Integrazione di una matrice di sensori di pressione per il controllo delle perturbazioni imposte al soggetto mediante una piattaforma a 3 GdL per il training e per la posturografia dinamica.

The present thesis born from a personnel formative path that has induced me to deal in rigorous way with the theme of the measure in rehabilitation. My formation begins with two years in engineering at the university "Sapienza" of Rome, then abandoned and concluded subsequently in completely different field with the attainment of the degree in Physiotherapy and later of the Master Degree in Sciences of the Rehabilitation. During the proceeding in my rehabilitative activity I realized that to introduce the rehabilitation in a scientific context was necessary to bring back observations in an objective and repeatable framework and it was necessary to conduct accurate measures under controlled conditions. I began building in house an opt-electronic system for the capture at distance of body movements, such to allow observations on the multi-joint body segments during the execution of natural movements. This experience brought me to migrate to the children’s hospital “Bambino Gesù” when in 1999 the realization of a laboratory of movement analysis inside the Department of Paediatric Rehabilitation was planned. Although, the study of the movement was useful to characterize the strategies of movement and their modifications following cycles of therapy, further tools were necessary to dose and to verify the efficacy of the therapeutic exercise. It was perfectly clear to me that was possible to start to use robots able to provide a controlled context for the exercises, this is the foundation for several attempts to separate and to correlate the trainings with their effects. A robot in its simpler conception is a tool that consists of sensors, actuators and of a control system. The presence of sensors opens two perspectives: the first is that they allow an interactive control of the instrument on the basis of the measure of the force exchanged with the subject; the latter is the possibility to measure some characteristics of the interaction between man and robot, in our case the therapeutic relationship. That is, measuring to control (the therapy) and measuring for characterizing (motor behaviours). In such context the meeting with a group of experienced engineers changed the perspective. Their initial purpose was to guarantee the reliability and accuracy of the measures conducted in the movement analysis laboratory but then the solicitations operated in the direction of the emergent perspectives in rehabilitation led them to move towards robotics perspective. We begin to transform of mechanical passive equipment in active robotics objects for the rehabilitation. From this collaboration the 3D robotics platform that rotate around a central pivot was realized. Initially it was controlled in position with pre-programmed movements exploiting the control characteristics of the three linear actuators. In this initial phase the strength's sensors was used only to detect the interactions dynamic strengths and moments of the subjects placed over the platform. Afterwards, the information of the load cells were been used to realize a control in strength tuneable through two variables, the stiffness and the damping. In such way the point of equilibrium of the platform is regulated simulating its suspension on springs with defined elastic characteristics. This evolution of the platform has opened the possibility to define rules of dynamic interaction between the subject and the support base. The transfer of the load from one limb to the other and from the toe to the heel of the foot could be an essential information in the cycle of useful measures to complete the potentialities of interactive control. We decided to proceed to insert on the surface of the platform a pressure matrix with the purpose to detect information on the Centre of Pressure (CoP) that should conduct us to the individualization of useful laws of dynamic interaction in order to recovery the most efficient distribution of the interactions with the ground that was always altered under pathological conditions. The interaction with different balance tasks and contexts will allow to train the "animate tower" among the multiple equilibrium solutions and the diversified use of its informational channels. This process promises to conduct us to the selection and individualization of specific rehabilitative trials that can also be exported in other contexts where the use of the robotics platform is not quite necessary. Furthermore, the platform projected is ready to be transformed in a lower part object cost in order to facilitate its distribution. This brief introduction contains history, motivations and perspectives of the present work and an implicit plan of this thesis. In the Chapter 1, the relationship between equilibrium and perception is described. In the Chapter 2 is analysed the literature on postural motor control mechanisms. In the Chapter 3 is studied the mechanisms of postural control under normal and pathological conditions. Fig. A.1 