Clinician’s Ability to Identify Non-Melanoma Skin Cancer on 3D-Total Body Photography Sectors that Were Initially Identified during In-Person Skin Examination with Dermoscopy

Introduction: Non-melanoma skin cancer (NMSC) is a cause of significant morbidity and mortality in high-risk individuals. Total body photography (TBP) is currently used to monitor melanocytic lesions in patients with high risk for melanoma. The authors examined if three-dimensional (3D)-TBP could be useful for diagnosis of NMSC. Methods: Patients (n = 129; 52 female, 77 male) with lesions suspicious for NMSC who had not yet had a biopsy underwent clinical examination followed by examination of each lesion with 3D-TBP Vectra®WB360 (Canfield Scientific, Parsippany, NJ, USA) and dermoscopy. Results: The 129 patients had a total of 182 lesions. Histological examination was performed for 158 lesions; the diagnoses included basal cell carcinoma (BCC; n = 107), squamous cell carcinoma (SCC; n = 27), in-situ SCC (n = 15). Lesions were located in the head/neck region (n = 138), trunk (n = 21), and limbs (n = 23). Of the 182 lesions examined, 12 were not visible on 3D-TBP; reasons for not being visible included location under hair and on septal of nose. Two lesions appeared only as erythema in 3D-TBP but were clearly identifiable on conventional photographs. Sensitivity of 3D-TBP was lower than that of dermoscopy for BCC (73% vs. 79%, p = 0.327), higher for SCC (81% vs. 74%, p = 0.727), and lower for in-situ SCC (0% vs. 33%, p = 125). Specificity of 3D-TBP was lower than that of dermoscopy for BCC (77% vs. 82%, 0.581), lower for SCC (75% vs. 84%, p = 0.063), and higher for in-situ SCC (97% vs. 94%, p = 0.344). Diagnostic accuracy of 3D-TBP was lower than that of dermoscopy for BCC (75% vs. 80%), lower for SCC (76% vs. 82%), and lower for in-situ SCC (88% vs. 89%). Lesion location was not associated with diagnostic confidence in dermoscopy (p = 0.152) or 3D-TBP (p = 0.353). If only lesions with high confidence were included in the calculation, diagnostic accuracy increased for BCC (n = 27; sensitivity 85%, specificity 85%, diagnostic accuracy 85%), SCC (n = 10; sensitivity 90%, specificity 80%, diagnostic accuracy 83%), and for in-situ SCC (n = 2; sensitivity 0%, specificity 100%, diagnostic accuracy 95%). Conclusion: Diagnostic accuracy appears to be slightly lower for 3D-TBP in comparison to dermoscopy. However, there is no statistically significant difference in the sensitivity and specificity of 3D-TBP and dermoscopy for NMSC. Diagnostic accuracy increases, if only lesions with high confidence are included in the calculation. Further studies are necessary to determine if 3D-TBP can improve management of NMSC.


Introduction
The incidence of non-melanoma skin cancer (NMSC) is increasing [1][2][3].Regular monitoring is essential for highrisk individuals such as organ transplant patients and those with familial cancer syndromes and occupational sun exposure.For all skin cancers, early diagnosis ensures better outcomes [4][5][6]; therefore, patients with high risk for NMSC might benefit from additional monitoring with total body photography (TBP).TBP is currently used for monitoring of patients with high risk for melanoma and has been shown to result in earlier diagnosis [7][8][9][10][11][12][13][14][15][16][17][18][19].To the best of our knowledge, no study to date has investigated the value of TBP for diagnosis of NMSC.
Three-dimensional TBP (3D-TBP) is a technique that uses a large number of overlapping images of the body surface taken from different angles to create a 3D image of the body surface.In contrast to conventional non-3D TBP obtained with one camera, with 3D-TBP, the calculated vectors allow the allocation of each skin area in three dimensions.Furthermore, modern software can detect lesions and display them sorted by different characteristics for easier diagnosis.Thus, 3D-TBP allows structured diagnosis of all skin lesions from a vertical viewing angle and the direct visualization of changes in skin lesions over time.The aims of the present study were to examine whether NMSC and its precursors can be detected and reliably diagnosed with 3D-TBP, to compare the diagnostic value of 3D-TBP versus dermoscopy, and to identify the limitations of 3D-TBP in the diagnosis of NMSC.

