Dermatoscopic Images Research Articles

Application of Teledermoscopy in the Diagnosis of Pigmented Lesions.

Background Dermatology, due to the peculiar characteristic of visual diagnosis, is suitable for the application of modern telemedicine techniques, such as mobile teledermoscopy. Objectives To evaluate the feasibility and reliability of the technique for the diagnosis of pigmented lesions. Methods Through the storage and routing method, 41 pigmented lesions were analyzed. After the selection of the lesions during the outpatient visit, the clinical and dermatoscopic images were obtained by the resident physician through the cellphone camera and sent to the assistant dermatologist by means of an application for exchange of messages between mobile platforms. Firstly, the assistant dermatologist described the visualized dermatoscopic structures and defined its diagnosis and conduct, based solely on the evaluation of the clinical and dermatoscopic images, without having the knowledge of the anamnesis data. Afterwards, the same assistant dermatologist evaluated the patient face to face, defining the dermatoscopic structures, diagnosis, and conduct. The data obtained through teledermoscopy and face-to-face assessments were compared and accuracy was defined as the concordance between the diagnoses. Results A match rate of 90% between teledermoscopic and face-to-face diagnosis was demonstrated (McNemar's statistical analysis, whose p value was 0.1366, showed no evidence to support the inferiority of the teledermoscopic method).

Multimodal skin lesion classification using deep learning.

While convolutional neural networks (CNNs) have successfully been applied for skin lesion classification, previous studies have generally considered only a single clinical/macroscopic image and output a binary decision. In this work, we have presented a method which combines multiple imaging modalities together with patient metadata to improve the performance of automated skin lesion diagnosis. We evaluated our method on a binary classification task for comparison with previous studies as well as a five class classification task representative of a real-world clinical scenario. We showed that our multimodal classifier outperforms a baseline classifier that only uses a single macroscopic image in both binary melanoma detection (AUC 0.866 vs 0.784) and in multiclass classification (mAP 0.729 vs 0.598). In addition, we have quantitatively showed the automated diagnosis of skin lesions using dermatoscopic images obtains a higher performance when compared to using macroscopic images. We performed experiments on a new data set of 2917 cases where each case contains a dermatoscopic image, macroscopic image and patient metadata.

Automated skin melanoma diagnostics based on mathematical model of artificial convolutional neural network

In the last 10 years there has been a revolu on in the fi eld of computer image analysis and pa ern recogni on. Modern algorithms of computer vision equaled and even in some problems surpassed human capabili es. This jerk is largely due to the emergence and development of the technology of deep convolu onal neural networks.Recent developments in the fi eld of image processing and machine learning open up the prospect of crea ng systems based on ar fi cial neural convolu onal networks, superior to humans in problems of image classifi ca on, in par cular, in solving problems of analysis of various medical images. Among the most promising applica ons: automated recogni on and classifi ca on of skin diseases, detec on of pathologies on X-ray, CT, MRI, ultrasound imaging. In the proposed project, we will focusour a en on on the diagnosis of human skin diseases.At the moment, melanoma is one of the most dangerous types of malignant tumors of the skin with a lot of deaths due to rapid metastasis, which is difficult to treat. The development of computer vision technology has allowed the development of technical vision systems that allow detec on and classifi ca on of skin diseases with a quality that is comparable and in some cases exceeds the values a ained by man.In this paper, the authors propose an algorithm for the primary diagnosis of skin melanoma based on deep neural networks, achieving an accuracy of 91% for melanoma in dermatoscopic images. At the moment, the algorithm is implemented programma cally and is used in the test version of the online system for detec ng and classifying skin diseases, available at skincheckup.online.Thanks to this development, the prospect of a signifi cantincrease in the propor on of people subjected to preven ve examina on for the presence of skin diseases opens up. Along with this, an addi onal source of informa on for specialized professionals can also play a role in seng the right diagnosis.

