The presence of uniform and regularly distributed reticular, globular, starburst, and homogeneous blue patterns identifies a given lesion as a melanocytic nevus (i.e., per definition absence of melanoma-specific patterns) [4]. Each pattern corresponds to a specific underlying histopathologic correlate.

Based on the most common dermoscopic patterns associated with melanocytic nevi, a new classification has been recently proposed that includes four main categories: globular, reticular, starburst, and homogenous blue nevi [22, 23]. In this classification system, small congenital nevi, compound nevi, and dermal nevi are lumped together in the globular category based on their common dermoscopic-histopathologic features (globules correspond to predominantly dermal nests of melanocytes). The reticular pattern typically corresponds to junctional or lentiginous nevi, with the exception of congenital nevi in the lower extremities [24, 25]. The starburst category includes both pigmented Spitz nevi and Reed nevi based on their striking dermoscopic features [26] (regular peripheral streaks). Homogenous structureless blue pigmentation without additional dermoscopic features is the pattern commonly observed in blue nevi [17]. Nevi with a homogeneous structureless brown pattern are not considered within the group of blue nevi because they reveal histopathologic correlates similar to reticular nevi [27, 28].

Based on the work of Kittler et al. [29, 30] in digital follow-up, growing nevi can be easily recognized by their striking dermoscopic hallmark of a peripheral rim of small brown globules frequent in adolescents and young adults (Fig. 13.2). The same pattern in individuals older than 60 years should raise suspicion and preferably be excised and studied to rule out malignancy.

Spitz and Reed nevi reveal different patterns, depending on the growth phase of the lesion [31–33]. After an initial globular pattern, Spitz and Reed nevi tend to show the classic starburst pattern, which represents an intermediate pattern of their evolution, showing finally a reticular or homogeneous pattern at the end. Since no single criterion allows differentiating such spitzoid-appearing melanomas from Spitz or Reed nevi with sufficient accuracy, excision of all spitzoid lesions, in adults, is always recommended [34, 35].

Conversely, homogeneous blue nevi seem to be highly stable lesions. This stability of blue nevi is an important, although subjective, clue for the diagnosis because blue color alone is a highly unspecific feature that may occur also in nodular melanoma, melanoma metastases, or pigmented basal cell carcinoma (4–17–23). Therefore, the diagnosis of blue nevi should always be based on the combination of the dermoscopic pattern and a convincing subjective history of no changes. When a reliable history is difficult to obtain, excision, not monitoring, must be performed.

On the contrary, melanoma tends to show multiple colors and structures. A given melanocytic lesion with multiple colors, asymmetry, and multicomponent pattern (three or more structures) is highly suggestive of melanoma. For those lesions that we do not recognize at a first glance as melanoma with global pattern analyses, various dermoscopic algorithms may be applied to differentiate nevi from melanoma [4, 36–38]. Pattern analyses consider the presence of some structures associated to melanoma as additional ten clues for the diagnosis of melanoma: the presence of atypical pigment network, atypical dots and globules, regression structures, crystalline structures, negative pigment network, irregular blotches, irregular pseudopods or streaks, blue-white veil, milky red areas, or polymorphous/dotted vessels (Figs. 13.3, 13.4, 13.5, 13.6, and 13.7).

According to the results of the Consensus Net Meeting on Dermoscopy, the sensitivity for melanoma of these algorithms (pattern analysis, the ABCD rule, 7-point checklist, or Menzies scoring method) was similar, but pattern analysis demonstrated a better specificity [4].

However, dermoscopy is not 100 % accurate, and a certain percentage of suspicious but benign lesions have to be excised in order not miss melanoma. It is worth highlighting again that malignant cutaneous melanoma may sometimes mimic benign melanocytic and non-melanocytic lesions. For this reason, it is important that a good clinical-dermoscopic correlation exists, and the presence of conflicting clinical- dermoscopic features should lead the clinician to perform a biopsy.

On the contrary of what happen in other body sites, the application of dermoscopy in facial lesions increases the number of biopsies and in consequence the number of early melanomas identified.