Materials and Methods
This prospective study was performed in 129 patients who were treated between April 21, 2021 and June 16, 2022 at the Department of Dermatology, University Hospital Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany.The study was conducted in accordance with the Declaration of Helsinki and with the approval of the Local Ethics Committee (Ethikkommission an der TU Dresden, BO-EK-29012021) and was registered with clinicaltrials.gov(NCT05842421).
Patients were enrolled for this study if they had lesions suspicious for NMSC and had not yet undergone biopsy.Each patient underwent a full clinical examination followed by examination by dermoscopy and 3D-TBP.Vectra ® WB360 (Canfield Scientific GmbH, Parsippany, NJ, USA) and DermaGraphix ® (Canfield Scientific GmbH) were used for image acquisition and diagnosis.Non-polarized flash was used during image acquisition.
The clinician noted the features of the lesion in dermoscopy and the primary diagnosis by dermoscopy.Dermoscopy was performed at the time of the clinical examination, and the images were interpreted by one clinician immediately at acquisition; the 3D-TBP images were interpreted separately by another clinician who was blinded to the results in dermoscopy.The clinician looked at the 3D-avatar both in an overview and zoomed-in image for diagnosis and could examine each lesion from different angles.The clinician got precise information about the number of lesions and lesion location per patient so that it was ensured that the same lesions were examined with both devices.For diagnostic accuracy, diagnoses in dermoscopy and the zoomed-in image of 3D-TBP were compared with histopathology.The clinician did not choose between different categories of diagnoses but noted a specific suspected diagnosis and a confidence level for each device.The diagnoses were afterwards distributed in diagnostic categories that were basal cell carcinoma (BCC), squamous cell carcinoma (SCC), in-situ SCC, other malignant lesions, which included skin cancers other than NMSCs, and benign lesions, which included all benign lesions.The final diagnosis was made by histological findings, if available.If the diagnosis was clear, tumors were excised, and benign lesions received local treatment without histopathological examination.In unclear cases, a punch biopsy was performed.Lesions suspicious for in-situ SCC and superficial BCC were only diagnosed by a punch biopsy or excision biopsy if the clinician had a low confidence level.For calculation of sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of the imaging devices, only lesions with histopathological examination were included; clinical diagnoses without histopathological results were excluded from this calculation.
Actinic keratosis, Bowen's disease, and actinic cheilitis were recorded as in-situ SCC.Patients' age, sex, Fitzpatrick skin type, reason for and type of immunosuppression, and history of occupational sun exposure were recorded.
Statistical Analysis SPSS 28 (IBM Corp., Armonk, NY, USA) was used for statistical analysis.The χ 2 test was used to compare diagnoses by histopathology (BCC, SCC, in-situ SCC, malignant lesions other than NMSC, benign lesions), dermoscopy, and 3D-TBP.The χ 2 test and Fisher's exact test (when number of cases in a group was <5) were used to explore association between confidence and tumor entity, skin type, and tumor location.The McNemar test was used to calculate the significance of differences in sensitivity and specificity.Two-sided p < 0.05 was considered statistically significant.

