Empirical validation of Conformal Prediction for trustworthy skin lesions classification.

BACKGROUND AND OBJECTIVE: Uncertainty quantification is a pivotal field that contributes to realizing reliable and robust systems. It becomes instrumental in fortifying safe decisions by providing complementary information, particularly within high-risk applications. existing studies have explored various methods that often operate under specific assumptions or necessitate substantial modifications to the network architecture to effectively account for uncertainties. The objective of this paper is to study Conformal Prediction, an emerging distribution-free uncertainty quantification technique, and provide a comprehensive understanding of the advantages and limitations inherent in various methods within the medical imaging field. METHODS: In this study, we developed Conformal Prediction, Monte Carlo Dropout, and Evidential Deep Learning approaches to assess uncertainty quantification in deep neural networks. The effectiveness of these methods is evaluated using three public medical imaging datasets focused on detecting pigmented skin lesions and blood cell types. RESULTS: The experimental results demonstrate a significant enhancement in uncertainty quantification with the utilization of the Conformal Prediction method, surpassing the performance of the other two methods. Furthermore, the results present insights into the effectiveness of each uncertainty method in handling Out-of-Distribution samples from domain-shifted datasets. Our code is available at: github.com/jfayyad/ConformalDx. CONCLUSIONS: Our conclusion highlights a robust and consistent performance of conformal prediction across diverse testing conditions. This positions it as the preferred choice for decision-making in safety-critical applications.

Autofluorescence Excitation Imaging of Nonmelanoma Skin Cancer for Margin Assessment Before Mohs Micrographic Surgery: A Pilot Study.

BACKGROUND: Autofluorescence photography can detect specific light-tissue interactions and record important pathophysiological changes associated with nonmelanoma skin cancer (NMSC), which has been ascribed to the fluorescence of an aromatic amino acid, tryptophan. OBJECTIVE: To assess the impact of a novel, autofluorescence imaging (AFI) device on margin control for NMSCs before Mohs micrographic surgery (MMS) in an effort to decrease overall operating time. METHODS: Before the initial stage of MMS, NMSCs were measured with a 2-mm margin as standard of care (normal margin). The tumor was then imaged with the AFI device. A 2-mm margin was drawn around the fluorescent area captured by the AFI device and was referred to as the camera margin. The tumor was excised based on the normal margin and evaluated on frozen histological section. RESULTS: Imaging based on the AFI device resulted in appropriate recommendations for margin control in 8 of 11 tumors. Four of these tumors did not fluoresce and demonstrated a lack of tumor residuum on stage I specimen, as anticipated. There were no side effects from the AFI device. CONCLUSION: This is an initial pilot study that supports the use of a novel, noninvasive imaging device to help with margin assessment before MMS. On optimization, this device has potential to extend applicability to surgical excisions for tumors that do not fulfill criteria for MMS.

Assessment of perineural spread in advanced cutaneous squamous cell carcinomas treated with immunotherapy.

BACKGROUND: Cutaneous squamous cell carcinoma (CSCC) has a propensity for perineural spread (PNS) which is associated with poorer treatment outcomes. Immunotherapy is the new standard of care treatment for advanced CSCC resulting in durable responses. PNS is not captured by traditional response assessment criteria used in clinical trials, e.g. RECIST 1.1, and there is limited literature documenting radiological PNS responses to immunotherapy. In this study we assess PNS responses to immunotherapy using a modified grading system. METHODS: This is an Australian single-center retrospective review of patients with advanced CSCC who were treated with immunotherapy between April 2018 and February 2022 who had evidence of PNS on pre-treatment magnetic-resonance imaging (MRI). The primary outcome was blinded overall radiological response in PNS using graded radiological criteria, post-commencement of immunotherapy. Three defined timepoints (< 5 months, 5-10 months, > 10 months) were reviewed. Secondary outcomes included a correlation between RECIST 1.1 and PNS assessments and the assessment of PNS on fluorodeoxyglucose (FDG)-positron emission tomography (PET)/computed tomography (CT). RESULTS: Twenty CSCC patients treated with immunotherapy were identified. Median age was 75.7 years and 75% (n = 15) were male. All patients had locoregionally advanced disease and no distant metastases. Median follow-up was 18.5 months (range: 2-59). 70% (n = 14) demonstrated a PNS response by 5 months. Three patients experienced pseudoprogression. One patient had PNS progression by the end of study follow up. RECIST 1.1 and PNS responses were largely concordant at > 10 months (Cohen's Kappa 0.62). 5/14 cases had features suspicious for PNS on FDG-PET/CT. CONCLUSIONS: PNS response to immunotherapy can be documented on MRI using graded radiological criteria. High response rates were seen in PNS with the use of immunotherapy in this cohort and these responses were largely concordant with RECIST 1.1 assessments. FDG-PET/CT demonstrated limited sensitivity in the detection of PNS.

