Skin Imaging Research Articles

P12 A periorbital and upper back rash on a female patient of West African descent: a case of NXP2 dermatomyositis initially misdiagnosed as an allergic reaction

Abstract Introduction Idiopathic inflammatory myopathies (IIM) are a group of conditions characterised by immune-driven muscular inflammation, often in the presence of specific autoantibodies. NXP2 dermatomyositis is a subtype of IIM characterised by antibodies to nuclear matrix protein 2. NXP2 dermatomyositis is associated with calcinosis, especially in younger patients and with malignancy, especially in older patients. We report a case of NXP2 dermatomyositis in a female patient of West African descent presenting with a periorbital and upper back rash which was initially misdiagnosed as an allergic reaction. Images of darker skin are underrepresented in medical education around dermatomyositis. Case description A 49-year-old female office worker presented to A&E with a two-week history of arthralgia and fatigue. Over the past two days, she had also noticed the development of periorbital swelling and a rash on her eyelids and her upper back. On examination in A&E, the rash was considered minor and it was noted that she had difficulty taking off her coat due to pain in her arms. She was discharged home with a working diagnosis of an allergic reaction, with advice to see her GP if symptoms persisted. She was reviewed by her GP, who suspected a vasculitic rash. She was found to have a significantly raised ANCA-MPO antibody titre and was referred to the acute medical take where she was seen urgently by rheumatology on-call. On assessment, she was noted to have pain and morning stiffness in her hands and wrists, with swollen DIPJ, MCPJ and wrists on examination. On assessment by a rheumatologist aware of the skin manifestations of dermatomyositis, she was also noted to have a heliotrope rash, shawl sign, periungual erythema, and a V sign rash and was suspected to have dermatomyositis. Her CK was raised at 3000, and she was found to have NXP2 autoantibodies which confirmed the diagnosis. She was started on hydroxychloroquine whilst an MR scan of her arms and legs was organised, which revealed evidence of myositis in the proximal muscles. Following MR scanning, she was started on high dose corticosteroids. Azathioprine was subsequently initiated as a steroid sparing agent, but was not tolerated due to nausea and so she was switched to subcutaneous methotrexate. Her CK has now normalised, and her skin lesions are healing. Discussion One of the key issues highlighted by this case is the lack of awareness of rashes associated with dermatomyositis generally, and especially in diverse skin types. We suspect this is because of a lack of education and exposure to skin rashes in diverse skin types, especially in highly melanated skin types. On her initial assessment in A&E, her rash was misdiagnosed as an allergic reaction. We suspect a lack of familiarity with skin rashes in darker skin types contributed to this initial misdiagnosis. However, we believe it was clear that the patient’s rash and history on initial presentation were not in keeping with urticaria, or an allergic skin rash. However, a further point to consider is that dermatomyositis is rare, and so those working in primary care are less familiar with its manifestations. Medical resources have historically had a tendency to use images of people with lighter skin tones to demonstrate skin diseases. Over the last few years, several textbooks and online resources have become available specifically showing skin conditions in different skin tones, for example, the Mind the Gap handbook. These resources are generally excellent; however, they tend to focus on common skin disorders. In addition, images of dermatomyositis in medical textbooks are predominantly of white skin. If medical education resources were to include examples of dermatomyositis in a variety of skin tones it is possible that diagnostic confidence and proficiency would improve, and racial disparities would be reduced. This is especially important in a multicultural society like the United Kingdom. Key learning points Lack of awareness of skin rashes in diverse skin types contributes to misdiagnosis and worse outcomes for patients with ethnic minority backgrounds, including those with dermatomyositis. Increasing awareness and education surrounding the skin manifestations in diverse skin types is a key priority for medical education, including in rheumatology. New initiatives aimed at improving our understanding of skin manifestations in diverse skin types are underway.

Comparison of humans versus mobile phone-powered artificial intelligence for the diagnosis and management of pigmented skin cancer in secondary care: a multicentre, prospective, diagnostic, clinical trial

Diagnosis of skin cancer requires medical expertise, which is scarce. Mobile phone-powered artificial intelligence (AI) could aid diagnosis, but it is unclear how this technology performs in a clinical scenario. Our primary aim was to test in the clinic whether there was equivalence between AI algorithms and clinicians for the diagnosis and management of pigmented skin lesions. In this multicentre, prospective, diagnostic, clinical trial, we included specialist and novice clinicians and patients from two tertiary referral centres in Australia and Austria. Specialists had a specialist medical qualification related to diagnosing and managing pigmented skin lesions, whereas novices were dermatology junior doctors or registrars in trainee positions who had experience in examining and managing these lesions. Eligible patients were aged 18-99 years and had a modified Fitzpatrick I-III skin type; those in the diagnostic trial were undergoing routine excision or biopsy of one or more suspicious pigmented skin lesions bigger than 3 mm in the longest diameter, and those in the management trial had baseline total-body photographs taken within 1-4 years. We used two mobile phone-powered AI instruments incorporating a simple optical attachment: a new 7-class AI algorithm and the International Skin Imaging Collaboration (ISIC) AI algorithm, which was previously tested in a large online reader study. The reference standard for excised lesions in the diagnostic trial was histopathological examination; in the management trial, the reference standard was a descending hierarchy based on histopathological examination, comparison of baseline total-body photographs, digital monitoring, and telediagnosis. The main outcome of this study was to compare the accuracy of expert and novice diagnostic and management decisions with the two AI instruments. Possible decisions in the management trial were dismissal, biopsy, or 3-month monitoring. Decisions to monitor were considered equivalent to dismissal (scenario A) or biopsy of malignant lesions (scenario B). The trial was registered at the Australian New Zealand Clinical Trials Registry ACTRN12620000695909 (Universal trial number U1111-1251-8995). The diagnostic study included 172 suspicious pigmented lesions (84 malignant) from 124 patients and the management study included 5696 pigmented lesions (18 malignant) from the whole body of 66 high-risk patients. The diagnoses of the 7-class AI algorithm were equivalent to the specialists' diagnoses (absolute accuracy difference 1·2% [95% CI -6·9 to 9·2]) and significantly superior to the novices' ones (21·5% [13·1 to 30·0]). The diagnoses of the ISIC AI algorithm were significantly inferior to the specialists' diagnoses (-11·6% [-20·3 to -3·0]) but significantly superior to the novices' ones (8·7% [-0·5 to 18·0]). The best 7-class management AI was significantly inferior to specialists' management (absolute accuracy difference in correct management decision -0·5% [95% CI -0·7 to -0·2] in scenario A and -0·4% [-0·8 to -0·05] in scenario B). Compared with the novices' management, the 7-class management AI was significantly inferior (-0·4% [-0·6 to -0·2]) in scenario A but significantly superior (0·4% [0·0 to 0·9]) in scenario B. The mobile phone-powered AI technology is simple, practical, and accurate for the diagnosis of suspicious pigmented skin cancer in patients presenting to a specialist setting, although its usage for management decisions requires more careful execution. An AI algorithm that was superior in experimental studies was significantly inferior to specialists in a real-world scenario, suggesting that caution is needed when extrapolating results of experimental studies to clinical practice. MetaOptima Technology.

Classification of skin lesion using deep convolutional neural network by applying transfer learning

<p>The early and accurate diagnosis of skin cancer is crucial for improving patient outcomes and reducing the need for invasive biopsies. This study proposes a deep learning model for classifying skin malignancy using transfer learning and data augmentation techniques to address limitations observed in previous models and enhance diagnostic accuracy. The approach involves applying transfer learning to a pre-trained ResNet152 architecture using tensorflow and keras. Data augmentation techniques are employed on a dataset consisting of 10,015 skin lesion images obtained from the international skin imaging collaboration (ISIC) 2018 challenge, which encompasses diverse lesion types, sizes, and colors, posing a challenging classification task. Binary cross-entropy serves as the loss function, and the Adam optimizer is utilized for training the model. The results demonstrate a specificity of 87.42% and an F1 score of 0.854, outperforming other models in the field. These statistical findings highlight the effectiveness of transfer learning and data augmentation techniques in improving the accuracy of skin cancer diagnosis. The novelty of this study lies in the combination of transfer learning and data augmentation methods to enhance diagnostic accuracy. However, it is important to acknowledge the limitations of this study, including the necessity for further investigation to evaluate the clinical practicality of the model and address potential biases. Future research could explore the application of this model in a clinical setting and the development of models for detecting other types of skin lesions. In conclusion, the proposed deep learning model based on the ResNet152 architecture showcases promising results in the classification of skin lesions, demonstrating its potential for accurate skin cancer diagnosis. With further research and improvement, these models have the potential to revolutionize healthcare, improving patient outcomes, reducing healthcare costs, and increasing accessibility to screening and diagnosis, particularly for underserved populations.</p>

Analysis of Artificial Intelligence-Based Approaches Applied to Non-Invasive Imaging for Early Detection of Melanoma: A Systematic Review.

Melanoma, the deadliest form of skin cancer, poses a significant public health challenge worldwide. Early detection is crucial for improved patient outcomes. Non-invasive skin imaging techniques allow for improved diagnostic accuracy; however, their use is often limited due to the need for skilled practitioners trained to interpret images in a standardized fashion. Recent innovations in artificial intelligence (AI)-based techniques for skin lesion image interpretation show potential for the use of AI in the early detection of melanoma. The aim of this study was to evaluate the current state of AI-based techniques used in combination with non-invasive diagnostic imaging modalities including reflectance confocal microscopy (RCM), optical coherence tomography (OCT), and dermoscopy. We also aimed to determine whether the application of AI-based techniques can lead to improved diagnostic accuracy of melanoma. A systematic search was conducted via the Medline/PubMed, Cochrane, and Embase databases for eligible publications between 2018 and 2022. Screening methods adhered to the 2020 version of the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Included studies utilized AI-based algorithms for melanoma detection and directly addressed the review objectives. We retrieved 40 papers amongst the three databases. All studies directly comparing the performance of AI-based techniques with dermatologists reported the superior or equivalent performance of AI-based techniques in improving the detection of melanoma. In studies directly comparing algorithm performance on dermoscopy images to dermatologists, AI-based algorithms achieved a higher ROC (>80%) in the detection of melanoma. In these comparative studies using dermoscopic images, the mean algorithm sensitivity was 83.01% and the mean algorithm specificity was 85.58%. Studies evaluating machine learning in conjunction with OCT boasted accuracy of 95%, while studies evaluating RCM reported a mean accuracy rate of 82.72%. Our results demonstrate the robust potential of AI-based techniques to improve diagnostic accuracy and patient outcomes through the early identification of melanoma. Further studies are needed to assess the generalizability of these AI-based techniques across different populations and skin types, improve standardization in image processing, and further compare the performance of AI-based techniques with board-certified dermatologists to evaluate clinical applicability.

Image-based Skin Disease Detection and Classification through Bioinspired Machine Learning Approaches

A self-learning disease detection model will be useful for identifying skin infections in suspected individuals using skin images of infected patients. To detect skin diseases, some AI-based bioinspired models employ skin images. Skin infection is a common problem that is currently faced due to various reasons, such as food, water, environmental factors, and many others. Skin infections such as psoriasis, skin cancer, monkeypox, and tomato flu, among others, have a lower death rate but a significant impact on quality of life. Neural Networks (NNs) and Swarm intelligence (SI) based approaches are employed for skin disease diagnosis and classification through image processing. In this paper, the convolutional neural networks-based Cuckoo search algorithm (CNN-CS) is trained using the well-known multi-objective optimization technique cuckoo search. The performance of the suggested CNN-CS model is evaluated by comparing it with three commonly used metaheuristic-based classifiers: CNN-GA, CNN-BAT, and CNN-PSO. This comparison was based on various measures, including accuracy, precision, recall, and F1-score. These measures are calculated using the confusion matrices from the testing phase. The results of the experiments revealed that the proposed model has outperformed the others, achieving an accuracy of 97.72%.

Skin lesion image segmentation by using backchannel filling CNN and level sets

Pigmented skin lesion image segmentation is a key step in the computer-aided diagnosis of melanoma. This image segmentation task is also difficult due to blurred skin lesion boundaries, uneven color, and diverse textures. To improve the accuracy of skin lesion image segmentation, this paper proposes an image segmentation method that combines convolutional neural networks (CNN) and level sets. First, a backchannel filling (BCF) method is proposed. Through the first forward propagation of the CNN, the target position estimated by the CNN is backfilled into the position prior channel. Then, end-to-end deep learning models based on BCF multibranch encoding–decoding CNN (BCF-CNN) and level sets are built. The initial level set and driving energy are predicted through the second forward propagation of the BCF-CNN and then input into the level set segmentation model to complete the evolution. Additionally, a learning algorithm that combines a joint BCF-CNN and a level set is designed to train the segmentation model. The experimental results of the proposed algorithm are compared with those of other state-of-the-art algorithms on both the International Skin Imaging Collaboration (ISIC) 2017 and 2016 skin lesion challenge datasets, demonstrating that the proposed algorithm outperforms the others in terms of accuracy with a Jaccard index of approximately 77.0% and 85.1% on the ISIC 2017 and 2016 skin lesion challenge datasets, respectively.

Deep learning-based pigment analysis model trained with optical approach and ground truth assistance.

This study introduces an integrated training method combining the optical approach with ground truth for skin pigment analysis. Deep learning is increasingly applied to skin pigment analysis, primarily melanin and hemoglobin. While regression analysis is a widely used training method to predict ground truth-like outputs, the input image resolution is restricted by computational resources. The optical approach-based regression method can alleviate this problem, but compromises performance. We propose a strategy to overcome the limitation of image resolution while preserving performance by incorporating ground truth within the optical approach-based learning structure. The proposed model decomposes skin images into melanin, hemoglobin, and shading maps, reconstructing them by solving the forward problem with reference to the ground truth for pigments. Evaluation against the VISIA system, a professional diagnostic equipment, yields correlation coefficients of 0.978 for melanin and 0.975 for hemoglobin. Furthermore, our model can produce pigment-modified images for applications like simulating treatment effects.

Artificial intelligence-based PRO score assessment in actinic keratoses from LC-OCT imaging using Convolutional Neural Networks.

The histological PRO score (I-III) helps to assess the malignant potential of actinic keratoses (AK) by grading the dermal-epidermal junction (DEJ) undulation. Line-field confocal optical coherence tomography (LC-OCT) provides non-invasive real-time PRO score quantification. From LC-OCT imaging data, training of an artificial intelligence (AI), using Convolutional Neural Networks (CNNs) for automated PRO score quantification of AK in vivo may be achieved. CNNs were trained to segment LC-OCT images of healthy skin and AK. PRO score models were developed in accordance with the histopathological gold standard and trained on a subset of 237 LC-OCT AK images and tested on 76 images, comparing AI-computed PRO score to the imaging experts' visual consensus. Significant agreement was found in 57/76 (75%) cases. AI-automated grading correlated best with the visual score for PROII (84.8%) vs. PROIII (69.2%) vs. PROI (66.6%). Misinterpretation occurred in 25% of the cases mostly due to shadowing of the DEJ and disruptive features such as hair follicles. The findings suggest that CNNs are helpful for automated PRO score quantification in LC-OCT images. This may provide the clinician with a feasible tool for PRO score assessment in the follow-up of AK.

A Novel Approach of Animal Skin Classification Using CNN Model with CLAHE and SUCK Method Support

This study describes the process of classifying animal skin images which are rather difficult to obtain optimal image characteristics. For this reason, in the pre-processing stage, we propose two methods to support feature extraction: sharpening using a convolutional kernel (SUCK-Sharpening) and adaptive histogram equalization with limited contrast (CLAHE-Equalized). SUCK works by operating on these pixel values using direct math to build a new image; this final value is the new value of the current pixel. CLAHE overcomes the limitations of the global approach by performing local contrast enhancement. Because of the advantages of the two methods, it becomes a solution to get features processed at the feature extraction and classification stage. The process of animal skin imagery has characteristics in terms of shape and texture, including the characteristics of animal skin color. In this study, some experiments have been carried out on several CNN models, with an average classification accuracy of more than 70% using the sharpened and equalized methods on six animal skins. More detail, the average classification accuracy using 3 CNN models supported by two methods, namely Sharpening and Equalize on the CNN Resnet 50V2 model is 67.73% and 73.78%, InceptionV3 model at 82.13%, and 74.76% and Densenet121 models were 87.64% and 87.46 %. This research can be continued to improve the accuracy of other animal skin images, including determining fake or genuine skin images.

Efficient disentangled representation learning for multi-modal finger biometrics

Most multi-modal biometric systems use multiple devices to capture different traits and directly fuse multi-modal data while ignoring correlation information between modalities. In this paper, finger skin and finger vein images are acquired from the same region of the finger and therefore have a higher correlation. To represent data efficiently, we propose a novel Finger Disentangled Representation Learning Framework (FDRL-Net) that is based on a factorization concept, which disentangles each modality into shared and private features, thereby improving complementarity for better fusion and extracting modality-invariant features for heterogeneous recognition. Besides, to capture as much finger texture as possible, we utilize three-view finger images to reconstruct full-view multi-spectral finger traits, which increases the identity information and the robustness to finger posture variation. Finally, a Boat-Trackers-based multi-task distillation method is proposed to migrate the feature representation ability to a lightweight multi-task network. Extensive experiments on six single-view multi-spectral finger datasets and two full-view multi-spectral finger datasets demonstrate the effectiveness of our approach.

Diagnostic Accuracy of Hyperpigmented Microcircles in Dermatoscopy of Non-Facial Non-Acral Melanomas: A Pilot Retrospective Study using a Public Image Database

Background: Grey perifollicular circles are a dermatoscopic clue to melanoma, especially in facial skin. So far, no other adnexal clues than follicular have been investigated in this diagnosis. Objectives: The study aimed to analyse the prevalence of hyperpigmented periadnexal microcircles (HMs) in melanoma and its common simulators at non-facial non-acral sites, analyse the relation between the presence of HM, regression and hypopigmentation, and evaluate the diagnostic accuracy of this structure in melanoma. Methods: International Skin Imaging Collaboration (69,445 images) was searched in April 2020 for the pathology-confirmed dermatoscopic images with metadata including sex, age bin, and declared non-acral non-facial anatomic site. The final study sample (5,408 images, 1,326 of which were melanomas) was evaluated by expert dermatoscopist blinded to the diagnosis and labelled for the presence of ≥3 HM distributed centrally (cHM) or peripherally (pHM), hypopigmentation, and classic dermatoscopic regression structures. A subset of 40 images was labelled by 7 raters (2 residents, 5 experts) to assess interobserver agreement. We compared the presence of pHM with the presence of regression as well as performed a set of independent χ2 tests to evaluate the discriminatory power and its fragility. Performance of the models was assessed using measures of discrimination and calibration. Results: HM were significantly more prevalent in melanomas than in non-melanomas and nevi. Fair/good interobserver agreement for HM was reached for all the raters, and moderate/good for experts only (single rater/average, respectively). Regardless of regression/hypopigmentation status, pHM were significantly more common in melanoma than in non-melanomas or nevi and were observed significantly more often in melanomas on sun-damaged skin (upper extremity, posterior torso). No significant differences between the groups were found for cHM. pHM proved a high odds ratio in the tests as to the classical indicators such as classic dermatoscopic regression structures. Conclusion: pHM could be considered a novel dermatoscopic clue to melanoma.

Skin cancer classification using explainable artificial intelligence on pre-extracted image features

Skin cancer is the most common type of cancer worldwide, affecting a large population recently. To date, various machine learning techniques exploiting skin images have been applied directly to skin cancer classification, showing promising results in improving diagnostic accuracy. This study aims to develop a machine learning-based model capable of accurately classifying skin cancer by utilizing extracted features from preprocessed images in the publicly available PH² dataset. Preprocessed features are known to provide more significant information than raw image data, as they capture specific characteristics of the images that are relevant to the classification task. The proposed model of this study can identify the most pertinent information in the images more accurately, thereby improving the performance and interpretability of the machine learning classification. Our simulation results illustrate that employing XG-boost yields an accuracy of 94% and an area under the curve value of 0.9947, further indicating that the proposed technique effectively distinguishes between non-melanoma and melanoma skin cancer. Explainable artificial intelligence provides some explanations by leveraging model-agnostic methods such as partial dependence plot, permutation importance, and SHAP. Moreover, the explainable artificial intelligence results show that asymmetry and pigment network features are the most important feature in the classification of skin cancer. These specific characteristics emerge as the most influential factors in distinguishing between different types of skin cancer.

Fractional WSD: Fractional war strategy dingo optimization with unified segmentation for detection of skin cancer

Skin cancer represents a general cancer type. Earlier treatment and diagnosis are important tasks in skin cancer. The models, like deep learning, are generally utilized for diagnosing skin lesions and aid in precise treatment. This paper devises a new optimized deep model for discovering skin cancer with a group of skin images. The model utilized pre-processing with a Gaussian filter to filter out the noise and include contrast enhancement to improve image quality. Then segmentation process is done and it is accomplished by fusing two networks, namely Pyramid Based Multi-Scale Encoder-Decoder Network (PMED-Net) and U-Net. The weight optimization takes place using Dingo War Strategy optimization (DWSO). Various operations, like flipping, brightness, shifting, and random erasing are performed for data augmentation to increase dimensionality. The features are extracted to acquire the imperative information from the pixels. The Deep Convolution Neural Network (DCNN) is adapted finally for discovering skin cancer and is subsequently trained with Fractional DWSO. The proposed Fractional DWSO-based DCNN outperformed with an utmost testing accuracy of 95.8 %, sensitivity of 96.9 %, and specificity of 96.6 % respectively.

A Prospective Study Demonstrating Early Prediction of Skin Toxicity From Radiation Therapy Using Radiomic Features From Optical and Infrared Images

Approximately 90% of patients undergoing breast cancer radiation therapyexperience skin toxicities that are difficult to classify and predict ahead of time. A prediction oftoxicityat the early stages of the treatment would provide clinicians with a prompt to intervene. The objectives of this study were to evaluate the correlation between skin toxicity and radiomic features extracted from optical and infrared (thermal) images of skin, and to develop a model for predicting a patient's skin response to radiation. Optical and infrared breast and chest-wall images were acquired daily during the course of radiation therapy, as well as weekly for 3 weeks after the end of treatment for 20 patients with breast cancer. Skin-toxicity assessments were conducted weekly until the patients' finalvisit. Skin color and temperature trends from histogram-based and texture-based radiomic features, extracted from the treatment area, were analyzed, reduced, and used in a cross-validation machine learning model to predict the patients' skin toxicity grades. A set of 9 independent color and temperature features with significant correlation to skin toxicitywere identified from 108 features. The cross-validation accuracy of a cubic Support Vector Machine remained >85% and area under the receiver operating characteristic curve remained>0.75,when reducing the input imaging data to include onlythe sessions with a biologically effective dosenot exceeding30 Gy(approximately the first third to first half of the total treatment dose). The quantitative analysis of radiomic featuresextracted from opticaland infrared (thermal) images of skinwas shown to be promising for predicting skin toxicities.

Pact-Net: Parallel CNNs and Transformers for medical image segmentation

Background and ObjectiveThe image segmentation of diseases can help clinical diagnosis and treatment in medical image analysis. Because medical images usually have low contrast and large changes in the size and shape of some structures, this will lead to over-segmentation and under-segmentation. These problems are particularly evident in the segmentation of skin damage. The blurring of the boundary in skin images and the specificity of patients will further increase the difficulty of skin lesion segmentation. Currently, most researchers use deep learning networks to solve these skin segmentation problems. However, traditional convolution methods often fail to obtain satisfactory segmentation performance due to their shortcomings in obtaining global features. Recently, Transformers with good global information extraction ability has achieved satisfactory results in computer vision, which brings new solutions to optimize the model of medical image segmentation further. MethodsTo extract more features related to medical image segmentation and effectively use features to further optimize the skin image segmentation model, we designed a network that combines CNNs and Transformers to improve local and global features, called Parallel CNNs and Transformers for Medical Image Segmentation (Pact-Net). Specifically, due to the advantages of Transformers in extracting global information, we create a novel fusion module CSMF, which uses channel and spatial attention mechanism and multi-scale mechanism to effectively fuse the global information extracted by Transformers into the local features extracted by CNNs. Therefore, our Pact-Net dual-branch runs in parallel to effectively capture global and local information. ResultsOur Pact-Net exceeds the models submitted on the three datasets ISIC 2016, ISIC 2017 and ISIC 2018, and the indicators required for the datasets reach 86.95%, 79.31% and 84.14%, respectively. We also conducted medical image segmentation experiments on cell and polyp datasets to evaluate the robustness, learning and generalization ability of the network. The ablation study of each part of Pact-Net proves the validity of each component, and the comparison with state-of-the-art methods on different indicators proves the predominance of the network. ConclusionsThis paper uses the advantages of CNNs and Transformers in extracting local and global features, and further integrates features for skin lesion segmentation. Compared with the state-of-the-art methods, Pact-Net can achieve the most advanced segmentation ability on the skin lesion segmentation dataset, which can help doctors diagnose and treat diseases.

44149 Visible improvements in skin pigmentation and redness with topical bakuchiol-containing sunscreen using noninvasive clinical skin imaging technique in photodamaged facial skin

44149 Visible improvements in skin pigmentation and redness with topical bakuchiol-containing sunscreen using noninvasive clinical skin imaging technique in photodamaged facial skin

Image Quality Assessment Using a Convolutional Neural Network for Clinical Skin Images

Background The quality of the images received for teledermatology evaluation is often suboptimal, with up to 50% of patients providing images that are poorly lit, off-center, or blurry. To ensure a similar level of care to in-person consultations, high-quality images are essential. Objective The aim of this study is to develop an image quality analysis tool to assess patient- and primary care physician (PCP)–derived images using a deep learning model leveraging multiple instance learning and ordinal regression for model predictions. Methods The data set used for this study was acquired from patient-derived images submitted to the Department of Dermatology, Duke University, between August 21, 2018, and December 31, 2019, and PCP-derived images between March 1, 2021, and June 30, 2022. Seven dermatology faculty members with a designation of professor, associate professor, and assistant professor evaluated 400 images each, and 2 dermatology residents evaluated 400 images, assuring that each image had 4 different quality labels. We used a pretrained model VGG16 architecture, further fine-tuned by updating weights based on the input data. The images were taken with cell phones (patients) or cameras (PCPs) in RGB scale, with the resolution being 76 pixels per inch for both height and width, and the average pixel size of the image being 2840×2793 (SD 986×983; 1471 inch2, SD 707 inch2). The optimal threshold was determined using the Youden index, which represents the best trade-off between sensitivity and specificity and balance the number of true positives and true negatives in the classification results. Once the model predicts the rank, the ordinal labels are transformed to binary labels by using a majority vote as the goal is to distinguish between 2 distinct categories (good vs bad quality) and not predict quality as a continuous variable. Results Based on the Youden index, we achieved a positive predicted value of 0.906, implying that the model will predict 90% of the good-quality images as such, while 10% of the poor-quality images are predicted as being of good quality to enhance clinical utility, with an area under the receiver operating characteristic curve (AUC) for the test set at 0.885 (95% CI 0.838-0.933) and sensitivity, specificity, and negative predictive value (NPV) of 0.829, 0.784, and 0.645, respectively. Further evaluation on independent validation consisting of 300 images from patients and 150 images from PCPs revealed AUCs of 0.864 (95% CI 0.818-0.909) and 0.902 (95% CI 0.85-0.95), respectively. The sensitivity, specificity, positive predicted value, and NPV for the 300 images were 0.827, 0.800, 0.959, and 0.450, respectively. Conclusions This study shows a practical approach to improve image quality for clinical decision-making. While patients and PCPs may have to capture additional images (due to lower NPV), this is offset by the reduced workload and improved efficiency of clinical teams due to the receipt of higher-quality images. Additional images can also be useful if all images (good or poor) are transmitted to medical records. Future studies need to focus on real-time clinical validation of our results. Conflicts of Interest None declared.

A Hybrid Model for Classification of Skin Cancer Images After Segmentation

For dermatoscopic skin lesion images, deep learning-based algorithms, particularly convolutional neural networks (CNN), have demonstrated good classification and segmentation capabilities. The impact of utilizing lesion segmentation data on classification performance, however, is still up for being subject to discussion. Being driven in this direction, in this work we propose a hybrid deep learning-based model to classify the skin cancer using segmented images. In the first stage, a fully convolutional encoder–decoder network (FCEDN) is employed to segment the skin cancer image and then in the second phase, a CNN is applied on the segmented images for classification. As the model’s success depends on the hyper-parameters it uses and fine-tuning these hyper-parameters by hand is time-consuming, so in this study the hyper-parameters of the hybrid model are optimized by utilizing an exponential neighborhood gray wolf optimization (ENGWO) technique. Extensive experiments are carried out using the International Skin Imaging Collaboration (ISIC) 2016 and ISIC 2017 datasets to show the efficacy of the model. The suggested model has been evaluated on both balanced and unbalanced datasets. With the balanced dataset, the proposed hybrid model achieves training accuracy up to 99.98%, validation accuracy up to 92.13% and testing accuracy up to 89.75%. It is evident from the findings that the proposed hybrid model outperforms previous known models in a competitive manner over balanced data.

The role of imaging technologies in the diagnosis of hidradenitis suppurativa

Hidradenitis suppurativa is a chronic, inflammatory, recurrent, and debilitating disease of the hair follicle. It presents with painful, deep-seated, inflamed lesions, such as nodules, abscesses, sinus tracts, and fistulas, generally located in the main folds. Clinical severity assessment alone can be reductive; noninvasive skin imaging technologies, such as ultrasound, magnetic resonance imaging, medical infrared thermography, computed tomography, and positron emission tomography, provide subclinical anatomical and functional details. These instrumental techniques confirm the clinical suspect, thus allowing an earlier diagnosis and improving patients’ clinical evaluation, staging, and management. Finally, they might be helpful for preoperative mapping. In this contribution, we provide an overview of the current knowledge about noninvasive skin imaging techniques with a particular focus on ultrasonography, which is widely used thanks to its precision, versatility, and availability.

Image Segmentation Based Automated Skin Cancer Detection Technique

Skin cancer is a prevalent and deadly disease that affects millions of people worldwide. Early detection and diagnosis of skin cancer can significantly improve the chances of successful treatment and recovery. This study proposes a skin cancer segmentation and detection system using image processing and deep learning techniques to automate the diagnosis process. The system is trained on a dataset of skin images and uses a deep learning algorithm to classify skin lesions as benign or malignant. The performance of the system is evaluated using various metrics, including accuracy, precision, recall, and F1 score. The results show that the proposed system achieves high accuracy in detecting and classifying skin lesions as benign or malignant. Additionally, the proposed system is compared with other state-of-the-art methods, and it is found that the proposed system outperforms them in terms of accuracy and speed. The study contributes to the advancement of deep learning and image-processing techniques for medical diagnosis and detection. The proposed system can have significant implications in improving the accuracy and speed of skin cancer diagnosis, thereby improving the chances of successful treatment and recovery.