Does Dermatoscopy Facilitate the Detection and Diagnosis of Vascular Skin Lesions? A Case–Control Study

During dermatoscope-guided surgical procedures, we noticed that vasculature was easily identified. This study investigated the use of dermatoscopy in detecting and diagnosing vascular skin lesions. We retrieved records of patients with vascular skin lesions who underwent dermatoscopy over a 3month period, in two outpatient clinics affiliated with a university teaching hospital. Our controls were similar patients where dermatoscopy was not performed. Our new findings are: 1, clinical and dermatoscopic examinations diagnosed significantly more patients with vascular skin lesions than clinical examinations alone (risk ratio: 1.36; 95% confidence interval: 1.10-1.67); 2, the detection rate increase was significant for cherry angiomas (p<0.001), telangiectasias (p<0.01) and spider angiomas (p<0.01); 3, qualitatively, dermatoscopy revealed characteristic configurations, hues and colour saturations of the vascular skin lesions; and 4, the first reported dermatoscopic images of focal essential telangiectasia and petechial angioma. In our setting, clinical and dermatoscopic examinations significantly facilitated detecting and diagnosing vascular skin lesions, compared to clinical examination alone.

A prospective diagnostic study on povidone-iodine retention in lesions suspected to be squamous cell carcinoma or keratoacanthoma.

While dermatoscopy improves diagnostic accuracy for raised nonpigmented lesions, those with white surface keratin can be problematical. We investigated whether retention of povidone-iodine by surface keratin provides a clue to benignity. We performed a retrospective pilot study (n=57) followed by a prospective study (n=117) on raised nonpigmented lesions with white surface keratin. An initial dermatoscopic image was taken of each lesion, povidone-iodine was applied and another image taken. Following lavage with 70% ethanol, a third image was acquired. The percentage surface area of residual povidone-iodine staining after lavage was recorded, and the results analysed. The optimal cut-off point of residual staining was 80%, where values of ≤80% pointed to malignancy. At this cut-off, the OR for lesions with values ≤80% to be truly malignant in the retrospective set was 4.03 (95% CI: 2.1-7.6) and the AUC was 0.7 (95% CI: 0.62-0.78). For the prospective set, the corresponding OR was 2.3 (95% CI: 1.4-3.7) and the AUC was 0.62 (95% CI: 0.55-0.68). This study presents evidence that povidone-iodine retention may have a degree of efficacy in distinguishing benign from malignant keratotic lesions. Further study is warranted.

The HAM10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions

Training of neural networks for automated diagnosis of pigmented skin lesions is hampered by the small size and lack of diversity of available datasets of dermatoscopic images. We tackle this problem by releasing the HAM10000 (“Human Against Machine with 10000 training images”) dataset. We collected dermatoscopic images from different populations acquired and stored by different modalities. Given this diversity we had to apply different acquisition and cleaning methods and developed semi-automatic workflows utilizing specifically trained neural networks. The final dataset consists of 10015 dermatoscopic images which are released as a training set for academic machine learning purposes and are publicly available through the ISIC archive. This benchmark dataset can be used for machine learning and for comparisons with human experts. Cases include a representative collection of all important diagnostic categories in the realm of pigmented lesions. More than 50% of lesions have been confirmed by pathology, while the ground truth for the rest of the cases was either follow-up, expert consensus, or confirmation by in-vivo confocal microscopy.

Sequential digital dermatoscopic imaging of patients with multiple atypical nevi.

Patients with multiple atypical nevi are at higher risk of developing melanoma. Among different techniques, sequential digital dermatoscopic imaging (SDDI) is a state-of-the art method to enhance diagnostic accuracy in evaluating pigmented skin lesions. It relies on analyzing digital dermatoscopic images of a lesion over time to find specific dynamic criteria inferring biologic behavior. SDDI can reduce the number of necessary excisions and finds melanomas in an early—and potentially curable—stage, but precautions in selecting patients and lesions have to be met to reach those goals.

The diagnostic accuracy of the mobile phone teledermatoscopy

The positive predictive value (PPV) of smart mobile phone teledermatoscopy is not known. The main purpose of the present study was to investigate the sensitivity and positive predictive values (PPVs) of smart mobile phone teledermatoscopy. Over a period of 6 months, up to three clinical and dermatoscopic images were obtained of 67 skin lesions from 67 patients using a mobile phone camera and standard pocket dermatoscopy device. Out of the 67 patients, 44 were men (65.67%) and 23 were women (34.32%). The mean age of the patients was 39.56 ± 22.19 years (ranging from 18 to 92). The majority of the lesions (71.64%; n=48) were benign, while 11.94% (n=8) of the biopsies were premalignant and 16.41% (n=11) of the lesions were malignant. The sensitivity for the diagnosis of benign, premalignant, and malignant lesions were 93.8%, 100%, and 100%, respectively. PPVs for the diagnosis of benign, premalignant, and malignant lesions were 93.8%, 100%, and 100%, respectively. The sensitivity and PPVs of all lesions were 95.9% and 95.7%. The accuracy of the teledermatoscopic consultation with a mobile phone is very high. We therefore think that it can be a cost effective and useful method in the consultation at distance.