Photoacoustic imaging for cutaneous melanoma assessment: a comprehensive review.

Significance: Cutaneous melanoma (CM) has a high morbidity and mortality rate, but it can be cured if the primary lesion is detected and treated at an early stage. Imaging techniques such as photoacoustic (PA) imaging (PAI) have been studied and implemented to aid in the detection and diagnosis of CM. Aim: Provide an overview of different PAI systems and applications for the study of CM, including the determination of tumor depth/thickness, cancer-related angiogenesis, metastases to lymph nodes, circulating tumor cells (CTCs), virtual histology, and studies using exogenous contrast agents. Approach: A systematic review and classification of different PAI configurations was conducted based on their specific applications for melanoma detection. This review encompasses animal and preclinical studies, offering insights into the future potential of PAI in melanoma diagnosis in the clinic. Results: PAI holds great clinical potential as a noninvasive technique for melanoma detection and disease management. PA microscopy has predominantly been used to image and study angiogenesis surrounding tumors and provide information on tumor characteristics. Additionally, PA tomography, with its increased penetration depth, has demonstrated its ability to assess melanoma thickness. Both modalities have shown promise in detecting metastases to lymph nodes and CTCs, and an all-optical implementation has been developed to perform virtual histology analyses. Animal and human studies have successfully shown the capability of PAI to detect, visualize, classify, and stage CM. Conclusions: PAI is a promising technique for assessing the status of the skin without a surgical procedure. The capability of the modality to image microvasculature, visualize tumor boundaries, detect metastases in lymph nodes, perform fast and label-free histology, and identify CTCs could aid in the early diagnosis and classification of CM, including determination of metastatic status. In addition, it could be useful for monitoring treatment efficacy noninvasively.

Cost-effective device for locating and circumscribing superficial tumors with contrast enhancement and fluorescence quantification.

BACKGROUND: Two Bispectral contrast enhancement approaches for the fluorescence diagnosis with chlorine-e6 and a wide field-of-view imaging system with fluorescence excitation at 405 nm and time-resolved background suppression were analyzed and compared. METHODS: Two techniques for the contrast enhancement of a fluorescent video system (Red/Green (R/G) ratio and Red-Green (R-G)) with time-resolved background suppression for fluorescent diagnosis (FD) were tested in four patients with basal cell carcinoma (BCC). RESULTS: The results of both contrast enhancement methods were compared for the diagnostic efficiency for FD of BCC. Both techniques successfully determined the boundaries of the lesions and the fluorescence intensity. CONCLUSIONS: Both contrast enhancement modes have proven effective in identifying tumor borders in cases of low contrast in BCC FD with Ce6. While the Red/Green (R/G) mode provides sharper lesion borders, the Red minus Green (R-G) mode visualizes more fluorescent features and makes it easier to assess the lesion margins.

Assessment of the Diagnostic Performance of Clinical Examinations and High-Frequency Ultrasound in Patients With Pigmented Skin Tumors.

