Tlemsafe Project Deliverable Report 1.2: Functional Measurements (Healthy)

TLEMsafe project Deliverable Report 1.2: Functional measurements (healthy)

Deliverable Report Work Package 1, deliverable 1.2

Database of functional measurements (healthy)

Start – End

M1 (01-03-2010) – M19 (01-11-2011)

Involved partner

Radboud University Nijmegen Medical Centre (RUNMC)

Summary

The work in Work Package 1 (WP1) revolves around functional measurements of healthy subjects. The data of these functional measurement serves as input to Work Package 3 of the project, in which the outcomes of subject-specific models are to be compared to their corresponding actual outcomes from the functional measurements.Work Package 1 contains a total of four deliverables. This report describes deliverable 1.2: ‘Database of functional measurements (healthy)’. First, a description of the work according to Annex 1 is given. Then, the work that has been performed towards the construction of the database is presented.

The database contains five measurement types (MRI scans, maximum strength tests, functional tasks, FDG-PET scans and oxygen consumption measurements) divided into three measurement sessions. All of the measurement sessions have been performedand the database is complete. Some of the data in the database is not directly suitable for placement on the website and needs to be processed before it can be used for validation of M-S models. This includes the functional task data, the PET/CT scans, and oxygen consumption measurements. Most of this data has already been processed and will be uploaded in the very near future (M23).

Table of contents

Start – End

Involved partner

Summary

Table of contents

Description of the work

Work performed

NIRS

Ultrasound

Oxygen consumption

FDG-PET

Introduction

Summary literature study

Conclusion literature study

Functional tasks

Strength measurements

Current status and finalization

Timing

Description of the work

Annex 1 describes an extensive set of measurements to be performed on 10 healthy subjects, to construct a database with which the project can validate musculo-skeletal models. They are:

Near-InfraRed Spectroscopy (NIRS);

Oxygen measurements;

Ultrasound;

FluoroDeoxyGlucose - Positron Emission Tomography (FDG-PET);

Electromyography;

Isometric tests;

Kinematic and kinetic analysis:

• Gait;
• Stepping over an obstacle;
• Slow/Fast walking;
• Stepping after perturbation;
• Stair negotiation;
• Getting up from a chair.

Work performed

The first tasks that were performed towards the construction of the database were a thorough literature study and consultations with experts in the field into the usefulness of each of these techniques for the TLEMsafe project. The results will be discussed in the following paragraphs.

NIRS

In light of the requirements of TLEMsafe from this technique (validating the relative contributions of muscles to joint torques), the most important disadvantage of NIRS is its high correlation with electromyography (EMG). EMG is already beforehand in the laboratory of the RUNMC. Several other and interrelated disadvantages are:

1. NIRS only measures 2-3 cm deep under the skin, so only superficial muscles can be measured with it. Note that NIRS shares this limitation with EMG.
2. Subcutaneous fat is a major problem. If the subject has a thick adipose tissue layer,

the light from the NIRS probe becomes scattered before it reaches the muscle. Those

subjects' muscle oxygenation simply cannot be measured with NIRS.

1. The relation between adipose tissue thickness and differential pathlength factor (DPF) is not well-studied. This makes it hard to come up with a subject-specific value for the DPF at a muscle site.
2. According to Hiroyuki et al. (2002), NIRS is not able to provide a quantitative value of deoxygenation and/or oxygenation. This is because the pathlength of NIRS radiation cannot be measured with the device. Optical path length is determined by subcutaneous

fat.

In conclusion, although the NIRS technique appeared useful for the TLEMsafe project for measurement of muscle oxygenation, and thus estimation of muscle energy consumption and developed force, we were unable to find evidence in the literature for an added value of NIRS over EMG in the measurement of muscle activation, ability to solve the load-sharing problem, and thus, validation of M-S models. Moreover, the purchase costs of the equipment are high (~30000 Euros). Therefore, the NIRS technique was not used in TLEMsafe. The exclusion of NIRS did not endanger the project objectives, nor did it delay them. Plenty of other measurements were performed that we can use to validate the musculoskeletal models.

