Ultrasound in dermatology

ARTICLE

Ultrasound was first introduced as a diagnostic tool in the early 1950s [1]. The major advantage of this technique is its non-invasive, non-ionizing nature of examination and its relatively low cost when compared to X-ray, CT- and scintigraphic scanning techniques. Abdominal, gynaecological, obstetrical, thyroid, and cardiological scanning are now routinely used for medical diagnoses. Dermatosonography, or the use of high-frequency ultrasound with frequencies of more than 20 MHz for the examination of human skin is now a fully developed matured technique offering a wide range of possibilities in clinical and experimental dermatology.

Background

In the late 1970s Alexander and Miller used high frequency ultrasound (15 MHz) to generate unidimensional scans of the skin [2]. Much progress in the development of high frequency scanners has occurred since then, leading to the introduction of the first 20 MHz scanner [3]. The equipment currently available allows both two and three dimensional imaging of the structures of human skin in vivo.

Ultrasound imaging is based on the different acoustic properties of different tissues. The equipment consists of an ultrasound probe (hosting transducer), an elaboration, and a visualisation system. The probe emits ultrasound waves which are transmitted into the tissue, where they are reflected, refracted, or inflected following the optical laws governing interfaces between structures of different acoustic impedance, and analysed after returning to the probe [4]. Ultrasound imaging is, therefore, highly influenced by any changes in the echogenicity of skin structures whereas the acoustic impedance of a tissue is defined as the product of its density and the velocity of sound in the tissue. In human skin ­ as an inhomogeneous tissue ­ the velocity of ultrasound is commonly calculated with 1,580 m/s.

Uni-dimensional scanning is referred to as amplitude mode or A-mode. By moving the transducer laterally in one direction along the skin surface, multiple A-scans can be obtained, with each echo being assigned a typical brightness level thus creating a two-dimensional image. This type of scanning is called brightness mode or B-mode. If the probe is moved in two dimensions (right to left, and forwards to backwards) across the skin surface a three dimensional image is similarly produced, and called computed mode or C-mode scanning.

The time lag between the emitted and reflected signal enables the calculation of the distance between reflecting objects, whereas the amount of reflected energy characterises the echogenicity of a given object. All three types of ultrasonography contain this information, but the visual subjective as well as objective interpretation of the data become progressively less abstract in B- and C-scanning. The distance between different structures as well as echogenicity can be objectively assessed by computerised image analysis. In the most commonly used system, the amplitudes of the echoes of single-image elements (pixels) are assigned to discrete values on a numerical scale (from 0 to 255 with increasing echogenicity), and a weighted average density of the tissue is then presented in a simple numerical form [5]. The frequency of the ultrasound used is a determining factor in both penetration and resolution of the imaging system. The general relationship is that higher frequencies give higher resolution but lower penetration, although other factors such as transducer focusing or gain are also important. The frequencies used for the examination of subcutaneous tumours and regional lymph nodes, which is increasingly performed by dermatologists, are in the range of 3-10 MHz. In contrast, dermatological ultrasonography is primarily carried out with high-frequency scanners of 20-50 MHz. The first results of scanners operating at 100 MHz have also been presented by el-Gammal and co-workers [6]. Sharp focusing transducers which are commercially available at a frequency of 20 MHz, allow for a lateral resolution of 130 mum and a axial of 60 mum, respectively. The use of sharp-focusing transducers, that focus the ultrasound waves in the upper dermis, makes it necessary to adjust the gain compensation curve in order to obtain an equal signal sampling from all skin layers. Since the transducers vary slightly between ultrasound instruments, gain compensation curves should be adjusted accordingly. Gniadecka and Jemec [7] have demonstrated that limb skin showed higher echogenicity than truncal skin. In their report the gain compensation curve was adjusted in the oblique position at 22-28 dB for buttock and forehead and parallelly at 17-26 dB for extremities to obtain quality images from those skin regions. Gain adjustment also allows easier determination of the interface between dermis and subcutis. In some cases the cut-off between the dermis and subcutis is aided by one-dimensional measurements (A-scans) superimposed on the B-mode image. A marked reduction of the peak height on A-scans in the lower dermis indicates the interface between dermis and subcutis. The gain when once adjusted should be kept constant if skin thickness or echogenicity is quantified.