Results
The 129 patients (77 male, 52 female; median age, 77 years; age range, 42-91 years) had a total of 182 lesions.Skin types included Fitzpatrick types I-III (Table 1).There were nine immunocompromised patients: 3 patients were receiving prednisolone, two were transplant patients, one had an immune defect, and three had leukemia or lymphoma.Eight patients had work-related ultraviolet light-induced lesions.
Histological examination was performed for 158 lesions; tissue was obtained by punch biopsy for 14 cases and excision biopsy for 144 cases (Table 1).The histological diagnoses included BCC (n = 107), SCC (n = 27), in-situ SCC (n = 15), malignant tumors other than NMSC (atypical fibroxanthomas, n = 2), and benign lesions (inflammatory lesion, nevus, sebaceous gland hyperplasia, sebaceous gland adenoma, cylindroma, n = 7).For 19 lesions, no therapy or only local therapy was necessary.Four of the 129 patients did not undergo excision in our hospital; while 2 patients had the procedure performed at a private clinic, 2 patients refused the procedure.The lesions were located in the head/neck region (n = 138), trunk (n = 21), and limbs (n = 23).Twenty-five lesions were not excised, and the diagnoses were based on clinical examination; these lesions included 10 in-situ SCCs, 10 BCCs, and one each of SCC, sebaceous gland hyperplasia, papule, inflammatory lesion, and histiocytoma.

Pitfalls of 3D-TBP
Of the 182 lesions, 12 lesions (in 12 patients) were not visible on 3D-TBP (confidence level 4); these lesions included three located on the upper head, seven located on the face, and one each located on the ear and trunk.Reasons for the lack of visibility of the lesions were located under hair (on forehead or upper head) or the septal of the nose.Two lesions (on forehead) were poorly visible because of incorrect representation of body contours and cutting off parts of the haired head in the image.Two lesions that were clearly visible in the patients' photographs were only visible as erythema on 3D-TBP.For 1 patient, no image was saved.

