A novel multi-task learning network for skin lesion classification based on multi-modal clues and label-level fusion.

In this paper, we propose a multi-task learning (MTL) network based on the label-level fusion of metadata and hand-crafted features by unsupervised clustering to generate new clustering labels as an optimization goal. We propose a MTL module (MTLM) that incorporates an attention mechanism to enable the model to learn more integrated, variable information. We propose a dynamic strategy to adjust the loss weights of different tasks, and trade off the contributions of multiple branches. Instead of feature-level fusion, we propose label-level fusion and combine the results of our proposed MTLM with the results of the image classification network to achieve better lesion prediction on multiple dermatological datasets. We verify the effectiveness of the proposed model by quantitative and qualitative measures. The MTL network using multi-modal clues and label-level fusion can yield the significant performance improvement for skin lesion classification.

Cutaneous squamous cell carcinoma characterized by MALDI mass spectrometry imaging in combination with machine learning.

Cutaneous squamous cell carcinoma (SCC) is an increasingly prevalent global health concern. Current diagnostic and surgical methods are reliable, but they require considerable resources and do not provide metabolomic insight. Matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) enables detailed, spatially resolved metabolomic analysis of tissue samples. Integrated with machine learning, MALDI-MSI could yield detailed information pertaining to the metabolic alterations characteristic for SCC. These insights have the potential to enhance SCC diagnosis and therapy, improving patient outcomes while tackling the growing disease burden. This study employs MALDI-MSI data, labelled according to histology, to train a supervised machine learning model (logistic regression) for the recognition and delineation of SCC. The model, based on data acquired from discrete tumor sections (n = 25) from a mouse model of SCC, achieved a predictive accuracy of 92.3% during cross-validation on the labelled data. A pathologist unacquainted with the dataset and tasked with evaluating the predictive power of the model in the unlabelled regions, agreed with the model prediction for over 99% of the tissue areas. These findings highlight the potential value of integrating MALDI-MSI with machine learning to characterize and delineate SCC, suggesting a promising direction for the advancement of mass spectrometry techniques in the clinical diagnosis of SCC and related keratinocyte carcinomas.

Comprehensive analysis of clinical images contributions for melanoma classification using convolutional neural networks.

BACKGROUND: Timely diagnosis plays a critical role in determining melanoma prognosis, prompting the development of deep learning models to aid clinicians. Questions persist regarding the efficacy of clinical images alone or in conjunction with dermoscopy images for model training. This study aims to compare the classification performance for melanoma of three types of CNN models: those trained on clinical images, dermoscopy images, and a combination of paired clinical and dermoscopy images from the same lesion. MATERIALS AND METHODS: We divided 914 image pairs into training, validation, and test sets. Models were built using pre-trained Inception-ResNetV2 convolutional layers for feature extraction, followed by binary classification. Training comprised 20 models per CNN type using sets of random hyperparameters. Best models were chosen based on validation AUC-ROC. RESULTS: Significant AUC-ROC differences were found between clinical versus dermoscopy models (0.661 vs. 0.869, p < 0.001) and clinical versus clinical + dermoscopy models (0.661 vs. 0.822, p = 0.001). Significant sensitivity differences were found between clinical and dermoscopy models (0.513 vs. 0.799, p = 0.01), dermoscopy versus clinical + dermoscopy models (0.799 vs. 1.000, p = 0.02), and clinical versus clinical + dermoscopy models (0.513 vs. 1.000, p < 0.001). Significant specificity differences were found between dermoscopy versus clinical + dermoscopy models (0.800 vs. 0.288, p < 0.001) and clinical versus clinical + dermoscopy models (0.650 vs. 0.288, p < 0.001). CONCLUSION: CNN models trained on dermoscopy images outperformed those relying solely on clinical images under our study conditions. The potential advantages of incorporating paired clinical and dermoscopy images for CNN-based melanoma classification appear less clear based on our findings.

Line-field confocal optical coherence tomography in dermato-oncology: A literature review towards harmonized histopathology-integrated terminology.

