Computational Model Based on Optical Coherence Tomography (OCT) Skin Scanning to Identify and Quantify Acute Radiation Dermatitis (ARD): A Prospective Diagnostic Study.

BACKGROUND: Acute radiation dermatitis (ARD) is the most widely reported radiotherapy-induced adverse event. Currently, there is no objective or reliable method to measure ARD. OBJECTIVE: Our main objective was to identify and quantify the effects of radiotherapy with a computational model using optical coherence tomography (OCT) skin scanning. Secondary objectives included determining the ARD impact of different radiotherapeutic schemes and adjuvant topical therapies. METHODS: We conducted a prospective, single-center case series study in a tertiary referral center of patients with breast cancer who were eligible for whole breast radiotherapy (WBRT). RESULTS: A total of 39 women were included and distributed according to the radiotherapeutic schemes (15, 20, and 25 fractions). A computational model was designed to quantitatively analyze OCT findings. After radiotherapy, OCT scanning was more sensitive revealing vascularization changes in 84.6% of the patients (vs 69.2% of the patients with ARD by clinical examination). OCT quantified an increased vascularization at the end of WBRT (P<.05) and a decrease after 3 months (P=.032). Erythematous skin changes by OCT were more pronounced in the 25-fraction regime. CONCLUSION: An OCT computational model allowed for the identification and quantification of vascularization changes on irradiated skin, even in the absence of clinical ARD. This may allow the design of standardized protocols for ARD beyond the skin color of the patients involved.

New regulation of medical devices in the EU: impact in dermatology.

Medical device (MD) is a broad term that encompasses products ranging from, for example, examination gloves to digital dermoscopy systems; all of which are regulated by a new regulatory framework in the EU from May 2021. The new Medical Device Regulation (MDR) (Regulation EU 2017/745) will have a significant effect on suppliers of MD and will have subsequent effects also for dermatologists and other clinicians. Medical device software and apps are reclassified leading to more stringent requirements on documentation within, e.g. clinical evidence, as well as regulatory authority control. The changes will likely have positive effects on quality, to the benefit of patients. There will, however, be implications affecting the availability and support of existing devices and the introduction of new devices, as well as a likely price increase due to the higher costs for suppliers. Dermatologists, other clinicians and administrators need to be aware of the effects of MDR to ensure that existing devices and new purchases can be used as planned. Specifically, clinicians need to be aware of the following: (i) improved quality of MD and follow-up of incidents can be expected. (ii) Only 'non-significant' updates will be permitted after May 2021 to many existing systems and devices unless approved under the new MDR. (iii) Existing devices that do not achieve approval under the new regulation will no longer be manufactured after May 2024. (iv) New products and methods will take longer time to be approved and available. (v) Prices will likely increase. (vi) Suppliers of products that do not fulfil the new regulation will disappear, and the availability of consumables, spare parts or upgrades might be discontinued. (vii) A trend to oligopoly may appear in the market. It is therefore important to check with your suppliers as to how and when they will adhere to the new MDR regulation.

Uso del aprendizaje automático en el diagnóstico del melanoma. Limitaciones por superar / Machine Learning in Melanoma Diagnosis. Limitations About to be Overcome

ANTECEDENTES: La clasificación automática de imágenes es una rama prometedora del aprendizaje automático (de sus siglas en inglés Machine Learning [ML]), y es una herramienta útil en el diagnóstico de cáncer de piel. Sin embargo, poco se ha estudiado acerca de las limitaciones de su uso en la práctica clínica diaria. OBJETIVO: Determinar las limitaciones que existen en cuanto a la selección de imágenes usadas para el análisis por ML de las neoplasias cutáneas, en particular del melanoma. MÉTODOS: Se diseñó un estudio de cohorte retrospectivo, donde se incluyeron de forma consecutiva 2.849 imágenes dermatoscópicas de alta calidad de tumores cutáneos para su valoración por un sistema de ML, recogidas entre los años 2010 y 2014. Cada imagen dermatoscópica fue clasificada según las características de elegibilidad para el análisis por ML. RESULTADOS: De las 2.849 imágenes elegidas a partir de nuestra base de datos, 968 (34%) cumplieron los criterios de inclusión. De los 528 melanomas, 335 (63,4%) fueron excluidos. La ausencia de piel normal circundante (40,5% de todos los melanomas de nuestra base de datos) y la ausencia de pigmentación (14,2%) fueron las causas más frecuentes de exclusión para el análisis por ML. DISCUSIÓN: Solo el 36,6% de nuestros melanomas se consideraron aceptables para el análisis por sistemas de ML de última generación. Concluimos que los futuros sistemas de ML deberán ser entrenados a partir de bases de datos más grandes que incluyan imágenes representativas de la práctica clínica habitual. Afortunadamente, muchas de estas limitaciones están siendo superadas gracias a los avances realizados recientemente por la comunidad científica, como se ha demostrado en trabajos recientes