relationship between sensory condition and body sway. From Motor Control, Woollacott (2003). Sensory conditions contribute differently to posture stability, figure A.1 resumes the sway increase related to different sensory conditions. Body segments sway with different amplitude in relation to task and context too. Figure A.2 showed different body segment sway area with eyes open and closed in response to 3D continuous wave perturbation. Fig. A.2 Sway areas of the center of the mass at floor and of the body segments following 3D continuous perturbation with eyes open (A) and eyes closed (B). The two sensory condition showed a different coordination respect the same perturbation, in the first condition we could observe that the head is the most stable segment, while in the latter condition the most steady section is the pelvis. In the Chapter 4 is described the 3D platform used in the present thesis. Figure A.3 illustrates the RotoBIT3D dynamic platform that rotates around the central pivot driven by three axial actuators with a workspace of about ±10° for each rotational axes, roll, pitch and yaw respectively. The position of the platform was calculated from its known geometry, illustrated in Figure A.4, while the force control is realized simuling a 3D spring with rigidity K and equilibrium point γ0 in parallel with a 3D damping of coefficient C. The dynamic system is influenced by the torque in the Cartesian space designed in relationship with the Cardanic components TTM. Eq. A.1 In the Chapter 5 is described the pressure matrix, Figure A.5, and its measure characteristic. The Centre of Pressure (CoP) was studied in comparative experiments with measures gathered both by the sensors placed on the 3D platform and the pressure matrix. Figure A.6 and A.7 illustrates CoP comparison with the Platform controlled in force (dynamic condition) for an healthy young and a subject with atassia, respectively. Fig. A.6: Compared CoP data for an healthy subject collected by mean of Platform, rb apex, and pressure matrix, tk apex. Fig. A.7: CoP compared measures between robotic platform and pressure matrix in dynamic condition in a patient with atassia. In the chapter 6 is illustrated the innovative protocols realized for the evaluation of the equilibrium in static and dynamics conditions using the information provided by the matrix of pressure, for healthy and pathological subjects. In static condition the platform is fixed in horizontal position while in dynamic condition the platform is controlled in force simulating the suspension of the platform on a spring. In Figure A.8 and A.9 are illustrated the mean oscillation frequency that remain constant for both dynamic and static condition, with the same tendency for all pathologies. Fig. A.8: Mean oscillation frequency for patients with spinal cord lesion M, atassia A, hemiplegia E, diplegia D, in static stance condition Fig. A.9: Mean oscillation frequency for patients with spinal cord lesion M, atassia A, hemiplegia E, diplegia D, in dynamic stance condition. * indicates statistic significance. The dynamic condition appears more sensible to detect the real balance competence and this hypothesis is confirmed when the oscillation frequency in static and dynamic condition is compared with gait velocity as shown in Figure A.10 and A.11. Fig. A.10 Relationship between frequency in static condition and gait speed in subjects with balance disorders. Fig. A.11 Relationship between frequency in dynamic condition and gait speed in subjects with balance disorders. In the Chapter 7 are described innovative procedures of dynamic interaction among the platform and subject finalized to the training of the equilibrium and the walking. The subjects were trained requiring them to reach object in different space position while they standing on the platform controlled in force. In that condition the subject experiment self and external source of unbalancing. To assess the improvement during training a new test condition were developed, Figure A.12. Fig. A.12 Schematic representation of the new test condition. The platform rotate of 6° along eight directions while it is controller in force. The subject had to move the platform back to the horizontal position without any feet movements. In figure A.13 it is possible to observe the results obtained after two different training, the first with traditional therapy, while the latter utilizing the new treatment proposed in this thesis. Fig. A.13 Eight equilibrium point plots repeated three times at the first day of treatment (T0), at 15 days of treatment and at 30 days of treatment in a subject with traumatic brain injury. Two different treatments were conducted between T0-T15 and T15-T30 periods. It is possible to recognize a reduction of the oscillation only at T30 assessment. In the Chapter 8 is discussed in comparative way the results related to the administrated tests and training. The fundamental purpose of the thesis is to furnish innovative therapeutic tools, while for me it represented an opportunity to deepen the knowledge on the fundamental chains of measure to achieve results of quality.