Dermatoscopic-pathological correlation of melanocytic skin lesions

There is no doubt that dermatopathology is the most important method to decide if amelanocytic lesion is benign or malignant; however, like most morphologic examinations, dermatopathology is subjective. Arecent study demonstrated that the pathologic diagnosis of melanocytic skin lesions has ahigh variability. Reports with false-positive or false-negative diagnoses are relatively common. The pathologic examination of melanocytic lesions also has observer-independent limitations and one has to accept that some melanocytic lesions cannot be classified as benign or malignant with confidence by dermatopathology alone. If aconfident diagnosis is not possible adermatoscopic-pathologic correlation may be helpful. This, however, is only possible if dermatoscopic images are available and if the dermatopathologist knows how to interpret dermatoscopic structures. Adermatoscopic-pathologic correlation is not useful in all difficult melanocytic lesions but it should be considered in difficult flat pigmented lesions. In these cases dermatoscopy may provide even more important additional information than molecular findings.

Morphologic characteristics of nevi associated with melanoma: a clinical, dermatoscopic and histopathologic analysis.

BackgroundThe aim of this retrospective study was to determine the frequency of nevus-associated melanomas and to better characterize the preexisting nevus from a histopathologic, clinical and dermatoscopic point of view.MethodsWe reviewed the histopathologic slides of a consecutive series of 357 melanomas and corresponding clinical and dermatoscopic images, if available.ResultsWe found that 31 (8.7%) melanomas were associated with a preexisting nevus, 284 (79.5%) melanomas developed de novo, and in 42 (11.8%) a preexisting nevus could not be excluded, although the alternative explanation that the entire lesion represented a melanoma was also possible. The preexisting nevus was a “superficial” or “superficial and deep” congenital nevus in 27 cases (87%) and a Clark nevus in 4 cases (13%). Clinical or dermatoscopic images were available in 149 (41.7%) cases. The preexisting nevus, if visible, looked inconspicuous clinically or dermatoscopically. The median invasion thickness of nevus-associated melanoma was not significantly different from “de novo” melanomas but the frequency of in situ melanomas was higher in the “de novo“ group (40.1% versus 16.1%). Patients with melanoma in association with a nevus were significantly younger (mean age=55 years, SD: 16 years) than patients with “de novo” melanomas (mean age=68 years SD: 15 years, p<0.001). When controlled for age and invasion thickness overall, survival did not differ significantly between patients with nevus-associated melanomas and patients with de novo melanomas.ConclusionsFrom a histomorphologic point of view, the majority of melanomas arise de novo. If melanomas develop in a preexisting nevus, they usually occur in association with a “superficial” or “superficial and deep” congenital nevus.

Machine learning-based diagnosis of melanoma using macro images.

Cancer bears a poisoning threat to human society. Melanoma, the skin cancer, originates from skin layers and penetrates deep into subcutaneous layers. There exists an extensive research in melanoma diagnosis using dermatoscopic images captured through a dermatoscope. While designing a diagnostic model for general handheld imaging systems is an emerging trend, this article proposes a computer-aided decision support system for macro images captured by a general-purpose camera. General imaging conditions are adversely affected by nonuniform illumination, which further affects the extraction of relevant information. To mitigate it, we process an image to define a smooth illumination surface using the multistage illumination compensation approach, and the infected region is extracted using the proposed multimode segmentation method. The lesion information is numerated as a feature set comprising geometry, photometry, border series, and texture measures. The redundancy in feature set is reduced using information theory methods, and a classification boundary is modeled to distinguish benign and malignant samples using support vector machine, random forest, neural network, and fast discriminative mixed-membership-based naive Bayesian classifiers. Moreover, the experimental outcome is supported by hypothesis testing and boxplot representation for classification losses. The simulation results prove the significance of the proposed model that shows an improved performance as compared with competing arts.