OBJECTIVES: To investigate whether the integration of high-frequency ultrasound (HFUS) to routine clinical examinations could improve diagnostic performance and management decision for pigmented skin tumors. METHODS: Three general practitioners trained previously and a dermatologist independently assessed pigmented skin tumors and rendered management decision based on clinical examinations alone or clinical examinations integrating HFUS. RESULTS: After integrating HFUS, the diagnostic area under the curve (AUC) (0.658-0.693 versus 0.848, all P < .05) and specificity (46.6-58.6% versus 89.7%, all P < .05) for pigmented skin malignancies were improved for general practitioners, meanwhile unnecessary biopsy rate reduced (42.9-53.6% versus 10.7%, P < .001). To the dermatologist, the diagnostic AUC (0.822 versus 0.949, P < .001), sensitivity (81.7% versus 96.7%, P = .012) and specificity (0.828 versus 0.931, P = .031) improved significantly, meanwhile both missed biopsy rate (14.5% versus 4.8%, P = .031) and unnecessary biopsy rate (19.6% versus 7.1%, P = .016) decreased. Additionally, the diagnostic performance of the general practitioner with integrating HFUS could be comparable with the dermatologist based on clinical examinations alone (all P > .05). CONCLUSIONS: As a complementary tool of clinical examinations, HFUS could help physicians differentiate pigmented skin malignancies and manage decision.

Combined ultrasound and photoacoustic C-mode imaging system for skin lesion assessment.

Accurate assessment of the size and depth of infiltration is critical for effectively treating and removing skin cancer, especially melanoma. However, existing methods such as skin biopsy and histologic examination are invasive, time-consuming, and may not provide accurate depth results. We present a novel system for simultaneous and co-localized ultrasound and photoacoustic imaging, with the application for non-invasive skin lesion size and depth measurement. The developed system integrates an acoustical mirror that is placed on an ultrasound transducer, which can be translated within a flexible water tank. This allows for 3D (C-mode) imaging, which is useful for mapping the skin structure and determine the invasion size and depth of lesions including skin cancer. For efficient reconstruction of photoacoustic images, we applied the open-source MUST library. The acquisition time per 2D image is <1 s and the pulse energies are below the legal Maximum Permissible Exposure (MPE) on human skin. We present the depth and resolution capabilities of the setup on several self-designed agar phantoms and demonstrate in vivo imaging on human skin. The setup also features an unobstructed optical window from the top, allowing for simple integration with other optical modalities. The perspective towards clinical application is demonstrated.

Comparative dermoscopy assessment of nevus-associated versus de novo in situ melanoma.

BACKGROUND: Dermoscopic features differentiating in situ nevus-associated melanoma (NAM) versus in situ de novo melanoma (DNM) are inconclusive. OBJECTIVES: The aim of the study was to investigate the dermoscopic features associated with in situ NAM versus DNM. MATERIALS & METHODS: This was a retrospective observational study. All consecutive in situ melanomas diagnosed in adult patients were retrieved and stratified as NAM vs DNM, and clinical and dermoscopic data were compared between the two. RESULTS: A total of 183 patients with in situ melanoma were collected, of whom 98 (54%) were male with a mean age of 64±14 years. For 129 patients, standardized dermoscopic images were collected (51 for NAM and 78 for de novo MM). The most common dermoscopic features were an atypical pigment network (85%), atypical globules (63%) and regression (42%). No significant differences were found except for regression, which was detected in 54.9% NAM vs 33.3% DNM (p=0.016). Multivariate logistic regression confirmed the association between dermoscopic regression and NAM (OR=2.34, 95% CI: 1.15-4.91). CONCLUSION: Currently, the use of dermoscopy to determine whether a melanoma is associated with a nevus is unreliable, however, the presence of regression adjacent to atypical lesions may raise suspicion of in situ NAM.

Ultrasound Assessment of Skin Tumors Local Recurrence.

Skin cancer may recur at or around the surgical site despite wide excisions. Prompt clinical and sonographic detection of local recurrence is important since subjects with relapsing melanomas or nonmelanoma malignancies can be managed efficaciously, with a relevant impact on morbidity and survival. Ultrasound is being employed with increasing frequency in the assessment of skin tumors, but most of the published articles relate to initial pretherapeutic diagnosis and staging. This review aims to offer an illustrated guide to the sonographic evaluation of locally recurring skin cancer. We introduce the topic, then we provide some sonographic tips for patient follow-up, then we describe the ultrasound findings in case of local recurrence, illustrating the main mimickers, and finally, we mention the role of ultrasound in guiding diagnostic and therapeutic percutaneous procedures.

Multispectral Imaging for Skin Diseases Assessment-State of the Art and Perspectives.

