SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

Accurate and objective measurement of volume changes in infantile hemangiomas (IHs) is essential in routine clinical practice and clinical studies, particularly in the changing therapeutic landscape after the discovery of propranolol. Several bedside techniques for volume measurement have been described in the literature, but an objective method of measurement of this variable, dynamic vascular tumor is lacking. Three-dimensional (3D) phototechnology with data analysis is an up-and-coming technique in the objective measurement of facial volume changes. In this pilot study, the usability and clinical relevance of two methods of 3D stereophotogrammetry in the volume measurement of IH were explored.

Infantile hemangiomas (IHs) are the most common benign tumors of infancy, characterized by rapid growth during the proliferation phase in the first year of life followed by slow regression [1]. The clinical appearance of IHs is variable, which makes assessment of growth and regression difficult, especially in voluminous IHs. Therefore no standard method for measuring IH dynamics during growth and involution exists.

Accurate objective assessment of volume changes in IHs is essential in routine clinical practice and clinical studies. With the discovery of the effectiveness of β-blockers for IHs, there is an increasing need to evaluate and compare therapeutic effects.

In the past few years, three-dimensional (3D) phototechnology has evolved rapidly. Three-dimensional cameras (3D stereophotogrammetry) in combination with specialized software seems useful in assessing objective and quantitative evaluation of volume changes in IHs. A number of reproducibility and validity studies of 3D stereophotogrammetry have been performed [2, 3]. It can be concluded from earlier studies that surface-based registration is an accurate method of comparing 3D photographs of the same individual at different times [2].

To the best of our knowledge, no studies have been performed to investigate the relevance of 3D photographs for IHs. The usability and clinical relevance of two methods of 3D stereophotogrammetry for volume quantification in IHs were explored in this pilot study.

Patients and Methods

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

This prospective study sample comprised 11 IH patients who visited our vascular anomalies clinic. Inclusion criteria were patients with an IH in the head and neck region of the superficial nodular, deep, or mixed type with an indication for propranolol treatment. Three-dimensional photographs of the patients were taken at the start of propranolol treatment (T0) and at the first control visit (T1). A 3D stereophotogrammetric camera setup (3dMDface System, 3dMD Ltd., Atlanta, GA) was used. The camera setup consisted of two pods, each equipped with three digital cameras and a flash [4]. During acquisition, patients were carefully positioned in a natural head position [5]. A trained photographer took all 3D photographs (Fig. 1). Two methods are described for the 3D measurement of the changes between two different moments in time.

Method 1: Superimposing Images

The 3D photographs taken at T0 and T1 were superimposed using the surface-based matching tool of the Maxilim software [2]. In medical imaging, this matching procedure is referred to as surface based registration. After this registration procedure with volume subtraction, a color map (distance map) can be calculated illustrating the volume differences between T0 and T1 as a color scale image indicating the unchanged areas (in white), decreased volume (in red discoloration), and increased volume (in green discoloration). A higher intensity of discoloration corresponds with a larger change in facial volume. The lighter red areas indicate a small difference (decrease) between the two 3D images, lighter green areas indicate a small increase. Areas with a more intense red or green color indicate a larger decrease or increase, respectively. From this color map, the region covering the IH (which showed a red discoloration) was selected and the mean difference between the photographs could be calculated (Fig. 2).

image

Figure 1. Three-dimensional (3D) camera setup and 3D photograph of a patient with IH.

Download figure to PowerPoint

image

Figure 2. Flowchart method 1: superimposing images.

Download figure to PowerPoint

Method 2: Mirroring Images

Surface-based registration was applied in the second method as well, but in another way. The 3D photograph taken at T0 was mirrored and aligned with the original 3D photograph. The color map was subsequently computed and the volumetric difference between T0 and T1 was calculated. This mirroring procedure was validated earlier and described by Verhoeven et al [6]. The procedure was repeated with the 3D photograph taken at T1, resulting in two mean differences and two volumes. By subtracting the post-treatment volumes from the pretreatment volumes a difference could be computed (Fig. 3).

image

Figure 3. Flowchart method 2: mirroring images.

Download figure to PowerPoint

Results

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

This study included 11 patients (8 girls, mean age 4.3 months at the time of the first photograph; range 2.0–12.5 months). T0 was the time of the first photograph at the start of propranolol treatment, and T1 was the time of the first control visit (mean 9.1 weeks after starting treatment, range 3.1–16.6 weeks).

