Automatic detection of basal cell carcinoma using telangiectasia analysis in dermoscopy skin lesion images

Basal Cell Carcinoma Diagnosis with Fusion of Deep Learning and Telangiectasia Features

In recent years, deep learning (DL) has been used extensively and successfully to diagnose different cancers in dermoscopic images. However, most approaches lack clinical inputs supported by dermatologists that could aid in higher accuracy and explainability. To dermatologists, the presence of telangiectasia, or narrow blood vessels that typically appear serpiginous or arborizing, is a critical indicator of basal cell carcinoma (BCC). Exploiting the feature information present in telangiectasia through a combination of DL-based techniques could create a pathway for both, improving DL results as well as aiding dermatologists in BCC diagnosis. This study demonstrates a novel "fusion" technique for BCC vs non-BCC classification using ensemble learning on a combination of (a) handcrafted features from semantically segmented telangiectasia (U-Net-based) and (b) deep learning features generated from whole lesion images (EfficientNet-B5-based). This fusion method achieves a binary classification accuracy of 97.2%, with a 1.3% improvement over the corresponding DL-only model, on a holdout test set of 395 images. An increase of 3.7% in sensitivity, 1.5% in specificity, and 1.5% in precision along with an AUC of 0.99 was also achieved. Metric improvements were demonstrated in three stages: (1) the addition of handcrafted telangiectasia features to deep learning features, (2) including areas near telangiectasia (surround areas), (3) discarding the noisy lower-importance features through feature importance. Another novel approach to feature finding with weak annotations through the examination of the surrounding areas of telangiectasia is offered in this study. The experimental results show state-of-the-art accuracy and precision in the diagnosis of BCC, compared to three benchmark techniques. Further exploration of deep learning techniques for individual dermoscopy feature detection is warranted.

Hybrid Topological Data Analysis and Deep Learning for Basal Cell Carcinoma Diagnosis

A critical clinical indicator for basal cell carcinoma (BCC) is the presence of telangiectasia (narrow, arborizing blood vessels) within the skin lesions. Many skin cancer imaging processes today exploit deep learning (DL) models for diagnosis, segmentation of features, and feature analysis. To extend automated diagnosis, recent computational intelligence research has also explored the field of Topological Data Analysis (TDA), a branch of mathematics that uses topology to extract meaningful information from highly complex data. This study combines TDA and DL with ensemble learning to create a hybrid TDA-DL BCC diagnostic model. Persistence homology (a TDA technique) is implemented to extract topological features from automatically segmented telangiectasia as well as skin lesions, and DL features are generated by fine-tuning a pre-trained EfficientNet-B5 model. The final hybrid TDA-DL model achieves state-of-the-art accuracy of 97.4% and an AUC of 0.995 on a holdout test of 395 skin lesions for BCC diagnosis. This study demonstrates that telangiectasia features improve BCC diagnosis, and TDA techniques hold the potential to improve DL performance.

Improving Automatic Melanoma Diagnosis Using Deep Learning-Based Segmentation of Irregular Networks

Deep learning has achieved significant success in malignant melanoma diagnosis. These diagnostic models are undergoing a transition into clinical use. However, with melanoma diagnostic accuracy in the range of ninety percent, a significant minority of melanomas are missed by deep learning. Many of the melanomas missed have irregular pigment networks visible using dermoscopy. This research presents an annotated irregular network database and develops a classification pipeline that fuses deep learning image-level results with conventional hand-crafted features from irregular pigment networks. We identified and annotated 487 unique dermoscopic melanoma lesions from images in the ISIC 2019 dermoscopic dataset to create a ground-truth irregular pigment network dataset. We trained multiple transfer learned segmentation models to detect irregular networks in this training set. A separate, mutually exclusive subset of the International Skin Imaging Collaboration (ISIC) 2019 dataset with 500 melanomas and 500 benign lesions was used for training and testing deep learning models for the binary classification of melanoma versus benign. The best segmentation model, U-Net++, generated irregular network masks on the 1000-image dataset. Other classical color, texture, and shape features were calculated for the irregular network areas. We achieved an increase in the recall of melanoma versus benign of 11% and in accuracy of 2% over DL-only models using conventional classifiers in a sequential pipeline based on the cascade generalization framework, with the highest increase in recall accompanying the use of the random forest algorithm. The proposed approach facilitates leveraging the strengths of both deep learning and conventional image processing techniques to improve the accuracy of melanoma diagnosis. Further research combining deep learning with conventional image processing on automatically detected dermoscopic features is warranted.

A deep learning approach to detect blood vessels in basal cell carcinoma

PURPOSE

Blood vessels called telangiectasia are visible in skin lesions with the aid of dermoscopy. Telangiectasia are a pivotal identifying feature of basal cell carcinoma. These vessels appear thready, serpiginous, and may also appear arborizing, that is, wide vessels branch into successively thinner vessels. Due to these intricacies, their detection is not an easy task, neither with manual annotation nor with computerized techniques. In this study, we automate the segmentation of telangiectasia in dermoscopic images with a deep learning U-Net approach.

METHODS

We apply a combination of image processing techniques and a deep learning-based U-Net approach to detect telangiectasia in digital basal cell carcinoma skin cancer images. We compare loss functions and optimize the performance by using a combination loss function to manage class imbalance of skin versus vessel pixels.

RESULTS

We establish a baseline method for pixel-based telangiectasia detection in skin cancer lesion images. An analysis and comparison for human observer variability in annotation is also presented.

CONCLUSION

Our approach yields Jaccard score within the variation of human observers as it addresses a new aspect of the rapidly evolving field of deep learning: automatic identification of cancer-specific structures. Further application of DL techniques to detect dermoscopic structures and handle noisy labels is warranted.

Assessing Skin Biopsy Rates for Histologic Findings Indicative of Nonpathological Cutaneous Disease

BACKGROUND

Recent increase in skin biopsies has been attributed to an epidemic of skin cancer. This may be avoidable, with potential savings.

OBJECTIVE

To determine whether the increase in skin biopsies is attributable to increasing frequency of biopsies associated with histology lacking pathological cutaneous disease. Pathological cutaneous disease was defined as (1) a malignancy, precancerous lesion, or lesion of uncertain behavior; or (2) disease symptomatic or associated with adverse quality of life impact.

PATIENTS AND METHODS

Retrospective cohort study, 2006 to 2013 of dermatology practice serving Florida and Ohio. Data were a consecutive sample of skin biopsies for diagnosis of dermatologic disease.

RESULTS

A total of 267,706 biopsies by an average of 52 providers per month from January 06 to December 13 were analyzed. Number of biopsies per visit increased 2% per year (RR: 1.02, CI: 1.00-1.04). Likelihood of biopsy associated with histology indicative of nonpathological cutaneous disease did not increase over time (OR: 0.99, CI: 0.95-1.03, p = .6302).

CONCLUSION

Rates of biopsies associated with nonpathological cutaneous disease is not increasing. Overall biopsy rates per visit have gradually increased; this seems attributable to greater rates of detection of pathological dermatologic disease.

Automated detection of nonmelanoma skin cancer using digital images: a systematic review

BACKGROUND

Computer-aided diagnosis of skin lesions is a growing area of research, but its application to nonmelanoma skin cancer (NMSC) is relatively under-studied. The purpose of this review is to synthesize the research that has been conducted on automated detection of NMSC using digital images and to assess the quality of evidence for the diagnostic accuracy of these technologies.

METHODS

Eight databases (PubMed, Google Scholar, Embase, IEEE Xplore, Web of Science, SpringerLink, ScienceDirect, and the ACM Digital Library) were searched to identify diagnostic studies of NMSC using image-based machine learning models. Two reviewers independently screened eligible articles. The level of evidence of each study was evaluated using a five tier rating system, and the applicability and risk of bias of each study was assessed using the Quality Assessment of Diagnostic Accuracy Studies tool.

RESULTS

Thirty-nine studies were reviewed. Twenty-four models were designed to detect basal cell carcinoma, two were designed to detect squamous cell carcinoma, and thirteen were designed to detect both. All studies were conducted in silico. The overall diagnostic accuracy of the classifiers, defined as concordance with histopathologic diagnosis, was high, with reported accuracies ranging from 72 to 100% and areas under the receiver operating characteristic curve ranging from 0.832 to 1. Most studies had substantial methodological limitations, but several were robustly designed and presented a high level of evidence.

CONCLUSION

Most studies of image-based NMSC classifiers report performance greater than or equal to the reported diagnostic accuracy of the average dermatologist, but relatively few studies have presented a high level of evidence. Clinical studies are needed to assess whether these technologies can feasibly be implemented as a real-time aid for clinical diagnosis of NMSC.

Vascular patterns in basal cell carcinoma: Dermoscopic, confocal and histopathological perspectives

Basal cell carcinoma (BCC) is the most prevalent skin cancer in the Caucasian population. A variety of different phenotypic presentations of BCC are possible. Although BCCs rarely metastasize, these tumors commonly destroy underlying tissues and should therefore be treated promptly. As vascular formation and angiogenesis are indicators of tumor development and progression, the presence of blood vessels, their morphology and architecture are important markers in skin lesions, providing critical information towards pathogenesis and diagnosis. BCC commonly lacks pigmentation, therefore it is important to emphasize the usefulness of vascular feature detection, recognition, quantification and interpretation. To answer the question of whether vascular patterns observed on dermoscopy, reflectance confocal microscopy (RCM) and histopathology might reflect the biologic behavior of BCCs, we undertook this review article. Several studies have sought, by various means, to identify vascular features associated with the more aggressive BCC phenotypes. Dermoscopic vascular pattern assessment can facilitate diagnostic discrimination between BCC subtypes, more aggressive BCCs displaying less or no pink coloration and a relative absence of central tumor vessels. RCM, a novel, non-invasive imaging technique, allows for the quantification of blood vessel size, density, and flow intensity in BCCs. BCCs are distinguished on RCM chiefly by vessels that branch and intertwine between neoplastic aggregates, a pattern strongly reflecting tumor neo-angiogenesis. The analysis of these vascular morphological and distribution patterns can provide further support in the diagnosis, assessment, or monitoring of BCCs. Histopathology shows significantly higher microvessel densities in the peritumoral stroma of BCCs, when compared to normal skin or benign tumors. This angiogenic response in the stroma is associated with local aggressiveness, therefore the quantification of peritumoralmicrovessels may further assist with tumor evaluation. How dermoscopy and RCM vascular patterns in BCC correlate with histopathological subtype and thus help in discriminating aggressive subtypes definitely deserves further investigation.

Deep Learning and Handcrafted Method Fusion: Higher Diagnostic Accuracy for Melanoma Dermoscopy Images

This paper presents an approach that combines conventional image processing with deep learning by fusing the features from the individual techniques. We hypothesize that the two techniques, with different error profiles, are synergistic. The conventional image processing arm uses three handcrafted biologically inspired image processing modules and one clinical information module. The image processing modules detect lesion features comparable to clinical dermoscopy information-atypical pigment network, color distribution, and blood vessels. The clinical module includes information submitted to the pathologist-patient age, gender, lesion location, size, and patient history. The deep learning arm utilizes knowledge transfer via a ResNet-50 network that is repurposed to predict the probability of melanoma classification. The classification scores of each individual module from both processing arms are then ensembled utilizing logistic regression to predict an overall melanoma probability. Using cross-validated results of melanoma classification measured by area under the receiver operator characteristic curve (AUC), classification accuracy of 0.94 was obtained for the fusion technique. In comparison, the ResNet-50 deep learning based classifier alone yields an AUC of 0.87 and conventional image processing based classifier yields an AUC of 0.90. Further study of fusion of conventional image processing techniques and deep learning is warranted.

Towards automatic detection of abnormal retinal capillaries in ultra-widefield-of-view retinal angiographic exams

Retinal capillary abnormalities include small, leaky, severely tortuous blood vessels that are associated with a variety of retinal pathologies. We present a prototype image-processing system for detecting abnormal retinal capillary regions in ultra-widefield-of-view (UWFOV) fluorescein angiography exams of the human retina. The algorithm takes as input an UWFOV FA frame and returns the candidate regions identified. An SVM classifier is trained on regions traced by expert ophthalmologists. Tests with a variety of feature sets indicate that edge features and allied properties differentiate best between normal and abnormal retinal capillary regions. Experiments with an initial set of images from patients showing branch retinal vein occlusion (BRVO) indicate promising area under the ROC curve of 0.950 and a weighted Cohen's Kappa value of 0.822.

Analysis of clinical and dermoscopic features for basal cell carcinoma neural network classification

BACKGROUND

Basal cell carcinoma (BCC) is the most commonly diagnosed cancer in the USA. In this research, we examine four different feature categories used for diagnostic decisions, including patient personal profile (patient age, gender, etc.), general exam (lesion size and location), common dermoscopic (blue-gray ovoids, leaf-structure dirt trails, etc.), and specific dermoscopic lesion (white/pink areas, semitranslucency, etc.). Specific dermoscopic features are more restricted versions of the common dermoscopic features.

METHODS

Combinations of the four feature categories are analyzed over a data set of 700 lesions, with 350 BCCs and 350 benign lesions, for lesion discrimination using neural network-based techniques, including evolving artificial neural networks (EANNs) and evolving artificial neural network ensembles.

RESULTS

Experiment results based on 10-fold cross validation for training and testing the different neural network-based techniques yielded an area under the receiver operating characteristic curve as high as 0.981 when all features were combined. The common dermoscopic lesion features generally yielded higher discrimination results than other individual feature categories.

CONCLUSIONS

Experimental results show that combining clinical and image information provides enhanced lesion discrimination capability over either information source separately. This research highlights the potential of data fusion as a model for the diagnostic process.

Automatic Detection of Arrow Annotation Overlays in Biomedical Images

Images in biomedical articles are often referenced for clinical decision support, educational purposes, and medical research. Authors-marked annotations such as text labels and symbols overlaid on these images are used to highlight regions of interest which are then referenced in the caption text or figure citations in the articles. Detecting and recognizing such symbols is valuable for improving biomedical information retrieval. In this research, image processing and computational intelligence methods are integrated for object segmentation and discrimination and applied to the problem of detecting arrows on these images. Evolving Artificial Neural Networks (EANNs) and Evolving Artificial Neural Network Ensembles (EANNEs) computational intelligence-based algorithms are developed to recognize overlays, specifically arrows, in medical images. For these discrimination techniques, EANNs use particle swarm optimization and genetic algorithm for artificial neural network (ANN) training, and EANNEs utilize the number of ANNs generated in an ensemble and negative correlation learning for neural network training based on averaging and Linear Vector Quantization (LVQ) winner-take-all approaches. Experiments performed on medical images from the imageCLEFmed’08 data set, yielded area under the receiver operating characteristic curve and precision/recall results as high as 0.988 and 0.928/0.973, respectively, using the EANNEs method with the winner-take-all approach.