BACKGROUND: Automated image classification is a promising branch of machine learning (ML) useful for skin cancer diagnosis, but little has been determined about its limitations for general usability in current clinical practice. OBJECTIVE: To determine limitations in the selection of skin cancer images for ML analysis, particularly in melanoma. METHODS: Retrospective cohort study design, including 2,849 consecutive high-quality dermoscopy images of skin tumors from 2010 to 2014, for evaluation by a ML system. Each dermoscopy image was assorted according to its eligibility for ML analysis. RESULTS: Of the 2,849 images chosen from our database, 968 (34%) met the inclusion criteria for analysis by the ML system. Only 64.7% of nevi and 36.6% of melanoma met the inclusion criteria. Of the 528 melanomas, 335 (63.4%) were excluded. An absence of normal surrounding skin (40.5% of all melanomas from our database) and absence of pigmentation (14.2%) were the most common reasons for exclusion from ML analysis. DISCUSSION: Only 36.6% of our melanomas were admissible for analysis by state-of-the-art ML systems. We conclude that future ML systems should be trained on larger datasets which include relevant non-ideal images from lesions evaluated in real clinical practice. Fortunately, many of these limitations are being overcome by the scientific community as recent works show

Machine Learning in Melanoma Diagnosis. Limitations About to be Overcome. / Uso del aprendizaje automático en el diagnóstico del melanoma. Limitaciones por superar.

BACKGROUND: Automated image classification is a promising branch of machine learning (ML) useful for skin cancer diagnosis, but little has been determined about its limitations for general usability in current clinical practice. OBJECTIVE: To determine limitations in the selection of skin cancer images for ML analysis, particularly in melanoma. METHODS: Retrospective cohort study design, including 2,849 consecutive high-quality dermoscopy images of skin tumors from 2010 to 2014, for evaluation by a ML system. Each dermoscopy image was assorted according to its eligibility for ML analysis. RESULTS: Of the 2,849 images chosen from our database, 968 (34%) met the inclusion criteria for analysis by the ML system. Only 64.7% of nevi and 36.6% of melanoma met the inclusion criteria. Of the 528 melanomas, 335 (63.4%) were excluded. An absence of normal surrounding skin (40.5% of all melanomas from our database) and absence of pigmentation (14.2%) were the most common reasons for exclusion from ML analysis. DISCUSSION: Only 36.6% of our melanomas were admissible for analysis by state-of-the-art ML systems. We conclude that future ML systems should be trained on larger datasets which include relevant non-ideal images from lesions evaluated in real clinical practice. Fortunately, many of these limitations are being overcome by the scientific community as recent works show.

Basal cell carcinoma characterization using fusion ex vivo confocal microscopy: a promising change in conventional skin histopathology.

BACKGROUND: Ex vivo confocal microscopy (CM) works under two modes, fluorescence and reflectance, allowing the visualization of different structures. Fluorescence CM (FCM) requires a contrast agent and has been used for the analysis of basal cell carcinomas (BCCs) during Mohs surgery. Conversely, reflectance CM (RCM) is mostly used for in vivo diagnosis of equivocal skin tumours. Recently, a new, faster ex vivo confocal microscope has been developed which simultaneously uses both lasers (fusion mode). OBJECTIVES: To describe the BCC features identified on reflectance, fluorescence and fusion modes using this novel device. To determine the best mode to identify characteristic BCC features. To develop a new staining protocol to improve the visualization of BCC under the different modes. METHODS: From September 2016 to June 2017, we prospectively included consecutive BCCs which were excised using Mohs surgery in our department. The lesions were evaluated using ex vivo CM after routine Mohs surgery. The specimens were first stained with acridine orange and then stained using both acetic acid and acridine orange. RESULTS: We included 78 BCCs (35 infiltrative, 25 nodular, 12 micronodular, 6 superficial). Most features were better visualized with the fusion mode using the double staining. We also identified new CM ex vivo features, dendritic and plump cells, which have not been reported previously. CONCLUSIONS: Our results suggest that nuclei characteristics are better visualized in FCM but cytoplasm and surrounding stroma are better visualized in RCM. Thus, the simultaneous evaluation of reflectance and fluorescence seems to be beneficial due to its complementary effect. What's already known about this topic? Ex vivo fluorescent confocal microscopy (FCM) is an imaging technique that allows histopathological analysis of fresh tissue. FCM is faster - at least one-third of the time - than conventional methods. FCM has a sensitivity of 88% and a specificity of 99% in detecting basal cell carcinomas (BCCs). What does this study add? Reflectance and fluorescence modes can be used simultaneously in a new ex vivo CM device. Each mode complements the other, resulting in an increase in the detection of BCC features in fusion mode. A combined staining using acetic acid and acridine orange enhances the visualization of tumour and stroma without damaging the tissue for further histopathological analysis.