L’argomento della tesi nasce dalla coniugazione di competenze riabilitative, di misura e robotiche. Solo così è stato possibile formulare l’ipotesi che una matrice di pressione potesse essere integrata sulla pedana robotica RotoBiT3D (capitolo 5), già sviluppata dal gruppo di ricerca (capitolo 4), con finalità di valutazione e trattamento in un contesto di riabilitazione sensori-motoria dell’equilibrio e del cammino (capitolo 1, 2 e 3). Il dottorando ha fornito il disegno del protocollo riabilitativo della pedana RotoBiT3D contribuendo in tal modo alla definizione del disegno complessivo finale (capitolo 4). Il dottorando ha fornito indicazioni sulla modalità di controllo della piattaforma in forza utile alle finalità di uso in riabilitazione (paragrafo 4.4). L’integrazione della matrice di forza alla pedana così come la realizzazione dell’interfaccia software è stata condotta dal gruppo di ricerca. Il dottorando ha contribuito definendo la struttura di raccolta dei dati e gli indicatori da utilizzare per l’analisi del Centro di Pressione (CoP) (capitolo 5 e 6). Il dottorando ha contribuito alla analisi ed alla interpretazione delle prove di misura comparata dinamica del CoP sulla piattaforma, condotte su soggetti sani (capitolo 5). Il dottorando ha effettuato in autonomia la conduzione, analisi e interpretazione delle misure comparate del CoP in condizioni statiche e dinamiche su soggetti con disturbi di equilibrio (capitolo 5 e 6). Il dottorando ha condotto in completa autonomia: i) la selezione dalla letteratura corrente degli indicatori di analisi del CoP (paragrafo 6.4); ii) la selezione dei pazienti con disturbi dell’equilibrio in esito a diverse patologie (paragrafo 6.3); iii) la conduzione delle prove di stabilometria con pedana robotica statica e dinamica (paragrafo 6.2), iv) l’analisi dei risultati collezionati, organizzandoli secondo gli indicatori selezionati (paragrafo 6.5); v) l’analisi statistica dei risultati, comparando le prove statiche con quelle dinamiche e dividendo i risultati per patologia dei soggetti esaminati (paragrafo 6.5); vi) la formulazione delle ipotesi interpretative (paragrafo 6.6); vii) l’individuazione di ipotesi di trattamento congruenti con i risultati ottenuti e le attuali teorie sensori-motorie di controllo del movimento (paragrafo 7.1 e 7.2); viii) la conduzione di prove preliminari di verifica del trattamento ipotizzato su un paziente test (paragrafo 7.3). ix) la verifica dei risultati di addestramento ottenuti (paragrafo 7.3); I risultati ottenuti da questa ricerca hanno permesso: i) la realizzazione di un contesto di valutazione e trattamento dei disturbi di equilibrio fra i più naturali tra quelli proposti in letteratura (capitolo 7 e 8); ii) l’individuazione della frequenza di oscillazione come possibile variabile di controllo del bilanciamento, risultato emerso dalla comparazione delle misure del CoP in condizioni statiche e dinamiche (paragrafo 6.5, 6.6, 7.1 e 7.2); iii) di evidenziare il maggiore contenuto di informazione delle prove dinamiche rispetto a quelle statiche (paragrafo 6.5, 6.6 e capitolo 7 e 8); iv) di approfondire la relazione esistente tra la capacità di bilanciamento in condizioni statiche e dinamiche e l’abilità di cammino (paragrafo 6.5); v) di evidenziare l’efficacia dell’innovazione terapeutica nei disturbi di equilibrio permessa dal controllo in forza della pedana robotica RotoBiT3D (capitolo 7 e 8).