Skin optical inspection is an imperative procedure for a suspicious dermal lesion since very early skin cancer detection can guarantee total recovery. Dermoscopy, confocal laser scanning microscopy, optical coherence tomography, multispectral imaging, multiphoton laser imaging, and 3D topography are the most outstanding optical techniques implemented for skin examination. The accuracy of dermatological diagnoses attained by each of those methods is still debatable, and only dermoscopy is frequently used by all dermatologists. Therefore, a comprehensive method for skin analysis has not yet been established. Multispectral imaging (MSI) is based on light-tissue interaction properties due to radiation wavelength variation. An MSI device collects the reflected radiation after illumination of the lesion with light of different wavelengths and provides a set of spectral images. The concentration maps of the main light-absorbing molecules in the skin, the chromophores, can be retrieved using the intensity values from those images, sometimes even for deeper-located tissues, due to interaction with near-infrared light. Recent studies have shown that portable and cost-efficient MSI systems can be used for extracting skin lesion characteristics useful for early melanoma diagnoses. This review aims to describe the efforts that have been made to develop MSI systems for skin lesions evaluation in the last decade. We examined the hardware characteristics of the produced devices and identified the typical structure of an MSI device for dermatology. The analyzed prototypes showed the possibility of improving the specificity of classification between the melanoma and benign nevi. Currently, however, they are rather adjuvants tools for skin lesion assessment, and efforts are needed towards a fully fledged diagnostic MSI device.

Study protocol for a randomised controlled trial to evaluate the use of melanoma surveillance photography to the Improve early detection of MelanomA in ultra-hiGh and high-risk patiEnts (the IMAGE trial).

INTRODUCTION: Melanoma surveillance photography (MSP) is a comprehensive surveillance method that comprises two- or three-dimensional total body photography with tagged digital dermoscopy, performed at prescribed intervals. It has the potential to reduce unnecessary biopsies and enhance early detection of melanoma, but it is not yet standard care for all high-risk patients in Australia. This protocol describes a randomised controlled trial (RCT) designed to evaluate the clinical impact and cost-effectiveness of using MSP for the surveillance of individuals at ultra-high or high risk of melanoma from a health system perspective. METHODS AND DESIGN: This is a registry-based, unblinded, multi-site, parallel-arm RCT that will be conducted over 3 years. We aim to recruit 580 participants from three Australian states: Victoria, New South Wales and Queensland, via state cancer registries or direct referral from clinicians. Eligible participants within 24 months of a primary cutaneous melanoma diagnosis will be randomised 1:1 to receive either MSP in addition to their routine clinical surveillance (intervention group) or routine clinical surveillance without MSP (control group). Most participants will continue surveillance with their usual care provider, and the frequency of follow-up visits in both groups will depend on the stage of their primary melanoma and risk factors. The primary outcome measure of the study is the number of unnecessary biopsies (i.e. false positives, being cases where a lesion is biopsied due to suspected melanoma on clinical examination, either with or without MSP, but the resulting histopathology finding is negative for melanoma). Secondary outcomes include the evaluation of health economic outcomes, quality of life and patient acceptability. Two sub-studies will explore the benefit of MSP in high-risk patients prior to a melanoma diagnosis and the diagnostic performance of MSP in the teledermatology setting compared to the en face clinical setting. DISCUSSION: This trial will determine the clinical efficacy, cost-effectiveness and affordability of MSP to facilitate policy decision-making at the national and local levels, across primary and specialist care. TRIAL REGISTRATION: ClinicalTrials.gov NCT04385732 . Registered on May 13, 2020.

Timing and necessity of staging imaging in clinical stage II cutaneous melanoma: Cost-effectiveness and clinical decision analysis.

BACKGROUND: Preoperative imaging in clinical stage II melanoma is not indicated per National Comprehensive Cancer Network (NCCN) guidelines but remains common in clinical practice. METHODS: Patients presenting with cutaneous clinical stage II melanoma from 2007 to 2019 were retrospectively reviewed. A clinical decision analysis with cost data was designed to understand ideal practice patterns in managing stage II melanoma, with pre-versus selective post-operative imaging as the initial decision node. RESULTS: There were 277 subjects included, and 143 underwent preoperative imaging (49.5%). This changed management (i.e. no surgery) in one patient (0.4%). Overall, 16 patients had additional findings on imaging (5.8%). Upfront surgery with selective postoperative imaging was a more cost-effective strategy than routine performance of preoperative imaging, with savings of $1677 per patient. CONCLUSION: Preoperative imaging is a low yield, costly approach for patients with clinical stage II melanoma with minimal impact on the decision to proceed with surgical management.