Ultrasound

In the past decade, ultrasound has become a popular measurement technique for imaging muscle parameters such as fibre length, muscle belly thickness, tendon length and pennation angle during the performance of various exercises such as walking, cycling and stair negotiation. However, recently, critical questions have been raised about the reliability of these measurements (e.g. Bénard 2009). In that study, using standard measurement protocols yielded errors of fascicle length and fascicle angle measurements up to 14% and 23% from cadaveric measurements. These results indicate that if the measurements are to be taken in a reliable way, a lot of specialized knowledge and equipment is required. Even if such commodities would be available, ultrasound can only be used on superficial muscles that are not covered by thick fatty tissue. These requirements exclude almost all hip and thigh muscles, which are of utmost importance to measure for the TLEMsafe project; the afflictions that will benefit most likely from the results of the project are those that concern the hip and thigh regions. Moreover, only a single muscle can be imaged during a task, and therefore the influence of performing this measurement, even if it were reliable and valid, would be small on the whole musculoskeletal model predictions. To measure more muscles, the ultrasound probe would need to be moved a lot between trials to image several muscles for a single exercise. This would greatly increase the experimental burden and duration. Concluding,it was decided that ultrasound would not be used in the TLEMsafe project. The exclusion of ultrasound measurements did not endanger the project objectives, nor did it delay them. Plenty of other measurements were performed that we can use to validate the musculoskeletal models.

Oxygen consumption

This technique has been widely used in the literature for many decades. It measures total body oxygen consumption and, if the exercise is aerobic, thereby provides a measure of total body energy consumption. With the energy consumption information, the total musculoskeletal model energy consumption can be validated. The most suitable exercises that can be monitored with this technique are periodic, low-intensity exercises such as walking or cycling. We measured oxygen consumption during walking, because walking is a more important activity in daily life than cycling. Even though this measurement technique provides no information about individual muscle energy consumption, the measurement is very easy to perform and harmless to the patient, and has therefore found a place in the measurement protocol in the same session as the PET/CT scans.Example results are shown in figure 1.1.

FDG-PET

Introduction

In the validation of subject-specific musculoskeletal (M-S) models, such as created in the TLEMsafe project, some of the most important aspects that need to be validated are muscle forces and muscle energy consumption. Unfortunately, these are not easy to measure in vivo. Therefore, muscle activity is often estimated from indirect measurement techniques such as electromyograpy (EMG), which measures electrical activity of action potentials, or near-infrared spectroscopy (NIRS), which measures oxygenation of muscle tissue. Though both of these techniques can be used to produce an estimate of produced muscle force and muscle energy consumption, they share several disadvantages. The number of muscles that can be measured simultaneously is limited by the number of electrodes/optodes available, inferior muscles cannot be measured, and adipose tissue is a major confounding factor. These disadvantages lead us to investigate PET as a technique with which we can validate M-S models. A summary of the literature study is given in the following paragraph.

Summary literature study

Pallas et al. (2001) found that FDG uptake was significantly correlated with the number of repetitions of a dynamic elbow flexion exercise. In that same study, the ratio of the slopes of the regression lines was 4:94 when a five times heavier weight was used in the exercise, indicating that FDG uptake is also related to exercise intensity. Finally, FDG-PET was also able to differentiate functional differences between the recruitment of the soleus and gastrocnemius muscles as a function of knee angle. These results indicate that the level of detail that can be obtained with FDG-PET technology suffices to functionally distinguish contributions of individual muscles to, for instance, moment production around a joint. In an FDG-PET study that investigated level walking, it was found that aerobic activity was higher in the lower leg than in the thigh (Oi et al. 2003). This shows that literature findings on the activity of muscles during level walking found with EMG (e.g. Perry, Gait Analysis), and kinematic and kinetic (e.g. Winter (1983)) methods, can be confirmed with PET. Moreover, Oi et al. (2003) reported a high uptake of FDG in the gluteus minimus compared to the gluteus maximus and medius, suggesting an important role of that muscle in pelvis stabilization during level walking. The gluteus minimus is one of the inferior muscles that is almost impossible to monitor with EMG. This indicates that PET is capable of measuring muscles that are situated inferiorly.

[18]FluoroDeoxyGlucose Positron Emission Tomography (FDG-PET) is a technique that offers a solution to these problems. PET images provide metabolic information and are usually displayed on top of an MRI image or a CT image, so that the combination of the two provides both anatomical and physiological information (see figure 1.2 and figure 1.3). In TLEMsafe, the MRI images of subjects are already available since those scans were made anyway, so the use of PET did not entail any additional MRI scans. The muscle regions of interest on the PET images can be found by referring to the earlier made MRI scans of that subject. Then, muscle energy consumption can be estimated from the metabolic information of the PET image.