Non-invasive diagnostics like line-field confocal optical coherence tomography (LC-OCT) are being implemented in dermato-oncology. However, unification of terminology in LC-OCT is lacking. By reviewing the LC-OCT literature in the field of dermato-oncology, this study aimed to develop a unified terminological glossary integrated with traditional histopathology. A PRISMA-guided literature-search was conducted for English-language publications on LC-OCT of actinic keratosis (AK), keratinocyte carcinoma (KC), and malignant melanoma (MM). Study characteristics and terminology were compiled. To harmonize LC-OCT terminology and integrate with histopathology, synonymous terms for image features of AK, KC, and MM were merged by two authors, organized by skin layer and lesion-type. A subset of key LC-OCT image-markers with histopathological correlates that in combination were typical of AK, squamous cell carcinoma in situ (SCCis), invasive squamous cell carcinoma (SCC), basal cell carcinoma (BCC), and MM in traditional histopathology, were selected from the glossary by an experienced dermatopathologist. Seventeen observational studies of AK (7 studies), KC (13 studies), MM (7 studies) utilizing LC-OCT were included, with 117 terms describing either AK, KC, or MM. These were merged to produce 45 merged-terms (61.5% reduction); 5 assigned to the stratum corneum (SC), 23 to the viable epidermis, 2 to dermo-epidermal junction (DEJ) and 15 to the dermis. For each lesion, mandatory key image-markers were a well-defined DEJ and presence of mild/moderate but not severe epidermal dysplasia for AK, severe epidermal dysplasia and well-defined DEJ for SCCis, interrupted DEJ and/or dermal broad infiltrative strands for invasive SCC, dermal lobules connected and/or unconnected to the epidermis for BCC, as well as single atypical melanocytes and/or nest of atypical melanocytes in the epidermis or dermis for MM. This review compiles evidence on LC-OCT in dermato-oncology, providing a harmonized histopathology-integrated terminology and key image-markers for each lesion. Further evaluation is required to determine the clinical value of these findings.

A novel SpaSA based hyper-parameter optimized FCEDN with adaptive CNN classification for skin cancer detection.

Skin cancer is the most prevalent kind of cancer in people. It is estimated that more than 1 million people get skin cancer every year in the world. The effectiveness of the disease's therapy is significantly impacted by early identification of this illness. Preprocessing is the initial detecting stage in enhancing the quality of skin images by removing undesired background noise and objects. This study aims is to compile preprocessing techniques for skin cancer imaging that are currently accessible. Researchers looking into automated skin cancer diagnosis might use this article as an excellent place to start. The fully convolutional encoder-decoder network and Sparrow search algorithm (FCEDN-SpaSA) are proposed in this study for the segmentation of dermoscopic images. The individual wolf method and the ensemble ghosting technique are integrated to generate a neighbour-based search strategy in SpaSA for stressing the correct balance between navigation and exploitation. The classification procedure is accomplished by using an adaptive CNN technique to discriminate between normal skin and malignant skin lesions suggestive of disease. Our method provides classification accuracies comparable to commonly used incremental learning techniques while using less energy, storage space, memory access, and training time (only network updates with new training samples, no network sharing). In a simulation, the segmentation performance of the proposed technique on the ISBI 2017, ISIC 2018, and PH2 datasets reached accuracies of 95.28%, 95.89%, 92.70%, and 98.78%, respectively, on the same dataset and assessed the classification performance. It is accurate 91.67% of the time. The efficiency of the suggested strategy is demonstrated through comparisons with cutting-edge methodologies.

Mapping cutaneous field carcinogenesis of nonmelanoma skin cancer using mesoscopic imaging of pro-inflammation cues.

Nonmelanoma skin cancers remain the most widely diagnosed types of cancers globally. Thus, for optimal patient management, it has become imperative that we focus our efforts on the detection and monitoring of cutaneous field carcinogenesis. The concept of field cancerization (or field carcinogenesis), introduced by Slaughter in 1953 in the context of oral cancer, suggests that invasive cancer may emerge from a molecularly and genetically altered field affecting a substantial area of underlying tissue including the skin. A carcinogenic field alteration, present in precancerous tissue over a relatively large area, is not easily detected by routine visualization. Conventional dermoscopy and microscopy imaging are often limited in assessing the entire carcinogenic landscape. Recent efforts have suggested the use of noninvasive mesoscopic (between microscopic and macroscopic) optical imaging methods that can detect chronic inflammatory features to identify pre-cancerous and cancerous angiogenic changes in tissue microenvironments. This concise review covers major types of mesoscopic optical imaging modalities capable of assessing pro-inflammatory cues by quantifying blood haemoglobin parameters and hemodynamics. Importantly, these imaging modalities demonstrate the ability to detect angiogenesis and inflammation associated with actinically damaged skin. Representative experimental preclinical and human clinical studies using these imaging methods provide biological and clinical relevance to cutaneous field carcinogenesis in altered tissue microenvironments in the apparently normal epidermis and dermis. Overall, mesoscopic optical imaging modalities assessing chronic inflammatory hyperemia can enhance the understanding of cutaneous field carcinogenesis, offer a window of intervention and monitoring for actinic keratoses and nonmelanoma skin cancers and maximise currently available treatment options.