Onychoscopy in the diagnosis of the distal subungual onychomycosis and traumatic onycholysis

Aim: Distal subungual onychomycosis is a common public health problem which composes a large part of nail diseases. The exact diagnosis of the disease is often unfeasible without mycological investigations. In this study, we aimed to reveal sensitivity and specificity of previously described onychoscopic (nail dermoscopic) features that support and facilitate diagnosis of distal subungual onychomycosis. Material and Methods: Big toe nail of 53 patients who present with distal onycholysis were examined by digital dermatoscopy and were photographed. Specific dermatoscopic features were recorded. Mycological investigations were performed in all patients and dermatoscopic diagnosis was compared with mycological diagnosis. Threepatients were excluded from the study whose final diagnoses were psoriasis. Results: Two specific findings previously described as longitudinal stria and spike were obtained as a result of examination of the dermatoscopic imaging. Sensitivity and specificity of longitu-dinal stria sign was found as 75.7% and 88.2% respectively. Sensitivity and specificity of spike sign was found as 57.5% and 94.1% respectively. And finally sensitivity and specificity of the presence of at least one of the two signs was found as 96,9% and 88,2%, respectively. Conclusion: Dermoscopic examination is very useful and cost-effective method in diagnosis of the distal subungual onychomycosis

Application of automatic statistical post-processing method for analysis of ultrasonic and digital dermatoscopy images

ABSTRACTUltrasonic and digital dermatoscopy diagnostic methods are used in order to estimate the changes of structure, as well as to non-invasively measure the changes of parameters of lesions of human tissue. These days, it is very actual to perform the quantitative analysis of medical data, which allows to achieve the reliable early-stage diagnosis of lesions and help to save more lives. The proposed automatic statistical post-processing method based on integration of ultrasonic and digital dermatoscopy measurements is intended to estimate the parameters of malignant tumours, measure spatial dimensions (e.g. thickness) and shape, and perform faster diagnostics by increasing the accuracy of tumours differentiation. It leads to optimization of time-consuming analysis procedures of medical images and could be used as a reliable decision support tool in the field of dermatology.

Dermatoscopic Patterns of Nonfacial Actinic Keratosis: Characterization of Pigmented and Nonpigmented Lesions.

Actinic keratoses (AKs) are dysplastic proliferations of keratinocytes. Studies evaluating nonfacial dermatoscopic pattern of AKs are scarce. This study aimed to evaluate the dermatoscopic patterns of AKs located in nonfacial sites and to compare their patterns with facial lesions. Patients with concomitant facial and nonfacial AKs were recruited to participate and evaluated by clinical and dermatoscopic images of their AKs. Sixty-eight patients were included in the study. A total of 258 nonfacial AKs and 68 facial AKs were analyzed. The most frequent nonfacial AK dermatoscopic structures were white opaque scales (97.3%) and erythema (57.4%). When analyzed in combination, white scales plus erythema were found in 55.4% of nonfacial AKs. Pigmented structures were observed in 22.5% nonfacial AKs. Homogeneous brown pigmentation was the most prevalent pigmented structure in nonfacial pigmented AK (pAK) (93.1%). There was a positive association between patients having concomitant pigmented facial and nonfacial AKs (p < .001). The combinations of erythema and white opaque scales or yellow opaque scales and homogeneous pigmentation are suggestive, respectively of nonpigmented and pigmented nonfacial AKs. Pigmented AKs occur concomitantly in facial and nonfacial areas.

Dermatoscopic evaluation and histopathological correlation of acquired dermal macular hyperpigmentation.

Acquired dermal macular hyperpigmentation (ADMH) is a hypernym encompassing Riehl's melanosis, lichen planus pigmentosus, and ashy dermatoses that show significant clinicopathological overlap. We sought to describe the dermatoscopic features of ADMH and correlate them with histopathological findings. This was a prospective observational study performed in two phases. A detailed clinical and dermatoscopic examination was performed, and skin biopsies were obtained in 51 patients. Two dermatologists and a blinded dermatopathologist studied archived dermatoscopic images and histopathology specimens, respectively. Dermatoscopic features noted were (i) pigment structures; dots (82.4%), globules (66.7%) and blotches (56.9%) that spared the eccrine and hair follicle openings; (ii) telangiectasia (82.4%); (iii) accentuation of the normal pseudoreticular pigmentary network (33.3%); (iv) owl's eye structures (15.7%). Four dermatoscopic grades of disease severity were identified: grade 1 - dotted; grade 2 - Chinese letter; grade 3 - reticulate; and grade 4 - diffuse. Density of melanin incontinence on histopathology correlated positively with size of pigment structures (r=0.7, P<0.000) and grades of disease severity (r=0.75, P<0.000) on dermatoscopy. Increasing grades of disease severity can be detected dermatoscopically, which correlate well with histopathological features. A carefully performed dermatoscopy aids in better patient counseling regarding disease severity.