Lesions on the face show specific dermoscopic patterns due to the anatomy of the region, being the classical pattern of melanocytic lesions the pseudo-network.

Early lentigo maligna melanomas (LMMs) may show annular granular pattern (with gray dots around hair follicles), irregular perifollicular pigmentation, short lines, and rhomboidal structures. When lentigo maligna progresses, occlusion of follicular openings and specific criteria of melanoma as blue whitish veil may appear [39].

The histopathological correlation of each dermoscopy criteria of LMM is also well known: the annular granular pattern corresponds to melanophages in papillary dermis and around hair follicles; perifollicular pigmentation corresponds to the proliferation/invasion of atypical melanocytes in the hair follicle; short lines and rhomboidal structures correspond to the proliferation of atypical melanocytes in the dermoepidermal junction; and finally, the occlusion of follicular openings corresponds to the progression of the melanoma and complete invasion of the hair follicle.

Multiple gray dots (granularity) may also be seen in pigmented actinic keratosis and in liken planus-like keratosis (LPLK). Sometimes, in LPLK the granularity is coarser, but in other occasions the diagnosis is just performed after a biopsy. A difficult differential diagnosis with dermoscopy is between early LMM and pigmented actinic keratosis because both may show multiple gray dots, and again the diagnosis is performed by histopathology. Dermoscopy is useful not only to increase the index of suspicion in lesions on the face but also to select the best area to be biopsied to obtain a diagnosis.

For the diagnosis of melanoma in acral sites, the recognition of parallel ridge pattern [40] (Fig. 13.8) is highly specific, while benign patterns are also well characterized being the more frequent benign pattern the parallel furrow pattern [41]. Dermoscopy is very useful in the recognition of benign patterns and permits to avoid unnecessary excision.

In clinical practice, physicians are dealing with patients with multiple lesions and not with isolated lesions. In these situations comparative recognition process is used. Most individuals have a predominant lesion’s pattern; therefore, the examination of all lesions is an essential step in the recognition process because it allows the identification of lesions deviating from the individual’s prevailing benign pattern (concept of the “ugly duckling sign”) [42–45] (Fig. 13.9).

The importance of surveillance by digital dermoscopy follow-up programs to identify melanoma at an early stage in high-risk populations such as patients with hundreds of nevi, atypical mole syndrome (AMS), personal or family history of melanoma, and carriers of genetic mutations as in CDKN2A has been clearly demonstrated [14, 46] (Fig. 13.10). In a recent study comparing melanomas identified in patients included in digital dermoscopy follow up (DFU) program with melanomas diagnosed in patients referred to a melanoma unit, the multicomponent pattern (defined by the combination of three or more distinctive dermoscopic structures within a given lesion) was noted in 60 % of referred melanomas but in only 16 % of those identified in the DFU surveillance program. By contrast, the reticular pattern was the most frequent pattern observed in melanomas that appeared in patients included in the digital dermoscopic follow-up program [46]. This reticular pattern is the characteristic pattern of slow-growing melanomas described by Argenziano et al. [47] and that may be slow growing in a radial phase for years or decades being diagnosed when this pigment network is altered and a homogeneous structure appears [48] associated with a disruption of the dermoepidermal architecture or the occurrence of regression. The inclusion of high-risk population in follow-up programs allows the detection of melanomas in early stages, with good prognosis (mainly in situ, if not AJCC stage IA), even in the absence of clinical and dermoscopic features of melanoma. In a selected study population, with 90 % of the patients displaying AMS and almost 45 % having had a previous melanoma, one of eight developed melanoma during surveillance, which is more than 1,500 times higher than expected in the general population. According to these results, MM can be diagnosed at any time once a high-risk patient is included in the DFU program. Thus, maintained surveillance is required in individuals at high risk, and the “two-step method of digital follow-up” allows detection of melanomas with few dermatoscopic criteria by digital dermoscopy follow-up (60 % of the new melanomas diagnosed in DFU) and also the detection of melanoma that present as a new lesion or arising from nevi not monitored by dermatoscopy (40 % of melanomas detected in digital follow-up) [14].