Dermoscopy to differentiate clinically similar inflammatory and neoplastic skin lesions.

INTRODUCTION: Over the few last decades, dermoscopy has become an invaluable and popular imaging technique that complements the diagnostic armamentarium of dermatologists, being employed for both tumors and inflammatory diseases. Whereas distinction between neoplastic and inflammatory lesions is often straightforward based on clinical data, there are some scenarios that may be troublesome, e.g., solitary inflammatory lesions or tumors superimposed to a widespread inflammatory condition that may share macroscopic morphological findings. EVIDENCE ACQUISITION: We reviewed the literature to identify dermoscopic clues to support the differential diagnosis of clinically similar inflammatory and neoplastic skin lesions, also providing the histological background of such dermoscopic points of differentiation. EVIDENCE SYNTHESIS: Dermoscopic differentiating features were identified for 12 relatively common challenging scenarios, including Bowen's disease and basal cell carcinoma vs. psoriasis and dermatitis, erythroplasia of Queyrat vs. inflammatory balanitis, mammary and extramammary Paget's disease vs. inflammatory mimickers, actinic keratoses vs. discoid lupus erythematosus, squamous cell carcinoma vs. hypertrophic lichen planus and lichen simplex chronicus, actinic cheilitis vs. inflammatory cheilitis, keratoacanthomas vs. prurigo nodularis, nodular lymphomas vs. pseudolymphomas and inflammatory mimickers, mycosis fungoides vs. parapsoriasis and inflammatory mimickers, angiosarcoma vs granuloma faciale, and Kaposi sarcoma vs pseudo-Kaposi. CONCLUSIONS: Dermoscopy may be of aid in differentiating clinically similar inflammatory and neoplastic skin lesions.

Diagnosis of invasive cutaneous squamous cell carcinoma, imaging and staging.

The assessment of patients with a lesion raising the suspicion of an invasive cutaneous squamous cell carcinoma (cSCC) is a frequent clinical scenario. The management of patients with cSCC is a multistep approach, starting with the correct diagnosis. The two main diagnostic goals are to differentiate from other possible diagnoses and correctly recognize the lesion as cSCC, and then to determine the tumor spread (perform staging), that is if the patient has a common primary cSCC or a locally advanced cSCC, or a metastatic cSCC (with in-transit, regional lymph nodal, or rarely distant metastasis). The multistep diagnostic approach begins with the clinical characteristics of the primary cSCC, it is complemented with features with dermoscopy and, if available, reflectance confocal microscopy and is confirmed with histopathology. The tumor spread is assessed by physical examination and, in some cases, ultrasound and/or computed tomography or magnetic resonance imaging, mainly to investigate for regional lymph node metastasis or for local infiltration into deeper structures. In the last step, the clinical, histologic and radiologic findings are incorporated into staging systems.

A non-invasive approach to skin cancer diagnosis via graphene electrical tattoos and electrical impedance tomography.

Objective.Making up one of the largest shares of diagnosed cancers worldwide, skin cancer is also one of the most treatable. However, this is contingent upon early diagnosis and correct skin cancer-type differentiation. Currently, methods for early detection that are accurate, rapid, and non-invasive are limited. However, literature demonstrating the impedance differences between benign and malignant skin cancers, as well as between different types of skin cancer, show that methods based on impedance differentiation may be promising.Approach.In this work, we propose a novel approach to rapid and non-invasive skin cancer diagnosis that leverages the technologies of difference-based electrical impedance tomography (EIT) and graphene electronic tattoos (GETs).Main results.We demonstrate the feasibility of this first-of-its-kind system using both computational numerical and experimental skin phantom models. We considered variations in skin cancer lesion impedance, size, shape, and position relative to the electrodes and evaluated the impact of using individual and multi-electrode GET (mGET) arrays. The results demonstrate that this approach has the potential to differentiate based on lesion impedance, size, and position, but additional techniques are needed to determine shape.Significance.In this way, the system proposed in this work, which combines both EIT and GET technology, exhibits potential as an entirely non-invasive and rapid approach to skin cancer diagnosis.