The use of 3-dimensional C-scanning images in clinical dermatology is still hampered by the resolution of the systems available and the relatively high expenditure of time for a measurement. The technique, therefore, remains experimental at present.

Skin, skin appendages and their abnormalities on sonographic image

Initially, A-mode ultrasonography was used for non-invasive evaluation of simple anatomy. This was done primarily by measuring distances between characteristic echoes, e.g., in skin thickness, but with the advent of B- and C-mode scanning its application has expanded into examination of hair follicles [8], skin tumours [9], oedema [10-12], irritant and allergic reactions [13-15], scleroderma [16, 17], psoriasis [18-20], wound healing [21], and skin ageing [22-24].

The typical skin image shows the following layers: an epidermal entrance echo, the dermal layer, and the low- or non-echogenic subcutaneous tissue with obliquely oriented echogenic lines caused by connective tissue bundles (Fig. 1). Depending on the anatomical site the muscle fascia is visualized as an echogenic band. The thickness of the epidermal entrance echo is not correlated with the thickness of the epidermis. Because of the low resolution at 20 MHz, epidermal structures are only visible when they are thickened, e.g., in psoriasis, where hyperkeratosis is prominent, and, consequently, the epidermal echo is enhanced. Special experimental transducers have been developed for the study of epidermis using very high ultrasound frequencies of 50 to 100 MHz [25]. They demonstrate that the epidermis is low-reflectant in its internal structure.

The dermis is less echogenic than the epidermal entrance echo and contains many different echoes of various intensities. It is thought that dermal echoes arise as a result of the reflection of the ultrasound waves from the interface between collagen fibres and the surrounding intercellular matrix and the cells [26]. Some information about the nature of the dermal echo can be deduced from ultrasonography of various diseases involving the dermis. The influence of collagen fibres and their organisation on ultrasound imaging may be illustrated by studying scleroderma and keloids. In scleroderma and morphea there is an accumulation of collagen fibres in the skin, and the echogenicity of the dermis is increased [16]. However, keloids and hypertrophic scars, which are also fibrous tissue, give a homogenous, low-echogenic image. It is thought that the difference is due to the alignment of the collagen fibres. The collagen fibres are tightly packed in the keloids, with only a minimal amount of intercellular substance, whereas there is more intercellular substance in scleroderma, allowing for a greater number of echogenic interfaces [27]. Similarly, it has also been shown that water influx reduces the echogenicity of the dermis, most probably due to a distension of the fibre network.

Diurnal variation of skin echogenicity has been observed which is thought to be caused by the changes in dermal hydration due to postural changes [28]. This phenomenon has also been extensively investigated in studies of skin oedema [12]. As expected, the echogenicity of the dermis in oedema is decreased, but the distribution of the low-echogenic areas in various forms of oedema differ significantly. In the postthrombotic syndrome and lipodermatosclerosis, the reduction in echogenicity is noted primarily in upper dermis, whereas it spans the entire dermis in lymphoedema, and in cardiac insufficiency occurs mainly in the lower portion of the dermis adjacent to the subcutaneous tissue [12]. Removal of oedema with compressive therapy can be monitored by ultrasonography [11] (Fig. 2).

Another reason for a reduced echogenicity of the upper dermis is the infiltration with inflammatory cells observed in inflammatory skin diseases, e.g., lichen ruber planus or acute and chronic eczema. The sonographic image of these diseases is generally that of an echopoor area underneath the entry echo. Because of increasing echogenicity a sonographical monitoring of the therapeutic effect (e.g., after local treatment with corticosteroids) is possible [29]. In contrast, differential diagnosis of various diseases ­ to date ­ is not possible.

Imaging of photodamaged skin revealed a subepidermal low-echogenic band (SLEB), the thickness of which has been argued to be correlated with the severity of photodamage and used to measure the efficacy of anti-ageing preparations [24, 30, 31]. Recent studies of skin ageing where ultrasound images were collected from a number of sun exposed and sun protected skin sites have conclusively shown that the SLEB is present only in sun-exposed dermis [7] (Fig. 4). The origin of SLEB is probably multifactorial and the main contributing factors are alterations in skin fibre structure (collagen and elastin), and accumulation of glycosaminoglycans in the subepidermal region which results in the increased water-binding capacity.