Discussion
Individuals at high risk for NMSC require close highquality follow-up, but this cannot always be guaranteed, especially in rural areas [18,[20][21][22][23][24][25].Transplant patients tend to have larger tumors and more postoperative complications due to immunosuppression, and therefore early diagnosis is crucial [18,[20][21][22][23][24][25].Meanwhile, patients with occupational ultraviolet-induced skin lesions need regular monitoring for early detection of actinic keratoses and in-situ SCC or NMSC [26].These patients often develop multiple in-situ SCCs and SCCs in sun-damaged skin and receive frequent biopsies and surgery.They are also likely to develop high-risk subtypes of NMSC [27].
Monitoring of high-risk individuals with TBP enables early diagnosis of melanoma and reduces healthcare costs by avoiding unnecessary biopsies [7-9, 14-17, 28].Several studies have demonstrated better diagnostic accuracy with dermoscopy than with conventional clinical examinations, and Congalton et al. [29] found that healthcare costs could be reduced by use of teledermoscopy [29][30][31][32][33]. Whether similar benefits can be achieved with use of TBP for diagnosis and monitoring of NMSC has not yet been investigated [30][31][32][33].
In the present study, compared to dermoscopy, 3D-TBP had slightly lower sensitivity for BCC, higher sensitivity for SCC, and lower sensitivity for in-situ SCC; meanwhile, it had slightly lower specificity for BCC and for SCC, and higher specificity for in-situ SCC.However, the differences between sensitivity and specificity of 3D-TBP and dermoscopy for BCC and SCC were not statistically significant.For the management of NMSC and its precursors, it is most important that lesions are not mislabeled as nonmalignant and, therefore, high specificity for BCC and SCC is important; however, the sensitivity and specificity of dermoscopy and 3D-TBP were both at a similarly high level.The lower resolution of 3D-TBP images compared to dermoscopy images accounts for the slightly reduced diagnostic accuracy of 3D-TBP.Nonetheless, clinicians using 3D-TBP can zoom in and observe each lesion from multiple angles.Compared to conventional telemedicine images in two dimensions, the three-dimensional avatar provides supplementary information such as the morphology of the lesion (nodus or plaque), thereby facilitating finding the right diagnosis.
For in-situ SCC, both dermoscopy and TBP showed low sensitivity.Out of the examined lesions, only 15 insitu SCC lesions were examined in histopathology; 10 cases were diagnosed clinically and were therefore not included in this calculation.In the present study, biopsies of in-situ SCCs and superficial BCCs were performed only in unclear cases, as in clear cases, local treatment was initiated immediately without previous histopathological examination.For calculation of diagnostic accuracy, the diagnoses of dermoscopy and 3D-TBP were compared to the diagnosis in histopathology only and not to the clinical diagnoses.Several lesions (n = 25, including 10 insitu SCCs, 10 BCCs, and one SCC) were diagnosed clinically because of a typical appearance.Dermoscopy and 3D-TBP might have performed more poorly than expected because these clinical diagnoses were excluded of the calculation of diagnostic accuracy.Furthermore, the clinician chose a specific diagnosis in each case.Afterward, diagnoses were distributed in different categories of diagnoses like BCC, SCC, in-situ SCC, other malignant tumors, and other benign lesions.If the clinician would have only chosen between a specific category of diagnosis like BCC ("yes" or "no"), diagnostic accuracy would have probably also been higher.However, in everyday routine, the clinician also notes a specific diagnosis of each lesion.Overall diagnostic accuracy of 3D-TBP was lower than that of dermoscopy for BCC, SCC, and in-situ SCC.
Diagnostic accuracy with both dermoscopy and 3D-TBP increased if only lesions with high confidence were included in the calculation.Reasons for low confidence with 3D-TBP may include lower resolution than with naked eye examination, presence of artifacts, poor visibility in body folds, and heavy body hair.On visible skin, lesion location did not influence confidence.However, overall confidence decreased if lesions that were not visible in 3D-TBP were included in the calculation.Thus, 3D-TBP may be especially useful for lesions with high diagnostic confidence; for other lesions, careful clinical examination may be better.
Sensitivity of 3D-TBP for diagnosing the right subtype was higher for nodular BCCs (68%) than for diagnosing superficial BCCs (57%), quite low for diagnosing sclerodermiform BCCs (47%) and zero for diagnosing keratotic, metatypical, adenoid, and pigmented BCCs.However, several BCCs had more than one subtype and the subtypes nodular-cystic, metatypical, keratotic, and pigmented were all combined with at least one of the other subtypes like nodular, superficial, or sclerodermiform.
An ongoing study in Australia is examining the use of 3D-TBP in an integrated telemedicine network of 15 systems in urban, regional, and remote areas [34].Among 3D-TBP versus Dermoscopy for Diagnosis of NMSC patients with melanoma, outcomes have been found to be poorer in rural regions than in urban regions [16,[35][36][37].Some earlier studies have shown that although patients have concerns regarding privacy, most patients would share identifiable images for their own personal medical record [7,13,14].Among patients with high risk for melanoma, Moye et al. [11] found that monitoring with TBP led to decrease in anxiety.Kravets et al. [36] reported similar diagnostic accuracy with teledermatological examination versus clinical examination and teledermoscopy versus dermoscopy [36,37].Common problems with conventional images sent to teledermatology consultations are poor image quality and wrong focus.Meanwhile, 3D-TBP provides a comprehensive view of all lesions from a perpendicular viewing angle, with exact localization on the skin surface.Technicians can obtain standardized images for teleconsultation [15], with image acquisition taking only a few seconds unlike traditional TBP which requires the patient to assume several standardized positions [38].Thus, 3D-TBP fulfills the prerequisites for teleconsultation.Teledermatology with 3D-TBP could also serve as a triage tool for appointments at the hospital [34].Certain areas of the body (e.g., soles of feet, haired head, retroauricular area, the genital area, the mucous area, and other body folds) cannot be viewed on 3D-TBP [15] and must be examined separately by a physician.For telemedicine consultations, photographs taken by trained nonphysician personnel can fill this gap.There are limitations with use of 3D-TBP.Some regions cannot be fully visualized with 3D-TBP [15].However, NMSC usually occurs in chronically sun-exposed areas [1][2][3] and only rarely affects regions with limited or no visibility [1][2][3].In the present study, despite specific information about lesion location, 12 of 182 lesions were not visible on 3D-TBP.In one case this was because the image had not been saved, while in two other cases, the lesions were not visible because of incorrect representation of body contours and cutting off parts of the haired head in the image (Fig. 1a, b); however, in such cases, the technician can easily correct the errors.In the present study, two lesions were only visible as erythema in 3D-TBP although they were clearly identifiable on conventional photographs (Fig. 1c, d).Inappropriate flash settings and/or overlapping of images may have been responsible.In our patients, we used nonpolarized flash as it is better for revealing scales and nodular tumors.Another limitation of 3D-TBP is the need for a standardized patient posture.
Elderly patients with mobility problems may not be able to assume this posture and cannot, therefore, be diagnosed or monitored reliably with 3D-TBP.Body movement may also cause artifacts.However, the several standardized postures necessary in traditional 2D-TBP to get the whole body represented might be even more error prone [38].Furthermore, the differential diagnoses for NMSC include inflammatory skin diseases such as prurigo nodularis, verrucous lichen planus, and ulcers.Differentiation between benign and erythematous malignant lesions might be difficult.
In the present study, the clinician examining 3D-TBP had precise information about lesion location.We therefore suggest 3D-TBP as an assistance in patients that have been examined by a general practitioner or by a technician before the telemedicine consultation.This clinical examination has two main objectives: to identify suspicious lesions and to examine body regions not readily visible with 3D-TBP.Although 3D-TBP cannot replace clinical examination with dermoscopy, the findings of this study indicate that diagnostic accuracy in preselected lesions is only slightly lower than that of dermoscopy.Therefore, 3D-TBP can potentially enhance patient care in a telemedicine context.There are limitations in teledermatology, and 3D-TBP will be most valuable as an adjunct to regular clinical examination and dermoscopy for monitoring patients with high risk for NMSC.