Cutaneous, Subcutaneous, and Bilateral Adrenal Gland Metastases in Progressive Prostate Carcinoma: Theranostic Potential of 68 Ga-PSMA-11 PET/CT Imaging and 177 Lu-PSMA-617 Therapy.

ABSTRACT: Prostate carcinoma (PC) is the second most common malignant tumor in males globally. The metastatic spread of PC usually involves the pelvic and abdominal lymph nodes and the skeletal system. Cutaneous metastases are exceedingly uncommon and typically manifest themselves late in the disease course, considered as ominous sign with limited treatment options and a poor prognosis. We describe a patient wherein 68 Ga-PSMA-11 PET/CT detected multiple uncommon metastatic sites in the cutaneous region of the scrotum, penis, and thigh, as well as in the subcutaneous region of anterior abdominal wall, and in bilateral adrenal glands. These findings served as a theranostic tool for selecting 177 Lu-PSMA-617 treatment for these extremely rare metastatic sites.

Prostate-Specific Membrane Antigen-Avid Neurofibroma Mimicking Cutaneous Metastasis in Metastatic Castration-Resistant Prostate Cancer on 18 F-PSMA-1007 PET/CT.

ABSTRACT: The occurrence of cutaneous metastases in prostate cancer is exceedingly rare. Many benign lesions and nonprostatic cancers can express the prostate-specific membrane antigen (PSMA). They can potentially mimic metastasis of prostate cancer and lead to misinterpretation of PSMA PET/CT findings. Additionally, it has significant management and prognostic implications. We present a rare case of an 88-year-old man with metastatic castration-resistant prostate cancer who showed a PSMA-expressing subcutaneous nodule in the scalp on 18 F-PSMA-1007 PET/CT, raising the suspicion of cutaneous metastasis. However, its biopsy revealed a neurofibroma, altering the disease prognosis and management.

Conventional and contrast-enhanced ultrasound in the differential diagnosis of recurrent dermatofibrosarcoma protuberans and postoperative scar.

BACKGROUND: Dermatofibrosarcoma protuberans (DFSP) has a high recurrence rate after resection. Because of the lack of specific manifestations, recurrent DFSP is easily misdiagnosed as post-resection scar. A few series have reported ultrasound findings of recurrent DFSP; moreover, the usefulness of contrast-enhanced ultrasound in differentiating recurrent DFSP has not been studied. OBJECTIVE: We investigated conventional and contrast-enhanced ultrasound in the differential diagnosis of recurrent DFSP and post-resection scar. METHODS: We retrospectively evaluated the findings of conventional and contrast-enhanced ultrasound in 34 cases of recurrent DFSP and 38 postoperative scars examined between January 2018 and December 2022. RESULTS: The depth and vascular density of recurrent DFSP were greater than those of postoperative scars (P < 0.05). On gray-scale ultrasound, recurrent DFSP lesions were more commonly irregular, heterogeneous, and hypoechoic, with finger-like projections and ill-defined borders. Postoperative scar was more likely to appear as hypoechoic and homogeneous with well-defined borders (P < 0.05). On color Doppler ultrasound, recurrent DFSP was more likely to feature rich arterial and venous blood flow, and postoperative scar was more likely to display poor blood flow (P < 0.05). On contrast-enhanced ultrasound, recurrent DFSP was more likely to feature heterogeneous hyper-enhancement, and postoperative scar was more likely to display homogeneous iso-enhancement (P < 0.05). Recurrent DFSP presented a higher peak and sharpness than postoperative scar (P < 0.05). CONCLUSION: Conventional and contrast-enhanced ultrasound produced distinct features of recurrent DFSP and post-resection scar, which could improve the accuracy of differential diagnosis.

"Diffuse Hot Pleura" in a Case of Recurrent Squamous Cell Carcinoma of the Foot: 18 F-FDG PET/CT Findings.