Imaging technologies for presurgical margin assessment of basal cell carcinoma.

Basal cell carcinoma is the most common cancer worldwide, necessitating the development of techniques to decrease treatment costs through efficiency and efficacy. Mohs micrographic surgery, a specialized surgical technique involving staged resection of the tumor with complete histologic evaluation of the peripheral margins, is highly utilized. Reducing stages by even 5% to 10% would result in significant improvement in care and economic benefits. Noninvasive imaging could aid in both establishing the diagnosis of suspicious skin lesions and streamlining the surgical management of skin cancers by improving presurgical estimates of tumor sizes. Herein, we review the current state of imaging techniques in dermatology and their applications for diagnosis and tumor margin assessment of basal cell carcinoma prior to Mohs micrographic surgery.

Training dermatology residents in dermatoscopy: A case control lecture series assessment.

Lack of standardized dermatoscopy training limits confidence and accuracy. We assessed the effect of a dermatoscopy lecture series on the diagnostic accuracy of dermatology residents' biopsies. Additionally, we evaluated resident comfort with and knowledge of dermatoscopy before and after the curriculum. Twelve dermatology residents were enrolled in a 5-month dedicated dermatoscopy curriculum. To assess knowledge of and comfort with dermatoscopy, residents were given a 50-question assessment and 21-question survey before and after the curriculum. Change in diagnostic accuracy was assessed by comparing the suspected clinical diagnosis to the final histopathologic diagnosis of lesions biopsied by residents before and after the course. Upon completion of the curriculum, residents felt significantly more comfortable performing dermatoscopy (P = .002) and using dermatoscopy to identify melanocytic nevi (P = .037) and melanomas (invasive and in situ) (P = .012). Postgraduate year 2 residents also showed significantly improved diagnostic accuracy after the training course (odds ratio, 1.33; 95% confidence interval, 1.06-1.67; P = .013). Our study was limited by a small sample size of 12 residents from a single academic institution. A formal dermatoscopy course can effectively improve dermatology residents' knowledge, confidence, and diagnostic accuracy when using dermatoscopy.

Management of cutaneous melanoma: radiologists challenging and risk assessment.

Melanoma patient remains a challenging for the radiologist, due to the difficulty related to the management of a patient more often in an advanced stage of the disease. It is necessary to determine a stratification of risk, optimizing the means, with diagnostic tools that should be optimized in relation to the type of patient, and improving knowledge. Staging and risk assessment procedures are determined by disease presentation at diagnosis. Melanoma staging is a critical tool to assist clinical decision-making and prognostic assessment. It is used for clinical trial design, eligibility, stratification, and analysis. The current standard for regional lymph nodes staging is represented by the sentinel lymph node excision biopsy procedure. For staging of distant metastases, PET-CT has the highest sensitivity and diagnostic odds ratio. Similar trend is observed during melanoma surveillance. The advent of immunotherapy, which has improved patient outcome, however, has determined new issues for radiologists, partly due to atypical response patterns, partly due to adverse reactions that must be identified as soon as possible for the correct management of the patient. The main objectives of the new ir-criteria are to standardize the assessment between different trials. However, these ir-criteria do not take into account all cases of atypical response patterns, as hyperprogression or dissociated responses. None of these criteria has actually been uniformly adopted in routine. The immune-related adverse events (irAEs) can involve various organs from head to toe. It is crucial for radiologists to know the imaging appearances of this condition, to exclude recurrent or progressive disease and for pneumonitis, since it could be potentially life-threatening toxicity, resulting in pneumonitis-related deaths in early phase trials, to allow a proper patient management.

Assessment of melanoma thickness based on dermoscopy images: an open, web-based, international, diagnostic study.