Limitations
A major limitation of this study is the small sample size.Histopathological examination was performed for 107 BCCs but for only 27 SCCs and 15 in-situ SCCs.All patients in this study were preselected by dermatologists in the outpatient department and sent to our clinic with lesions suspicious for malignancy.The results might be different if 3D-TBP were to be used in routine clinical examination.Most patients were sent to our clinic with at least one lesion suspicious for NMSC detected by their dermatologist.Some of the patients had a regular skin examination performed by their dermatologist before, but in several patients, only the suspicious lesion had been examined by their dermatologist in the outpatient setting.As we performed a skin examination in all patients, we found 182 lesions in 129 patients that included also smaller and more subtle lesions that were not found in the outpatient department.Furthermore, the clinician examining the 3D-TBP had precise information about the number of lesions and lesion location to ensure that the same lesions were examined with both devices.Results might be different, if the clinician examining 3D-TBP had to identify suspicious lesions in a virtual clinical examination without prior clinical examination and without additional information about the patient.

Conclusion
Overall diagnostic accuracy appears to be slightly lower for 3D-TBP in comparison to dermoscopy.However, there is no statistically significant difference in the sensitivity and specificity of 3D-TBP and dermoscopy for BCC, SCC, and in-situ SCC.For both modalities, diagnostic accuracy increased, if only lesions with high confidence were evaluated.Additional diagnostic tests should be performed for lesions with moderate or low confidence.Further studies are necessary to establish the value of 3D-TBPused alone or as an adjunct to clinical examinationfor diagnosis and monitoring of NMSC and to determine whether 3D-TBP leads to improved management of lesions in patients with high-risk for NMSC.

Key Message
3D-Total body photography can be a useful tool for diagnosis of non-melanoma skin cancer.

Statement of Ethics
The study was conducted according to the Declaration of Helsinki and with the approval of the Local Ethics Committee (Ethikkommission an der TU Dresden, BO-EK-29012021).

Fig. 1 .
Fig. 1. a Section of a 3D-TBP of a BCC (arrow) on the left forehead of a female patient.b BCC (arrow) on the left forehead of the same patient.c section of a 3D-TBP of two BCCs (arrows) on the trunk of a male patient.The lower BCC is only visible as erythema (d) two BCCs (arrows) on the trunk of the same patient.

Table 2 .
Sensitivity, specificity, positive predictive value, and diagnostic accuracy of dermoscopy and 3D-TBP

Table 3 .
Sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of 3D-TBP for BCC *Several BCCs had two or three subtypes, resulting in two or three diagnoses.