**5.6 Adding intra-abdominal pressure**

Since LifeMOD and ADAMS provide tools that only generate concentrated or distributed forces, it is not possible to implement directly intra-abdominal pressure into the spine model. To overcome this difficulty, a new approach to intra-abdominal pressure modeling is proposed. Initially, an equivalent spring structure able to mimic all mechanical properties of intra-abdominal pressure such as tension/compression, anterior/posterior shear, lateral

Development of a Detailed Human Spine Model with Haptic Interface 181

and shear forces as well as the moment about the L5-S1 disc were calculated. Zee's model estimated an axial force of 4520 N and shear force of 639 N in the L5-S1 disc at a maximum extension moment of 238 Nm. Meanwhile, to obtain the same extension moment, the external force that needs to be applied in the present model is 1260 N. Corresponding with this force, axial and shear forces obtained in the model were 4582 N and 625 N, respectively. This is in accordance with the results presented by McGill et al. (1987) who found axial

In the second approach, a comparison was made with in-vivo intradiscal pressure measurements of the L4–L5 disc as reported by Wilke et al. (2001). They measured a pressure of 1.8 MPa in the L4–L5 disc while the subject (body mass: 70 kg; body height: 1.74 m) was holding a full crate of beer (19.8 kg) 60 cm away from the chest. The disc area was 18 cm2 and based on this the axial force was calculated to be 3240 N. The same situation was simulated using the spine model in this research. The estimated axial force was 3161.6 N. This is a good match considering the fact that no attempt was made to scale the model to the subject in this study. Body mass and body height of the subject in this study are quite

Haptic rendering is the process of applying forces to give the operators a sense of touch and interaction with physical objects. Typically, a haptic rendering algorithm consists of two parts: collision detection and collision response. Figure 18 illustrates in detail the procedure of haptic rendering. Note the update rate of haptic rendering has to be maintained at around 1000 Hz for stable and smooth haptic interaction. Otherwise, virtual surfces feel softer. Even

To observe the locomotion and study dynamic properties of the spine model quickly, conveniently and more realisticly, haptic technique can be integrated into a spine simulation system. In this study, for two types of spine models developed using finite element method as well as multi-body method in LifeMOD software, the haptic rendering process has two main stages: the rigid stage and the compliant stage. Without pressing the stylus button of the PHANToM device, the users can touch and explore the whole spine model since it is considered to be rigid throughout. After the users locate a specific vertebra where he/she wishes to apply force, they can then press the PHANToM stylus button and push or drag the vertebra in any direction to make the whole spine model deform. Once the stylus button is pressed, the system switches from rigid stage to compliant stage. The haptic rendering algorithms in these two stages will be clearly presented in the following sections. Figure 19

In real-time haptic simulation, users can only interact with the spine model by manipulating a rigid virtual object considered as a probe on the computer screen. At present, a simple probe such as a sphere is used in this study. In the rigid stage, a common haptic rendering method is used for the aforementioned types of spine models. Since this interaction carries out at a high update rate of 1 kHz, the chosen haptic rendering method needs to be

forces in the range of 3929–4688 N and shear forces up to 650 N.

similar to the body mass and height used in the model.

shows the complete haptic simulation process in the system.

**6.2 Haptic rendering method for the rigid stage of the spine models** 

**6. Haptic interface for spine modelling** 

**6.1 Haptic rendering** 

worse, the haptic device vibrates.

shear, flexion/extension, lateral bending and torsion is created (Figure 16). After that, the translational and torsional stiffnesses of the spring structure are determined. Finally, since adding this spring structure into the spine model is quite troublesome, a bushing element that can specify all stiffness properties of the structure is used instead (Figure 17).

Fig. 16. The spring structure used in the spine model
