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.

SharpRazor: Automatic removal of hair and ruler marks from dermoscopy images.

BACKGROUND: The removal of hair and ruler marks is critical in handcrafted image analysis of dermoscopic skin lesions. No other dermoscopic artifacts cause more problems in segmentation and structure detection. PURPOSE: The aim of the work is to detect both white and black hair, artifacts and finally inpaint correctly the image. METHOD: We introduce a new algorithm: SharpRazor, to detect hair and ruler marks and remove them from the image. Our multiple-filter approach detects hairs of varying widths within varying backgrounds, while avoiding detection of vessels and bubbles. The proposed algorithm utilizes grayscale plane modification, hair enhancement, segmentation using tri-directional gradients, and multiple filters for hair of varying widths. We develop an alternate entropy-based processing adaptive thresholding method. White or light-colored hair, and ruler marks are detected separately and added to the final hair mask. A classifier removes noise objects. Finally, a new technique of inpainting is presented, and this is utilized to remove the detected object from the lesion image. RESULTS: The proposed algorithm is tested on two datasets, and compares with seven existing methods measuring accuracy, precision, recall, dice, and Jaccard scores. SharpRazor is shown to outperform existing methods. CONCLUSION: The Shaprazor techniques show the promise to reach the purpose of removing and inpaint both dark and white hair in a wide variety of lesions.

ChimeraNet: U-Net for Hair Detection in Dermoscopic Skin Lesion Images.

Hair and ruler mark structures in dermoscopic images are an obstacle preventing accurate image segmentation and detection of critical network features. Recognition and removal of hairs from images can be challenging, especially for hairs that are thin, overlapping, faded, or of similar color as skin or overlaid on a textured lesion. This paper proposes a novel deep learning (DL) technique to detect hair and ruler marks in skin lesion images. Our proposed ChimeraNet is an encoder-decoder architecture that employs pretrained EfficientNet in the encoder and squeeze-and-excitation residual (SERes) structures in the decoder. We applied this approach at multiple image sizes and evaluated it using the publicly available HAM10000 (ISIC2018 Task 3) skin lesion dataset. Our test results show that the largest image size (448 × 448) gave the highest accuracy of 98.23 and Jaccard index of 0.65 on the HAM10000 (ISIC 2018 Task 3) skin lesion dataset, exhibiting better performance than for two well-known deep learning approaches, U-Net and ResUNet-a. We found the Dice loss function to give the best results for all measures. Further evaluated on 25 additional test images, the technique yields state-of-the-art accuracy compared to 8 previously reported classical techniques. We conclude that the proposed ChimeraNet architecture may enable improved detection of fine image structures. Further application of DL techniques to detect dermoscopy structures is warranted.

Automated Cervical Digitized Histology Whole-Slide Image Analysis Toolbox.

BACKGROUND: Cervical intraepithelial neoplasia (CIN) is regarded as a potential precancerous state of the uterine cervix. Timely and appropriate early treatment of CIN can help reduce cervical cancer mortality. Accurate estimation of CIN grade correlated with human papillomavirus type, which is the primary cause of the disease, helps determine the patient's risk for developing the disease. Colposcopy is used to select women for biopsy. Expert pathologists examine the biopsied cervical epithelial tissue under a microscope. The examination can take a long time and is prone to error and often results in high inter-and intra-observer variability in outcomes. METHODOLOGY: We propose a novel image analysis toolbox that can automate CIN diagnosis using whole slide image (digitized biopsies) of cervical tissue samples. The toolbox is built as a four-step deep learning model that detects the epithelium regions, segments the detected epithelial portions, analyzes local vertical segment regions, and finally classifies each epithelium block with localized attention. We propose an epithelium detection network in this study and make use of our earlier research on epithelium segmentation and CIN classification to complete the design of the end-to-end CIN diagnosis toolbox. RESULTS: The results show that automated epithelium detection and segmentation for CIN classification yields comparable results to manually segmented epithelium CIN classification. CONCLUSION: This highlights the potential as a tool for automated digitized histology slide image analysis to assist expert pathologists.

EpithNet: Deep Regression for Epithelium Segmentation in Cervical Histology Images.

BACKGROUND: Automated pathology techniques for detecting cervical cancer at the premalignant stage have advantages for women in areas with limited medical resources. METHODS: This article presents EpithNet, a deep learning approach for the critical step of automated epithelium segmentation in digitized cervical histology images. EpithNet employs three regression networks of varying dimensions of image input blocks (patches) surrounding a given pixel, with all blocks at a fixed resolution, using varying network depth. RESULTS: The proposed model was evaluated on 311 digitized histology epithelial images and the results indicate that the technique maximizes region-based information to improve pixel-wise probability estimates. EpithNet-mc model, formed by intermediate concatenation of the convolutional layers of the three models, was observed to achieve 94% Jaccard index (intersection over union) which is 26.4% higher than the benchmark model. CONCLUSIONS: EpithNet yields better epithelial segmentation results than state-of-the-art benchmark methods.