ABSTRACT: Cutaneous squamous cell carcinoma (cSCC) is the second most common nonmelanoma skin cancer. Unlike basal cell carcinoma, regional lymph nodal metastases and subsequent distant site metastases are more common. Up to approximately 2% to 5% of cSCCs can result in distant metastases. Prognosis is dismal, and median survival is distinctly shortened in case of distant metastatic disease. Diffuse pleural metastases with distinctive overarching unilateral involvement are uncommon. Cutaneous SCC commonly metastasizes to lymph nodes, lungs, liver, bones, and skin. Diffuse unilateral pleural metastasis of cSCC of the foot is extremely rare. We report the case of a 54-year-old man with recurrent cSCC. On follow up restaging, 18 F-FDG PET/CT revealed diffuse nodular bipleural (visceral and parietal) hypermetabolic right pleural thickening, which was later biopsied and turned out to be diffuse pleural metastases from cSCC giving appearance of "hot pleura."

LSCS-Net: A lightweight skin cancer segmentation network with densely connected multi-rate atrous convolution.

The rising occurrence and notable public health consequences of skin cancer, especially of the most challenging form known as melanoma, have created an urgent demand for more advanced approaches to disease management. The integration of modern computer vision methods into clinical procedures offers the potential for enhancing the detection of skin cancer . The UNet model has gained prominence as a valuable tool for this objective, continuously evolving to tackle the difficulties associated with the inherent diversity of dermatological images. These challenges stem from diverse medical origins and are further complicated by variations in lighting, patient characteristics, and hair density. In this work, we present an innovative end-to-end trainable network crafted for the segmentation of skin cancer . This network comprises an encoder-decoder architecture, a novel feature extraction block, and a densely connected multi-rate Atrous convolution block. We evaluated the performance of the proposed lightweight skin cancer segmentation network (LSCS-Net) on three widely used benchmark datasets for skin lesion segmentation: ISIC 2016, ISIC 2017, and ISIC 2018. The generalization capabilities of LSCS-Net are testified by the excellent performance on breast cancer and thyroid nodule segmentation datasets. The empirical findings confirm that LSCS-net attains state-of-the-art results, as demonstrated by a significantly elevated Jaccard index.

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.

Performance of an automated total body mapping algorithm to detect melanocytic lesions of clinical relevance.

IMPORTANCE: Total body photography for skin cancer screening is a well-established tool allowing documentation and follow-up of the entire skin surface. Artificial intelligence-based systems are increasingly applied for automated lesion detection and diagnosis. DESIGN AND PATIENTS: In this prospective observational international multicentre study experienced dermatologists performed skin cancer screenings and identified clinically relevant melanocytic lesions (CRML, requiring biopsy or observation). Additionally, patients received 2D automated total body mapping (ATBM) with automated lesion detection (ATBM master, Fotofinder Systems GmbH). Primary endpoint was the percentage of CRML detected by the bodyscan software. Secondary endpoints included the percentage of correctly identified "new" and "changed" lesions during follow-up examinations. RESULTS: At baseline, dermatologists identified 1075 CRML in 236 patients and 999 CRML (92.9%) were also detected by the automated software. During follow-up examinations dermatologists identified 334 CRMLs in 55 patients, with 323 (96.7%) also being detected by ATBM with automated lesions detection. Moreover, all new (n = 13) or changed CRML (n = 24) during follow-up were detected by the software. Average time requirements per baseline examination was 14.1 min (95% CI [12.8-15.5]). Subgroup analysis of undetected lesions revealed either technical (e.g. covering by clothing, hair) or lesion-specific reasons (e.g. hypopigmentation, palmoplantar sites). CONCLUSIONS: ATBM with lesion detection software correctly detected the vast majority of CRML and new or changed CRML during follow-up examinations in a favourable amount of time. Our prospective international study underlines that automated lesion detection in TBP images is feasible, which is of relevance for developing AI-based skin cancer screenings.

Additional Findings of 18 F-AIF-FAPI-42 PET/CT in a Patient With Mycosis Fungoides-Type Cutaneous T-Cell Lymphoma : Comparisons With 18 F-FDG PET/CT.

ABSTRACT: A 44-year-old woman presented with extensive skin patches and pruritus persisting for 3 years. Histopathological examination of the skin from the right abdomen confirmed mycosis fungoides-type cutaneous T-cell lymphoma. Staging PET with 18 F-FDG PET/CT) showed increased uptake in the skin on the right abdomen and left hip. Subsequently 18 F-FAPI-42 PET/CT revealed additional foci of abnormal uptake on the skin of the chest and back.

The utility of ultrasound for the diagnosis and treatment of primary and metastatic melanoma - a remaining oncological challenge.

.