint gait patterns in Cerebral Palsy. METHODS: 65 children with spastic diplegia aged 10.6 (SD 3.7) years participated in standardized gait analysis trials; 31 typically developing adolescents aged 13.6 (4.4) years were also tested. To determine if posture affects gait patterns, patients were split into Crouch and knee Hyperextension group according to knee flexion angle at standing. 3D coordinates of hips, knees, ankles, metatarsal joints, pelvis and shoulders were submitted to Principal Component Analysis. FINDINGS: Four Principal Movements accounted for 99% of global variance; components 1-3 explained major sagittal patterns, components 4-5 referred to movements on frontal plane and component 6 to additional movement refinements. Dimensionality was higher in patients than in controls (p<0.01), and the Crouch group significantly differed from controls in the application of components 1 and 4-6 (p<0.05), while the knee Hyperextension group in components 1-2 and 5 (p<0.05). INTERPRETATION: Compensatory strategies of children with Cerebral Palsy (interactions between main and secondary movement patterns), were objectively determined. Principal Movements can reduce the effort in interpreting gait reports, providing an immediate and quantitative picture of the connections between movement components.