DeepCIN: Attention-Based Cervical histology Image Classification with Sequential Feature Modeling for Pathologist-Level Accuracy.

BACKGROUND: Cervical cancer is one of the deadliest cancers affecting women globally. Cervical intraepithelial neoplasia (CIN) assessment using histopathological examination of cervical biopsy slides is subject to interobserver variability. Automated processing of digitized histopathology slides has the potential for more accurate classification for CIN grades from normal to increasing grades of pre-malignancy: CIN1, CIN2, and CIN3. METHODOLOGY: Cervix disease is generally understood to progress from the bottom (basement membrane) to the top of the epithelium. To model this relationship of disease severity to spatial distribution of abnormalities, we propose a network pipeline, DeepCIN, to analyze high-resolution epithelium images (manually extracted from whole-slide images) hierarchically by focusing on localized vertical regions and fusing this local information for determining Normal/CIN classification. The pipeline contains two classifier networks: (1) a cross-sectional, vertical segment-level sequence generator is trained using weak supervision to generate feature sequences from the vertical segments to preserve the bottom-to-top feature relationships in the epithelium image data and (2) an attention-based fusion network image-level classifier predicting the final CIN grade by merging vertical segment sequences. RESULTS: The model produces the CIN classification results and also determines the vertical segment contributions to CIN grade prediction. CONCLUSION: Experiments show that DeepCIN achieves pathologist-level CIN classification accuracy.

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.

Enhancements in localized classification for uterine cervical cancer digital histology image assessment.

BACKGROUND: In previous research, we introduced an automated, localized, fusion-based approach for classifying uterine cervix squamous epithelium into Normal, CIN1, CIN2, and CIN3 grades of cervical intraepithelial neoplasia (CIN) based on digitized histology image analysis. As part of the CIN assessment process, acellular and atypical cell concentration features were computed from vertical segment partitions of the epithelium region to quantize the relative distribution of nuclei. METHODS: Feature data was extracted from 610 individual segments from 61 images for epithelium classification into categories of Normal, CIN1, CIN2, and CIN3. The classification results were compared against CIN labels obtained from two pathologists who visually assessed abnormality in the digitized histology images. In this study, individual vertical segment CIN classification accuracy improvement is reported using the logistic regression classifier for an expanded data set of 118 histology images. RESULTS: We analyzed the effects on classification using the same pathologist labels for training and testing versus using one pathologist labels for training and the other for testing. Based on a leave-one-out approach for classifier training and testing, exact grade CIN accuracies of 81.29% and 88.98% were achieved for individual vertical segment and epithelium whole-image classification, respectively. CONCLUSIONS: The Logistic and Random Tree classifiers outperformed the benchmark SVM and LDA classifiers from previous research. The Logistic Regression classifier yielded an improvement of 10.17% in CIN Exact grade classification results based on CIN labels for training-testing for the individual vertical segments and the whole image from the same single expert over the baseline approach using the reduced features. Overall, the CIN classification rates tended to be higher using the training-testing labels for the same expert than for training labels from one expert and testing labels from the other expert. The Exact class fusion- based CIN discrimination results obtained in this study are similar to the Exact class expert agreement rate.

Fuzzy logic color detection: Blue areas in melanoma dermoscopy images.

Fuzzy logic image analysis techniques were used to analyze three shades of blue (lavender blue, light blue, and dark blue) in dermoscopic images for melanoma detection. A logistic regression model provided up to 82.7% accuracy for melanoma discrimination for 866 images. With a support vector machines (SVM) classifier, lower accuracy was obtained for individual shades (79.9-80.1%) compared with up to 81.4% accuracy with multiple shades. All fuzzy blue logic alpha cuts scored higher than the crisp case. Fuzzy logic techniques applied to multiple shades of blue can assist in melanoma detection. These vector-based fuzzy logic techniques can be extended to other image analysis problems involving multiple colors or color shades.

A fusion-based approach for uterine cervical cancer histology image classification.

Expert pathologists commonly perform visual interpretation of histology slides for cervix tissue abnormality diagnosis. We investigated an automated, localized, fusion-based approach for cervix histology image analysis for squamous epithelium classification into Normal, CIN1, CIN2, and CIN3 grades of cervical intraepithelial neoplasia (CIN). The epithelium image analysis approach includes medial axis determination, vertical segment partitioning as medial axis orthogonal cuts, individual vertical segment feature extraction and classification, and image-based classification using a voting scheme fusing the vertical segment CIN grades. Results using 61 images showed at least 15.5% CIN exact grade classification improvement using the localized vertical segment fusion versus global image features.

Modified watershed technique and post-processing for segmentation of skin lesions in dermoscopy images.

In previous research, a watershed-based algorithm was shown to be useful for automatic lesion segmentation in dermoscopy images, and was tested on a set of 100 benign and malignant melanoma images with the average of three sets of dermatologist-drawn borders used as the ground truth, resulting in an overall error of 15.98%. In this study, to reduce the border detection errors, a neural network classifier was utilized to improve the first-pass watershed segmentation; a novel "edge object value (EOV) threshold" method was used to remove large light blobs near the lesion boundary; and a noise removal procedure was applied to reduce the peninsula-shaped false-positive areas. As a result, an overall error of 11.09% was achieved.

An image feature-based approach to automatically find images for application to clinical decision support.

The illustrations in biomedical publications often provide useful information in aiding clinicians' decisions when full text searching is performed to find evidence in support of a clinical decision. In this research, image analysis and classification techniques are explored to automatically extract information for differentiating specific modalities to characterize illustrations in biomedical publications, which may assist in the evidence finding process. Global, histogram-based, and texture image illustration features were compared to basis function luminance histogram correlation features for modality-based discrimination over a set of 742 manually annotated images by modality (radiological, photo, etc.) selected from the 2004-2005 issues of the British Journal of Oral and Maxillofacial Surgery. Using a mean shifting supervised clustering technique, automatic modality-based discrimination results as high as 95.57% were obtained using the basis function features. These results compared favorably to other feature categories examined. The experimental results show that image-based features, particularly correlation-based features, can provide useful modality discrimination information.

Detection of granularity in dermoscopy images of malignant melanoma using color and texture features.

Granularity, also called peppering and multiple blue-grey dots, is defined as an accumulation of tiny, blue-grey granules in dermoscopy images. Granularity is most closely associated with a diagnosis of malignant melanoma. This study analyzes areas of granularity with color and texture measures to discriminate granularity in melanoma from similar areas in non-melanoma skin lesions. The granular areas in dermoscopy images of 74 melanomas and 14 melanomas in situ were identified and manually selected. For 200 non-melanoma dermoscopy images, those areas which most closely resembled granularity in color and texture were similarly selected. Ten texture and twenty-two color measures were studied. The texture measures consisted of the average and range of energy, inertia, correlation, inverse difference, and entropy. The color measures consisted of absolute and relative RGB averages, absolute and relative RGB chromaticity averages, absolute and relative G/B averages, CIE X, Y, Z, X/Y, X/Z and Y/Z averages, R variance, and luminance. These measures were calculated for each granular area of the melanomas and the comparable areas in the non-melanoma images. Receiver operating characteristic (ROC) curve analysis showed that the best separation of melanoma images from non-melanoma images by granular area features was obtained with a combination of color and texture measures. Comparison of ROC results showed greater separation of melanoma from benign lesions using relative color than using absolute color. Statistical analysis showed that the four most significant measures of granularity in melanoma are two color measures and two texture measures averaged over the spots: relative blue, relative green, texture correlation, and texture energy range. The best feature set, utilizing texture and relative color measures, achieved an accuracy of 96.4% based on area under the receiver operating characteristic curve.

Concentric decile segmentation of white and hypopigmented areas in dermoscopy images of skin lesions allows discrimination of malignant melanoma.

Dermoscopy, also known as dermatoscopy or epiluminescence microscopy (ELM), permits visualization of features of pigmented melanocytic neoplasms that are not discernable by examination with the naked eye. White areas, prominent in early malignant melanoma and melanoma in situ, contribute to early detection of these lesions. An adaptive detection method has been investigated to identify white and hypopigmented areas based on lesion histogram statistics. Using the Euclidean distance transform, the lesion is segmented in concentric deciles. Overlays of the white areas on the lesion deciles are determined. Calculated features of automatically detected white areas include lesion decile ratios, normalized number of white areas, absolute and relative size of largest white area, relative size of all white areas, and white area eccentricity, dispersion, and irregularity. Using a back-propagation neural network, the white area statistics yield over 95% diagnostic accuracy of melanomas from benign nevi. White and hypopigmented areas in melanomas tend to be central or paracentral. The four most powerful features on multivariate analysis are lesion decile ratios. Automatic detection of white and hypopigmented areas in melanoma can be accomplished using lesion statistics. A neural network can achieve good discrimination of melanomas from benign nevi using these areas. Lesion decile ratios are useful white area features.

Watershed segmentation of dermoscopy images using a watershed technique.

BACKGROUND/PURPOSE: Automatic lesion segmentation is an important part of computer-based image analysis of pigmented skin lesions. In this research, a watershed algorithm is developed and investigated for adequacy of skin lesion segmentation in dermoscopy images. METHODS: Hair, black border and vignette removal methods are introduced as preprocessing steps. The flooding variant of the watershed segmentation algorithm was implemented with novel features adapted to this domain. An outer bounding box, determined by a difference function derived from horizontal and vertical projection functions, is added to estimate the lesion area, and the lesion area error is reduced by a linear estimation function. As a post-processing step, a second-order B-Spline smoothing method is introduced to smooth the watershed border. RESULTS: Using the average of three sets of dermatologist-drawn borders as the ground truth, an overall error of 15.98% was obtained using the watershed technique. CONCLUSION: The implementation of the flooding variant of the watershed algorithm presented here allows satisfactory automatic segmentation of pigmented skin lesions.

Detection of atypical texture features in early malignant melanoma.

BACKGROUND: The presence of an atypical (irregular) pigment network (APN) can indicate a diagnosis of melanoma. This study sought to analyze the APN with texture measures. METHODS: For 106 dermoscopy images including 28 melanomas and 78 benign dysplastic nevi, the areas of APN were selected manually. Ten texture measures in the CVIPtools image analysis system were applied. RESULTS: Of the 10 texture measures used, correlation average provided the highest discrimination accuracy, an average of 95.4%. Discrimination of melanomas was optimal at a pixel distance of 20 for the 768 x 512 images, consistent with a melanocytic lesion texel size estimate of 4-5 texels per mm. CONCLUSION: Texture analysis, in particular correlation average at an optimized pixel spacing, may afford automatic detection of an irregular pigment network in early malignant melanoma.

Analysis of globule types in malignant melanoma.

OBJECTIVE: To identify and analyze subtypes of globules based on size, shape, network connectedness, pigmentation, and distribution to determine which globule types and globule distributions are most frequently associated with a diagnosis of malignant melanoma. DESIGN: Retrospective case series of dermoscopy images with globules. SETTING: Private dermatology practices. PARTICIPANTS: Patients in dermatology practices. Intervention Observation only. Main Outcome Measure Association of globule types with malignant melanoma. RESULTS: The presence of large globules (odds ratio [OR], 5.25) and globules varying in size (4.72) or shape (5.37) had the highest ORs for malignant melanoma among all globule types and combinations studied. Classical globules (dark, discrete, convex, and 0.10-0.20 mm) had a higher risk (OR, 4.20) than irregularly shaped globules (dark, discrete, and not generally convex) (2.89). Globules connected to other structures were not significant in the diagnosis of malignant melanoma. Of the different configurations studied, asymmetric clusters have the highest risk (OR, 3.02). CONCLUSIONS: The presence of globules of varying size or shape seems to be more associated with a diagnosis of malignant melanoma than any other globule type or distribution in this study. Large globules are of particular importance in the diagnosis of malignant melanoma.

Detection of basal cell carcinoma using color and histogram measures of semitranslucent areas.

BACKGROUND: Semitranslucency, defined as a smooth, jelly-like area with varied, near-skin-tone color, can indicate a diagnosis of basal cell carcinoma (BCC) with high specificity. This study sought to analyze potential areas of semitranslucency with histogram-derived texture and color measures to discriminate BCC from non-semitranslucent areas in non-BCC skin lesions. METHODS: For 210 dermoscopy images, the areas of semitranslucency in 42 BCCs and comparable areas of smoothness and color in 168 non-BCCs were selected manually. Six color measures and six texture measures were applied to the semitranslucent areas of the BCC and the comparable areas in the non-BCC images. RESULTS: Receiver operating characteristic (ROC) curve analysis showed that the texture measures alone provided greater separation of BCC from non-BCC than the color measures alone. Statistical analysis showed that the four most important measures of semitranslucency are three histogram measures: contrast, smoothness, and entropy, and one color measure: blue chromaticity. Smoothness is the single most important measure. The combined 12 measures achieved a diagnostic accuracy of 95.05% based on area under the ROC curve. CONCLUSION: Texture and color analysis measures, especially smoothness, may afford automatic detection of BCC images with semitranslucency.