Table 1 shows the characteristics of the IHs, including location, type, and whether the IH crossed the facial midline. Table 2 illustrates the volume differences measured between T0 and T1 for both methods.

Table 1. Characteristics of IH and Time Between Photographs
PatientSexLocation of IHType of IHCrossing the facial midlineTime between photographs (wks)
  1. a

    Patient in Figs. 1, 2, and 3.

1aMaleUpper eyelidNodularNo3.1
2FemaleForeheadCombinedNo6.9
3FemaleOrbitalDeepNo16.6
4FemaleNostrilCombinedYes3.9
5FemaleCheekDeepYes14.6
6FemaleUpper eyelidDeepNo10.6
7MaleNose tipNodularNo6.9
8FemaleOcciputNodularYes13.4
9FemaleUpper eyelidNodularNo8.0
10FemaleCheekDeepNo7.4
11MaleUpper eyelidNodularNo8.6
Table 2. Absolute Volume Differences After Propranolol Treatment Between T1 and T0 in Both Methods
PatientMethod 1(cm3)Method 2 (cm3)
  1. a

    Patient in Figs. 1, 2, and 3.

1a−2.5−2.3
2−3.3−3.3
3−5.6−4.4
4−1.9Not suitable for method 2
5−4.0Not suitable for method 2
6−4.9−4.8
7−1.6−1.5
8−4.2Not suitable for method 2
9−2.6−2.6
10−2.0−2.8
11−0.6−0.6

Discussion

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

IHs have dynamic growth patterns and great clinical variability, complicating volume measurements. In most studies, 2D photographs have been taken to monitor the evolution of the IH, but accurate volumetric measurements cannot be derived from this technique, only overall evaluation and follow-up. To the best of our knowledge, only three publications[7-9] have described bedside techniques for estimating IH volume. In addition to the inevitable interobserver variation, these methods model IHs as perfect spheres, half spheres, or ellipsoids, making them moderately suitable for assessment of irregularly shaped IHs.

Radiologic diagnostics such as serial duplex ultrasonography may be applicable but are operator dependent. Other imaging techniques such as magnetic resonance imaging and computed tomography are generally not practical for measuring changes in volume because of costs, the need for sedation, and the involvement of invasive radiation.

Since 2008 propranolol emerged as a successful treatment option for complicated IHs. Subsequently several additional new therapies have been reported [10, 11]. With the advent of these novel therapies, there is an increasing need to evaluate and compare treatment effects and possible regrowth or relapse after cessation of treatment, but also to evaluate the course of untreated IHs in functionally important areas with possibly severe complications (e.g., eye, nose). These findings may be important in considering whether to treat.

Three-dimensional stereophotogrammetry can be useful in daily clinical practice in evaluating facial volume changes not only with subjective parameters, but also with objective measurements. An increasing number of hospitals have acquired this technology. The time involved in image acquisition is limited and is a function of correct positioning of the patient. The time needed to capture high-quality “external surface” photographs using this technique is less than 2 milliseconds, which makes it ideal for collecting 3D data from faces, even in children and babies. Reconstruction of the 3D image takes 30 seconds of computational time, and postprocessing of the images takes approximately 15 minutes per case. With adapted software, this technique is also applicable to IHs at some other sites on the body, but it is most suitable for the head and neck region.

Volumetric registration of IHs using two methods of 3D photography was explored in this pilot study. Method 1 is more basic and calculates the difference in the region of interest of two photographs taken at different times. This method can be used for voluminous IHs, even for IHs crossing the midline of the body. The major drawback of this technique is that the effect of the growth of the child cannot be excluded, although this is inevitable with any technique of volume measurements of tumors occurring in infancy and childhood.

Method 2 is slightly more complicated and uses mirroring of the face to calculate the volumetric difference at two different times. A disadvantage of this method is that it is applicable only for unilateral IHs, and it is based on facial symmetry as a baseline. The major advantage is that the effect of growth is minimized. The second method may be more accurate for IHs not crossing the midline, especially in the case of longer time intervals between the measuring points. When the times between photographs are closer together, either method may be suitable.

Given the small number of patients enrolled in this pilot study, no statistical calculations were performed to illustrate comparability of the techniques. The focus of the study was the applicability. A larger study with more patients will follow to determine the statistical difference between the techniques more accurately.

Conclusion

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References

Three-dimensional stereophotogrammetry is a promising, new, accurate, fast, noninvasive way to determine and compare volumetric changes in IHs.

References

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. Conclusion
  7. References