BACKGROUND: Preoperative assessment of whether a melanoma is invasive or in situ (MIS) is a common task that might have important implications for triage, prognosis and the selection of surgical margins. Several dermoscopic features suggestive of melanoma have been described, but only a few of these are useful in differentiating MIS from invasive melanoma. OBJECTIVE: The primary aim of this study was to evaluate how accurately a large number of international readers, individually as well as collectively, were able to discriminate between MIS and invasive melanomas as well as estimate the Breslow thickness of invasive melanomas based on dermoscopy images. The secondary aim was to compare the accuracy of two machine learning convolutional neural networks (CNNs) and the collective reader response. METHODS: We conducted an open, web-based, international, diagnostic reader study using an online platform. The online challenge opened on 10 May 2021 and closed on 19 July 2021 (71 days) and was advertised through several social media channels. The investigation included, 1456 dermoscopy images of melanomas (788 MIS; 474 melanomas ≤1.0 mm and 194 >1.0 mm). A test set comprising 277 MIS and 246 invasive melanomas was used to compare readers and CNNs. RESULTS: We analysed 22 314 readings by 438 international readers. The overall accuracy (95% confidence interval) for melanoma thickness was 56.4% (55.7%-57.0%), 63.4% (62.5%-64.2%) for MIS and 71.0% (70.3%-72.1%) for invasive melanoma. Readers accurately predicted the thickness in 85.9% (85.4%-86.4%) of melanomas ≤1.0 mm (including MIS) and in 70.8% (69.2%-72.5%) of melanomas >1.0 mm. The reader collective outperformed a de novo CNN but not a pretrained CNN in differentiating MIS from invasive melanoma. CONCLUSIONS: Using dermoscopy images, readers and CNNs predict melanoma thickness with fair to moderate accuracy. Readers most accurately discriminated between thin (≤1.0 mm including MIS) and thick melanomas (>1.0 mm).

Noncontrast-enhanced 3-Tesla MRI using surface coil as a complementary test for assessment of distribution and depth of locoregional cutaneous metastases of malignant melanoma.

Locoregional and distant metastases account for most cases of morbidity and mortality associated with melanoma. In addition, local recurrences of melanoma might be the onset of disseminated disease. Therefore, precise diagnosis and therapy are warranted to minimize morbidity and increase survival in a subset of patients. However, the correct distribution of the metastatic lesions on the skin is often difficult to estimate. We present the application of noncontrast-enhanced 3-Tesla MRI using surface coil to detect locoregional cutaneous metastases of malignant melanoma on the basis of the topographic assessment of skin lesions. Furthermore, in a systematic review, we summarize the current knowledge about application of MRI in assessment of location, distribution, and depth of cutaneous primary malignant melanoma. MRI might be applied to evaluate the location, distribution, size, and depth of the locoregional cutaneous metastasis of malignant melanoma to identify the optimal cost-effective treatment strategies and monitor their effects.

Assessment of malignant melanoma lesions using violet-light dermoscopy: A case report.

Malignant melanomas often present with irregular shapes and in multiple shades of brown under white light. Dermoscopy is used to diagnose malignant melanomas; nevertheless, it is often difficult to differentiate malignant melanoma from healthy pigmented skin. The DZ-D100 dermoscope (Casio Computer) is a digital camera equipped with a white light-emitting diode (LED) and a violet LED, which can capture non-polarized/polarized conventional dermoscopy images (CDS) as well as violet-light dermoscopy (VLD) images. Since the absorption wavelength of melanin approaches that of ultraviolet rays, VLD with a wavelength of 405 nm can be used to visualize it. This camera allows three images with the same composition to be captured simultaneously. In this case, we performed dermoscopy with DZ-D100 to determine the surgical resection margins of a melanoma of the heel in a 76-year-old woman. The pale-colored lesions that were difficult to demarcate by CDS were clearly visible by VLD, presenting as dark areas in the grayscale images. Preoperatively determined lesion boundaries with CDS in combination with VLD were histologically more accurate than those with conventional CDS alone. Therefore, the combination of CDS and VLD may reveal the distribution of subtle pigmentation of fine melanin in the skin, making it easier to distinguish between lesions and healthy skin. As one of the limitations, parts of the heel with thick stratum corneum were also observed to be dark gray in the VLD images. Therefore, the evaluation of pigment lesion should be performed by comparing both CDS and VLD.