Temporal Image Comparison (Serial Imaging) in Assessing Pigmented Lesions.

The large-scale deployment of low-cost, noninvasive mechanisms of early detection are needed to reduce the melanoma burden. A serial 2-step system could power mass screening efforts serving the uninsured and underinsured as well as the rural and remote US counties where melanoma mortality is doubled for lack of access to dermatologists. Furthermore, serial melanoma screening strategies, serial total body photography and serial digital dermatoscopy imaging may be performed as a telehealth service, and thus would be available in any location that can support activity compliant with the Health Insurance Portability and Accountability Act and has appropriate bandwidth.

Long-term evaluation of the efficacy of digital dermatoscopy monitoring at a tertiary referral center.

We examined the value of monitoring patients with multiple nevi using sequential digital dermatoscopy imaging at a tertiary referral center. This is a retrospective cohort study including 2,824 patients and 23,241 monitored lesions. We calculated trends in key parameters such as the number of melanomas and nevi monitored and excised. During follow-up, we excised 1,266 lesions in 709 patients, including 146 (11.5%) melanomas. The percentage of in situ melanomas detected at follow-up was significantly higher than at baseline (46.6% versus 23.4%, p≤0.001). The risk of detecting a melanoma during follow-up was higher for patients with a melanoma at baseline, compared to those without (relative risk: 3.59, 95% CI: 2.15 to 6.00). The number of documented lesions showed a positive correlation with the benign/malignant ratio, and excisions at baseline decreased significantly over the course of the study period. Digital dermatoscopy monitoring improves the detection of thin melanomas in patients with multiple nevi. Patients with a melanoma at baseline are at an increased risk of developing a melanoma during follow-up and should therefore be a target group for sequential dermatoscopy imaging.

The significance of blue color in dermatoscopy.

Skin lesions with blue color are frequently excised to rule out malignancy. The objective of the present study was to investigate the significance of blue color. We retrospectively scanned dermatoscopic images for blue color and classified them according to pattern analysis. Of 1,123 pigmented skin lesions, 144 (12.8%) showed blue color, 92 of which (63.9%) were malignant. Among lesions with blue color, the most common benign diagnoses were nevi (n=35, 24.3%) and seborrheic keratoses (n=8, 5.6%). Of 103 (71.5%) lesions with a structureless blue pattern, eight (7.8%) were entirely blue and 95 (92.2%) were partly blue, of which 81 (78.6%) showed peripheral or patchy and 14 (13.6%) central blue color. Most lesions with peripheral or patchy blue color were melanomas (n=47, 58%), whereas most lesions with central blue color were nevi (n=9, 64.3%). Of 28 lesions with blue clods, 17 (60.7%) were basal cell carcinomas. With respect to malignancy, the positive predictive value of blue color was 63.9% (95% CI: 56.0-71.8%). Among malignant lesions with blue color, structureless peripheral or patchy blue color is a clue for melanoma, while blue clods point to basal cell carcinoma. Pitfalls include seborrheic keratoses, which may show blue color, as well as some nevi, especially combined nevi.

"Twin lesions": Which one is the bad one? Improvement of clinical diagnosis with reflectance confocal microscopy.

BackgroundIn vivo reflectance confocal microscopy (RCM) is a novel non-invasive diagnostic tool, which is used to differentiate skin lesions. Even in lesions with similar dermatoscopic images, RCM may improve diagnostic accuracy.MethodsThree sets of false “twin lesions” with similar macroscopic and dermatoscopic images are matched. All lesions are evaluated with RCM and lesions are excised for further evaluation. Corresponding features in confocal images, dermatoscopy and histopathology are discussed.ResultsIn all matched pairs, one of the lesions was diagnosed as melanoma with the observation of melanoma findings such as: epidermal disarray, pagetoid cells in epidermis and cellular atypia at the junction. Benign lesions were differentiated easily with RCM imaging.ConclusionExamining dermatoscopically difficult and/or similar lesions with RCM facilitates diagnostic and therapeutic decision making. Using RCM in daily practice may contribute to a decrease in unnecessary excisions.

Dermatoscopy Image Characteristics and Differences among Commonly Used Standard Dermatoscopes.

Dermatoscopy Image Characteristics and Differences among Commonly Used Standard Dermatoscopes.