Figure 1.2: Typical FDG-PET image of the lower extremity. Note the extensive metabolic information. / Figure 1.3: Typical CT image that was taken during the same scan as the PET image. With reference to this image, regions of interest (i.e. muscles) can be selected.
/ Figure 1.3 Example of Regions Of Interest (ROI’s) drawn on a combined PET/MRI picture. Individual muscle boundaries can be seen very well on the TLEMsafe MRI scan. Then, the PET data shows the counts/ml of the ROI (i.e. muscle tissue). This is actual data of one of the measurements of healthy subjects.

Conclusion literature study

From the literature study, we concluded that PET can be used to indirectly ascertain the relative contribution of individual muscles to joint torques, which is very useful for validating the load-sharing solution implemented in TLEM. It offers an added value over EMG, because inferior muscles can be imaged with PET, and all the muscles of the lower extremity can be measured at once without additional burden. Moreover, PET recordings are not influenced by fatty tissue, whereas EMG signals are (e.g. for the gluteus maximus, an important muscle for walking, it is hard to obtain a good EMG signal). Finally, using EMG, the intensity of muscle contraction is calculated relative to the maximum voluntary contraction, making it impossible to compare the intensity of muscular activity among multiple muscles. PET circumvents this problem, because its glucose uptake values do not have to be calculated relative to a maximum voluntary contraction; they can be compared directly. A limitation inherent to the technique is its cumulative nature. In the validation of M-S models, the timing of muscle activities is of utmost importance, and PET is unsuitable for measuring temporal variations in its signal intensities. Therefore the technique will never be able to completely replace EMG for muscle activity measurements, but to complement it. Also, PET images do not have a high resolution when compared to, for instance, MRI images. However, when dealing with large muscles such as those in the lower extremity, this is not a major drawback.

Concluding, we decided to use FDG-PET in the TLEMsafe project.In Annex 1 it was mentioned that five subjects would undergo this measurement, but all 10 healthy subjects were included instead, because of the great value that the PET measurement has for the validation of musculoskeletal models.Besides the benefits of FDG-PET for the project, the technique has never been used in the validation of M-S models, providing interesting possibilities for publication.

Functional tasks

This session consists of exercises that reflect activities of daily living and exercises that help the model determine the range of motion. Although not all of these measurements could be called ‘functional’, they have been placed in a single category because they were measured in the same session of the measurement protocol. The exercises that were performed are:

-Ascending/descending stairs;

-Getting up/sitting down on a chair;

-Stepping over an obstacle;

-Initiation/termination of gait;

-Gait at moderate/slow/maximum speeds;

-Lunge, sideways direction;

-Lunge, forward direction;

-Lunge, forward direction with additional weight (e.g. weighted girdle)

-Deep squat with varying distance between heels;

-Ankle rise;

-Hips describe a semi-circle pattern;

-Hips make large motions through range of motion;

-Knee flexion and extension;

-
Feet inversion/eversion;

-Ankle flexion/extension;

During the execution of the measurement protocol, the subject has reflective markers and EMG transmitters attached to his/her body as shown in figure 1.4. Example plots of EMG and force data from these experiments can be found in figure 1.5.

Strength measurements

The maximum force measurements (called ‘isometric (static) tests’ in Annex 1) have been measured in a separate session. The isometric and isokinetic measurements were performed using a Biodex S4 isokinetic dynamometer (figure 1.6). This dynamometer was used because the dynamometer that was already present at the Rehabilitation lab and that we originally planned to use broke down in April 2011. A solution was found in the form of an added measurement session at the Sports Medical Center of the Sint Maartensclinic in Nijmegen. Although it forced us to add a measurement session to the protocol, this allowed us to use a more sophisticated machine to test isokinetic and isometric strength.

The tests that were performed are:

-Isometric maximum moment measurements at three angles per joint;

• Hip flexion/extension 0, 30, 60 degrees;
• Hip abduction 0, 15, 30 degrees;
• Kneeflexion/extension 90, 60, 30 degrees;
• Ankleflexion/extension 0, 15, 30 degrees;

-Isokinetic moment measurements of those same joints at a slow speed (60°/s) and a fast speed (120°/s).