Skin appendages can also be visualized by high frequency ultrasound. Early studies using A-mode scanning have provided insight into the functionally bilamellar structure of the human nail, by demonstrating the presence of a deep layer with an increased water holding capacity [32]. Later studies of nails in B-mode scanning have confirmed the basic bilamellar ultrasonographic structure of the nail, and suggested that this technique may be a useful tool in studies of nail pathology [33] (Fig. 3).

Similarly, both two and three dimensional imaging of hair follicles have been used to identify characteristic predisposing abnormalities in clinically uninvolved follicles in hidradenitis suppurativa patients [8, 34].

The subcutis is a non-echogenic structure and, therefore, can be differentiated from the dermal layer. However, on truncal and facial skin the interface between dermis and the subcutaneous space is not sharply defined. Visualisation of the dermal-subcutaneous boundary can be improved by the ultrasound gain adjustment.

Skin tumours on sonographic image

Skin tumours can be visualised using both two and three dimensional ultrasound imaging [9, 25, 34, 35]. They are almost exclusively characterised by low echogenic areas being more or less separated from the dermis, and while sound effects like subtumoral attenuation or extinction of the signal may be useful indicators, they do not allow definite tumour diagnosis based on sonographic image. Malignant melanomas have been extensively studied by this technique [35-39]. Melanomas are characterized by their spindle-shaped appearance being clearly separated from the surrounding tissue. There is a significant correlation between the preoperative, sonographically assessed thickness of the tumour and the one assessed by postoperative, histological morphometry offering some guidance in the preoperative assessment [35, 37-39]. A still unsolved problem, however, is the difficulty of distinguishing between tumour parenchyma and subtumoral inflammatory infiltrate, or nevocytic nevi. Consequently, some tumours are sonographically overestimated with a precise measurement of the tumour thickness sometimes being impossible [36, 39]. Recently, Ulrich et al. have shown that the correlation between sonometry and histometry increases appreciably if tumours with significant inflammatory infiltrate or nevus-associated melanomas were excluded from the analysis [37].

Upon sonography, melanocytic nevi present as spindle-shaped structures with low echogenic material, thus being hard to distinguish from malignant melanomas. Current technology is therefore better suited to quantification of pigmented skin lesion size than diagnosis.

Although the preoperative need for differential diagnosis is less pressing in non-melanoma skin cancer, basal cell carcinomas have also been studied with high frequency ultrasound. Depending on size and histological type they generally present as bizarely formed low echogenic foci often causing an enhancement of dorsal echoes [40].

In contrast, tumours with marked hyperkeratosis such as angiokeratomas or seborrhoic keratosis present with the phenomenon of attenuation or total extinction of reflected echoes. In the sonographic image they are characterised by a sound shadow and can, therefore, easily be distinguished from other tumors.

CONCLUSION

High frequency two- and three-dimensional ultrasound have been developed from simple unidimensional methods to give better qualitative and semi-quantitative information about the structures of the skin. Two-dimensional ultrasonography has become an established method for clinical and experimental dermatology. Dermatosonography offers an important possibility to assess skin morphology at an intermediate level between clinical assessment and microscopy; it offers the possibility of both non-invasive and in vivo studies, and subsequently the possibility of chronological dynamic recording of morphological events in the skin. These characteristics make it a unique tool for the examination of the skin. In contrast, the three-dimensional technique essentially remains experimental at the moment, although showing some promising results, in particular in the assessment of skin tumours. With continued technical development and refined resolution, three dimensional ultrasound imaging may also become a useful clinical and experimental tool in dermatology.

Acknowledgments

Original work cited in this review was supported by grants from Gerda og Aage Haenschs Foundation, Denmark; Leo Pharmaceutical Products Foundation, Sigvaris Ganzoni Foundation, Switzerland, and Deutsche Forschungsgemeinschaft (INK 15/A1).

Article accepted on 23/3/00

REFERENCES

1. Dreijer N. Diagnostic ultrasound. Naerum: Brüel and Kjær, 1979.

2. Alexander H, Miller DL. Determining skin thickness with pulsed ultrasound. J Invest Dermatol 1979; 72: 17-9.

3. Hoffmann K, Dirting K, Stücker M, el-Gammal S, Wilmert M, Altmeyer P. Geschichte der hochfrequenten Sonographie. Ultraschall Med 1994; 15: 192-7.

4. Serup J, Keiding J, Fullerton A, Gniadecka M, Gniadecki R. High-frequency ultrasound examination of skin: introduction and guide. In: Serup J, Jemec GBE, eds. Handbook of non-invasive methods and the skin. Boca Raton: CRC Press, 1995: 239-56.

5. Seidenari S. Ultrasound B-mode imaging and in vivo structure analysis. In: Serup J, Jemec GBE, eds. Handbook of non-invasive methods and the skin. Boca Raton: CRC Press, 1995: 257-68.

6. El-Gammal S, Altmeyer P, Auer T, Kaspar K, Hoffmann K, Paßmann C, Ermert H. Der Stellenwert der 20, 50 und 100 MHz Sonographie in der Dermatologie. Akt Dermatol 1995; 21: 11-21.

7. Gniadecka M, Jemec GBE. Quantitative evaluation of chronological aging and photoaging in vivo: studies on skin echogenicity and thickness. Br J Dermatol 1998; 139: 815-21.

8. Jemec G, Gniadecka M. Ultrasound examination of hair follicles in hidradenitis. Arch Dermatol 1997; 133: 967-72.

9. Harland CC, Bamber JC, Guesterson BA, Mortimer PS. High frequency, high resolution ultrasound in the assessment of skin tumours. Br J Dermatol 1993; 128: 525-32.

10. Gniadecka M, Quistorff B. Assessment of dermal water by high-frequency ultrasound: comparative studies with nuclear magnetic resonance. Br J Dermatol 1996; 135: 218-24.

11. Gniadecka M. Dermal oedema in lipodermatosclerosis: distribution, effects of posture and compressive therapy evaluated by high-frequency ultrasonography. Acta Derm Venereol 1995; 75: 120-4.

12. Gniadecka M. Localization of dermal oedema in lipodermatosclerosis, lymphedema and cardiac insufficiency. J Am Acad Derm 1996; 35: 37-41.

13. Serup J, Staberg B. Ultrasound for assessment of allergic and irritant patch test reactions. Contact Dermatitis 1987; 17: 80-4.

14. Agner T. Non invasive measuring methods for the investigation of irritant patch test reactions. Thesis. Acta Dermato-Venereol (Stockh) 1992; 173 (suppl.): 1-26.

15. Seidenari S, Di Nardo A, Pepe P, Gianetti A. Ultrasound B scanning with image analysis for assessment of allergic patch test reactions. Contact Dermatitis 1991; 24: 216-22.

16. Serup J. Localized scleroderma (Morphea): thickness of sclerotic plaques as measured by 15 MHz pulsed ultrasound. Acta Dermato-Venereol (Stockh) 1984; 64: 214-9.

17. Levy JJ, Gaßmüller J, Audring H, Brenke A, Albrecht-Nebe H. Darstellung der subkutanen Atrophie bei der zirkumscripten Sklerodermie im 20-MHz-B-scan Ultraschall. Hautarzt 1993; 44: 446-51.

18. Serup J. Non-invasive quantification of psoriasis plaques measurement of skin thickness with 15-MHz pulsed ultrasound. Clin Exp Dermatol 1984; 9: 502-8.

19. Gupta AK, Turnbull DH, Harasiewicz KA, Shum DT, Watteel GN, Foster FS, Sauder DN. The use of high-frequency ultrasound as a method of assessing the severity of a plaque of psoriasis. Arch Dermatol 1996; 132: 658-62.

20. Hoffmann K, Dirschka T, Schwarze H, el-Gammal S, Matthes U, Hoffmann A, Altmeyer P. 20 MHz sonography, colorimetry and image analysis in the evaluation of psoriasis vulgaris. J Derm Sci 1995; 9: 103-10.

21. Hoffmann K, Winkler K, el-Gammal S, Altmeyer P. A wound healing model with sonographic monitoring. Clin Exp Dermatol 1993; 18: 217-25.

22. Richard S, de Rigal J, de Lacharriere O, Berardesca E, Leveque JL. Noninvasive measurement of the effects of lifetime exposure to the sun on the aged skin. Photodermatol Photoimmunol Photomed 1994; 10: 164-9.

23. Gniadecka M, Gniadecki R, Serup J, Sondergaard J. Ultrasound structure and digital image analysis of the subepidermal low echogenic band in aged human skin: diurnal changes and interindividual variability. J Invest Dermatol 1994; 102: 362-5.

24. Hoffmann K, Dirschka TP, Stücker M, el-Gammal S, Altmeyer P. Assessment of actinic skin damage by 20-MHz sonography. Photoderm Photoimmunol Photomed 1994; 10: 97-101.

25. El-Gammal S, Auer T, Hoffman K, Altmeyer P, Passmann C, Ermert H. High-resolution ultrasound examination of the epidermis. In: Serup J, Jemec GBE, eds. Handbook of non-invasive methods and the skin. Boca Raton: CRC Press, 1995: 125-32.

26. Serup J, Keiding J, Fullerton A, Gniadecka M, Gniadecki R. High-frequency ultrasound examination of the skin: introduction and guide. In: Serup J, Jemec GBE, eds. Handbook of non-invasive methods and the skin. Boca Raton: CRC Press, 1995: 239-56.

27. Altmeyer P, Hoffman K, Stücker M. General phenomena of ultrasound in dermatology. In: Altmeyer P, el-Gammal S, Hoffmann K, eds. Ultrasound in dermatology. Berlin: Springer, 1992: 55-79.

28. Gniadecka M, Serup J, Sondergaard J. Age-related changes of skin oedema by high-frequnecy ultrasound. Br J Dermatol 1995; 131: 849-55.

29. Korting HC, Vieluf D, Kerscher M. 0.25% prednicarbat cream and the corresponding vehicle induce less skin atrophy than 0.1% betamethason-17-valerate cream and 0.05% clobetasol-17-propionate cream. Eur J Clin Pharmacol 1992; 42: 159-61.

30. Leveque JL. Noninvasive measurements on photodamaged skin. In: Gilchrest BA, ed. Photodamage. Oxford: Blackwell Scientific Publications, 1995: 192-200.

31. Marks R, Hill S, Barton SP. The effects of an abrasive agent on normal skin on photoaged skin in comparison with topical tretinoin. Br J Dermatol 1990; 123: 457-66.

32. Jemec GBE, Serup J. The ultrasound structure of the normal human nail. Arch Derm 1989; 125: 643-6.

33. Jemec GBE, Gniadecka M. In vivo studies of normal and psoriatic finger nails using high frequency ultrasound. Turk J Derm Pathol 2000; in press.

34. El-Gammal S, Hoffmann K, Kenkmann J, Altmeyer P, Hoss A, Ermert H. Principles of three dimensional reconstructions from high resolution ultrasound in dermatology. In: Altmeyer P, el-Gammal S, Hoffmann K, eds. Ultrasound in dermatology. Berlin: Springer, 1992: 355-84.

35. Hoffmann K, Jung J, el-Gammal S, Altmeyer P. Malignant melanoma in 20-MHz B scan sonography. Dermatology 1992; 1992: 49-55.

36. Salmhofer W, Rieger E, Soyer HP, Smolle J, Kerl H. Influence of skin tension and formalin fixation on sonographic measurement of tumour thickness. J Am Acad Dermatol 1996; 34: 34-9.

37. Ulrich J, Petereit S, Gollnick H. Präoperative sonographische Melanomdiagnostik ­ Vergleich von 7.5 und 20-MHz-Sonographie. Ultraschall Med 1999; 20: 197-200.

38. Hoffmann K, Happe M, Schüller S, Stücker M, Wiesner M, Gottlöber P, Schwarz M, Strahler J, Neubauer N, Jung C, Petereit S, Welzel J, Brautzsch N, Bohmeyer J, Wohlrab J, Freitag M, Altmeyer P. Stellenwert der 20-MHz-Sonographie des malignen Melanoms und pigmentierter Läsionen in der Routinediagnostik. Ultraschall Med 1999; 20: 104-9.

39. Tacke J, Haagen G, Hornstein OP, Huettinger G, Kiesewetter F, Schell H, Diepgen TL. Clinical relevance of sonometry-derived tumour thickness in malignant melanoma ­ a statistical analysis. Br J Dermatol 1995; 132: 209-14.

40. Hoffmann K, Stücker M, el-Gammal S, Altmeyer P. Digitale 20-MHz-Sonographie des Basalioms im b-scan. Hautarzt 1990; 41: 333-9.