Current status and finalization

Immediately after ethical permission was granted, we started including healthy subjects in the study starting from M16. Ten healthy subjects were included in the TLEMsafe project, whose data can be found in table 1. Five male and five female subjects of varying height and weight were included. TheBody Mass Index (BMI) was used as a screening tool; subjects that had a BMI over 30 were excluded from participation for practical reasons (difficulty in applying 3-D tracking markers and EMG electrodes, for instance). The healthy subject population was composed to mimic the patient populations we will include in later stages in the project. Particularly, between 18-30 and over 50 there are incidence peaks in the occurrence of sarcoma.

Gender / DOB / Preferred Leg / Weight (kg) / Height (cm) / BMI (kg/m^2) / CWS (km/h)
Healthy_001 / M / 23-2-1984 / right / 91.7 / 1800 / 28.3 / 5.2
Healthy_002 / M / 26-1-1988 / left / 83.1 / 1820 / 25.1 / 4.5
Healthy_003 / M / 23-10-1985 / right / 90.4 / 1950 / 23.8 / 4.4
Healthy_004 / F / 18-1-1984 / right / 58 / 1661 / 21.0 / 4.3
Healthy_005 / F / 3-6-1988 / right / 77.6 / 1875 / 22.1 / 3.7
Healthy_006 / F / 23-2-1952 / right / 64.5 / 1600 / 25.2 / 4.5
Healthy_007 / F / 22-11-1953 / right / 55.5 / 1625 / 21.0 / 5.1
Healthy_008 / F / 17-12-1950 / right / 70.7 / 1640 / 26.3 / 4.4
Healthy_009 / M / 20-4-1955 / right / 75.9 / 1783 / 23.9 / 4.6
Healthy_010 / M / 20-8-1967 / right / 81 / 1735 / 26.9 / 4.6
mean: / 74.8 / 1748.9 / 24.4 / 4.5
std: / 11.9 / 110.7 / 2.3 / 0.4

Table 1: Gender, date of birth, preferred leg, weight, height, body mass index (BMI) and comfortable walking speed (CWS) of all 10 healthy volunteers who participated in the TLEMsafe project. Averages with standard deviations are also given.

Each subject underwent three measurement sessions in the period from M17-M23, consisting of five different measurement techniques.The measurements haveall been performed and the database is complete. For a complete overview, see Table 2.

MRI / Functional tasks / Strength / PET/CT scan / Oxycon
Healthy_001 / completed / M19 / completed / M18 / completed / M17 / completed / M22 / completed / M22
Healthy_002 / completed / M19 / completed / M19 / completed / M19 / completed / M22 / completed / M22
Healthy_003 / completed / M18 / completed / M19 / completed / M19 / completed / M22 / completed / M22
Healthy_004 / completed / M18 / completed / M19 / completed / M20 / completed / M23 / completed / M23
Healthy_005 / completed / M19 / completed / M18 / completed / M18 / completed / M22 / completed / M22
Healthy_006 / completed / M19 / completed / M18 / completed / M19 / completed / M22 / completed / M22
Healthy_007 / completed / M19 / completed / M19 / completed / M19 / completed / M23 / completed / M23
Healthy_008 / completed / M18 / completed / M19 / completed / M18 / completed / M23 / completed / M23
Healthy_009 / completed / M21 / completed / M20 / completed / M21 / completed / M22 / completed / M22
Healthy_010 / completed / M21 / completed / M20 / completed / M21 / completed / M22 / completed / M22

Table 2: Overview of all the measurements that have taken place in WP1, and the month in which they were completed.

At the time of writing (M23), all the measurement data has been collected and is being placed on the project website. Some of the data in the database is not directly suitable for placement on the website and needs to be processed before it can be used for validation of M-S models. This includes the functional task data, the PET/CT scans, and oxygen consumption measurements.Most of this data has already been processed and will be uploaded in the very near future (see Table 3).

MRI scans / Strength / Functional tasks / PET/CT / Oxygen consumption
10/10 uploaded / 10/10 uploaded / 90% processed (expected upload M23) / 100% processed
(expected upload M23) / 90% processed (expected upload M23)

Table 3: Overview of the current status of data processing and upload status.

Timing

We currently regard deliverable 1.2 as ‘done in M23’, even though there are some data that require a few more weeks of processing. There were three factors that contributed to the delay in this deliverable: