Hyper-fusion network for semi-automatic segmentation of skin lesions

Segmentation of skin lesions is an important step for imaging-based clinical decision support systems. Automatic skin lesion segmentation methods based on fully convolutional networks (FCNs) are regarded as the state-of-the-art for accuracy. When there are, however, insufficient training data to cover all the variations in skin lesions, where lesions from different patients may have major differences in size/shape/texture, these methods failed to segment the lesions that have image characteristics, which are less common in the training datasets. FCN-based semi-automatic segmentation methods, which fuse user-inputs with high-level semantic image features derived from FCNs offer an ideal complement to overcome limitations of automatic segmentation methods. These semi-automatic methods rely on the automated state-of-the-art FCNs coupled with user-inputs for refinements, and therefore being able to tackle challenging skin lesions. However, there are a limited number of FCN-based semi-automatic segmentation methods and all these methods focused on 'early-fusion', where the first few convolutional layers are used to fuse image features and user-inputs and then derive fused image features for segmentation. For early-fusion based methods, because the user-input information can be lost after the first few convolutional layers, consequently, the user-input information will have limited guidance and constraint in segmenting the challenging skin lesions with inhomogeneous textures and fuzzy boundaries. Hence, in this work, we introduce a hyper-fusion network (HFN) to fuse the extracted user-inputs and image features over multiple stages. We separately extract complementary features which then allows for an iterative use of user-inputs along all the fusion stages to refine the segmentation. We evaluated our HFN on three well-established public benchmark datasets - ISBI Skin Lesion Challenge 2017, 2016 and PH2 - and our results show that the HFN is more accurate and generalizable than the state-of-the-art methods, in particular with challenging skin lesions.

Characterisation of black skin stratum corneum by digital macroscopic images analysis

Black skin medical images generally show very low contrast. Being in a global initiative of characterisation of black skin horny layer (stratum corneum) by digital images analysis, the authors in this study proposed a four-step approach. The first step consists of differentiation between probable healthy skin regions and those affected. For that, they used an automatic classification system based on multilayer perceptron artificial neural networks. The network has been trained with texture and colour features. Best features selection and network architecture definition were done using sequential network construction algorithm-based method. After classification, selected regions undergo a colour transformation, in order to increase the contrast with the lesion region. Thirdly, created colour information serves as the basis for a modified fuzzy c-mean clustering algorithm to perform segmentation. The proposed method, named neural network-based fuzzy clustering, was applied to many black skin lesion images and they obtained segmentation rates up to 94.67%. The last stage consists in calculating characteristics. Eight parameters are concerned: uniformity, standard deviation, skewness, kurtosis, smoothness, entropy, and average pixel values calculated for red and blue colour channels. All developed methods were tested with a database of 600 images and obtained results were discussed and compared with similar works.

The impact of patient clinical information on automated skin cancer detection

Skin cancer is one of the most common types of cancer worldwide. Over the past few years, different approaches have been proposed to deal with automated skin cancer detection. Nonetheless, most of them are based only on dermoscopic images and do not take into account the patient clinical information, an important clue towards clinical diagnosis. In this work, we present an approach to fill this gap. First, we introduce a new dataset composed of clinical images, collected using smartphones, and clinical data related to the patient. Next, we propose a straightforward method that includes an aggregation mechanism in well-known deep learning models to combine features from images and clinical data. Last, we carry out experiments to compare the models' performance with and without using this mechanism. The results present an improvement of approximately 7% in balanced accuracy when the aggregation method is applied. Overall, the impact of clinical data on models' performance is significant and shows the importance of including these features on automated skin cancer detection.

Segmentation of skin lesions in dermoscopy images using fuzzy classification of pixels and histogram thresholding

BACKGROUND AND OBJECTIVE

To ensure proper functioning of a Computer Aided Diagnosis (CAD) system for melanoma detection in dermoscopy images, it is important to accurately detect the border of the lesion. This paper proposes a method developed by the authors to address this problem.

METHODS

The algorithm for segmentation of skin lesions in dermoscopy images is based on fuzzy classification of pixels and subsequent histogram thresholding.

RESULTS

This method participated in the 2016 and 2017 ISBI (International Symposium on Biomedical Imaging) Challenges, hosted by the ISIC (International Skin Imaging Collaboration). It was tested against two public databases containing 379 and 600 images respectively, and compared using the same defined metrics (Accuracy, Dice Coefficient, Jaccard Index, Sensitivity and Specificity) with the rest of participating state-of-the-art work, obtaining good results: (0.934, 0.869, 0.791, 0.870 and 0.978) and (0.884, 0.760, 0.665, 0.869 and 0.923) respectively, ranking 9th and 15th out of a total of 21 and 28 participants respectively using the Jaccard Index (which was the indicator used as a basis for ranking) and the 1st in the 2017 Challenge using the Sensitivity.

CONCLUSION

The method has been proven to be robust and reliable. It's main contribution is the very design of the algorithm, highly innovative, which could also be used to deal with other segmentation problems of a similar nature.

Recognition of pigment network pattern in dermoscopy images based on fuzzy classification of pixels

BACKGROUND AND OBJECTIVE

One of the most relevant dermoscopic patterns is the pigment network. An innovative method of pattern recognition is presented for its detection in dermoscopy images.

METHODS

It consists of two steps. In the first one, by means of a supervised machine learning process and after performing the extraction of different colour and texture features, a fuzzy classification of pixels into the three categories present in the pattern's definition ("net", "hole" and "other") is carried out. This enables the three corresponding fuzzy sets to be created and, as a result, the three probability images that map them out are generated. In the second step, the pigment network pattern is characterised from a parameterisation process -derived from the system specification- and the subsequent extraction of different features calculated from the combinations of image masks extracted from the probability images, corresponding to the alpha-cuts obtained from the fuzzy sets.

RESULTS

The method was tested on a database of 875 images -by far the largest used in the state of the art to detect pigment network- extracted from a public Atlas of Dermoscopy, obtaining AUC results of 0.912 and 88%% accuracy, with 90.71%% sensitivity and 83.44%% specificity.

CONCLUSION

The main contribution of this method is the very design of the algorithm, highly innovative, which could also be used to deal with other pattern recognition problems of a similar nature. Other contributions are: 1. The good performance in discriminating between the pattern and the disturbing artefacts -which means that no prior preprocessing is required in this method- and between the pattern and other dermoscopic patterns; 2. It puts forward a new methodological approach for work of this kind, introducing the system specification as a required step prior to algorithm design and development, being this specification the basis for a required parameterisation -in the form of configurable parameters (with their value ranges) and set threshold values- of the algorithm and the subsequent conducting of the experiments.

Melanoma Is Skin Deep: A 3D Reconstruction Technique for Computerized Dermoscopic Skin Lesion Classification

Melanoma mortality rates are the highest amongst skin cancer patients. Melanoma is life threating when it grows beyond the dermis of the skin. Hence, depth is an important factor to diagnose melanoma. This paper introduces a non-invasive computerized dermoscopy system that considers the estimated depth of skin lesions for diagnosis. A 3-D skin lesion reconstruction technique using the estimated depth obtained from regular dermoscopic images is presented. On basis of the 3-D reconstruction, depth and 3-D shape features are extracted. In addition to 3-D features, regular color, texture, and 2-D shape features are also extracted. Feature extraction is critical to achieve accurate results. Apart from melanoma, in-situ melanoma the proposed system is designed to diagnose basal cell carcinoma, blue nevus, dermatofibroma, haemangioma, seborrhoeic keratosis, and normal mole lesions. For experimental evaluations, the PH2, ISIC: Melanoma Project, and ATLAS dermoscopy data sets is considered. Different feature set combinations is considered and performance is evaluated. Significant performance improvement is reported the post inclusion of estimated depth and 3-D features. The good classification scores of sensitivity = 96%, specificity = 97% on PH2 data set and sensitivity = 98%, specificity = 99% on the ATLAS data set is achieved. Experiments conducted to estimate tumor depth from 3-D lesion reconstruction is presented. Experimental results achieved prove that the proposed computerized dermoscopy system is efficient and can be used to diagnose varied skin lesion dermoscopy images.

Noninvasive Real-Time Automated Skin Lesion Analysis System for Melanoma Early Detection and Prevention

Melanoma spreads through metastasis, and therefore, it has been proved to be very fatal. Statistical evidence has revealed that the majority of deaths resulting from skin cancer are as a result of melanoma. Further investigations have shown that the survival rates in patients depend on the stage of the cancer; early detection and intervention of melanoma implicate higher chances of cure. Clinical diagnosis and prognosis of melanoma are challenging, since the processes are prone to misdiagnosis and inaccuracies due to doctors’ subjectivity. Malignant melanomas are asymmetrical, have irregular borders, notched edges, and color variations, so analyzing the shape, color, and texture of the skin lesion is important for the early detection and prevention of melanoma. This paper proposes the two major components of a noninvasive real-time automated skin lesion analysis system for the early detection and prevention of melanoma. The first component is a real-time alert to help users prevent skinburn caused by sunlight; a novel equation to compute the time for skin to burn is thereby introduced. The second component is an automated image analysis module, which contains image acquisition, hair detection and exclusion, lesion segmentation, feature extraction, and classification. The proposed system uses PH2 Dermoscopy image database from Pedro Hispano Hospital for the development and testing purposes. The image database contains a total of 200 dermoscopy images of lesions, including benign, atypical, and melanoma cases. The experimental results show that the proposed system is efficient, achieving classification of the benign, atypical, and melanoma images with accuracy of 96.3%, 95.7%, and 97.5%, respectively.

Model-based classification methods of global patterns in dermoscopic images

In this paper different model-based methods of classification of global patterns in dermoscopic images are proposed. Global patterns identification is included in the pattern analysis framework, the melanoma diagnosis method most used among dermatologists. The modeling is performed in two senses: first a dermoscopic image is modeled by a finite symmetric conditional Markov model applied to L∗a∗b∗ color space and the estimated parameters of this model are treated as features. In turn, the distribution of these features are supposed that follow different models along a lesion: a Gaussian model, a Gaussian mixture model, and a bag-of-features histogram model. For each case, the classification is carried out by an image retrieval approach with different distance metrics. The main objective is to classify a whole pigmented lesion into three possible patterns: globular, homogeneous, and reticular. An extensive evaluation of the performance of each method has been carried out on an image database extracted from a public Atlas of Dermoscopy. The best classification success rate is achieved by the Gaussian mixture model-based method with a 78.44% success rate in average. In a further evaluation the multicomponent pattern is analyzed obtaining a 72.91% success rate.

A cascade classifier for diagnosis of melanoma in clinical images

Computer aided diagnosis of medical images can help physicians in better detecting and early diagnosis of many symptoms and therefore reducing the mortality rate. Realization of an efficient mobile device for semi-automatic diagnosis of melanoma would greatly enhance the applicability of medical image classification scheme and make it useful in clinical contexts. In this paper, interactive object recognition methodology is adopted for border segmentation of clinical skin lesion images. In addition, performance of five classifiers, KNN, Naïve Bayes, multi-layer perceptron, random forest and SVM are compared based on color and texture features for discriminating melanoma from benign nevus. The results show that a sensitivity of 82.6% and specificity of 83% can be achieved using a single SVM classifier. However, a better classification performance was achieved using a proposed cascade classifier with the sensitivity of 83.06% and specificity of 90.05% when performing ten-fold cross validation.

Unified approach for lesion border detection based on mixture modeling and local entropy thresholding

BACKGROUND/PURPOSE

Computer-aided design (CAD) methods are highly valuable for the analysis of skin lesions using digital dermoscopy due to low rate of diagnostic accuracy of expert dermatologist. In computerized diagnostic methods, automatic border detection is the first and crucial step.

METHOD

In this study, a novel unified approach is proposed for automatic border detection (ABD). A preprocessing step is performed by normalized smoothing filter (NSF) to reduce background noise. Mixture models technique is then utilized to initially segment the lesion area roughly. Afterward, local entropy thresholding is performed to extract the lesion candidate pixels and the lesion border is smoothed using morphological reconstruction.

RESULTS

The overall ABD system is tested on a set of 100 dermoscopy images with ground truth. A comparative study was conducted with the other three state-of-the-art methods using statistical metrics. This ABD technique has the minimal average error probability rate of 5%, true detection of 92.10% and false positive rate of 6.41%.

CONCLUSION

Results demonstrate that the proposed method segments the lesion area accurately. Sample dataset and execute software are available online and can be downloaded from: http://cs.ntu.edu.pk/research.

Segmentation of dermoscopy images using wavelet networks

This paper introduces a new approach for the segmentation of skin lesions in dermoscopic images based on wavelet network (WN). The WN presented here is a member of fixed-grid WNs that is formed with no need of training. In this WN, after formation of wavelet lattice, determining shift and scale parameters of wavelets with two screening stage and selecting effective wavelets, orthogonal least squares algorithm is used to calculate the network weights and to optimize the network structure. The existence of two stages of screening increases globality of the wavelet lattice and provides a better estimation of the function especially for larger scales. R, G, and B values of a dermoscopy image are considered as the network inputs and the network structure formation. Then, the image is segmented and the skin lesions exact boundary is determined accordingly. The segmentation algorithm were applied to 30 dermoscopic images and evaluated with 11 different metrics, using the segmentation result obtained by a skilled pathologist as the ground truth. Experimental results show that our method acts more effectively in comparison with some modern techniques that have been successfully used in many medical imaging problems.

Lesion border detection in dermoscopy images using ensembles of thresholding methods

BACKGROUND

Dermoscopy is one of the major imaging modalities used in the diagnosis of melanoma and other pigmented skin lesions. Due to the difficulty and subjectivity of human interpretation, automated analysis of dermoscopy images has become an important research area. Border detection is often the first step in this analysis. In many cases, the lesion can be roughly separated from the background skin using a thresholding method applied to the blue channel. However, no single thresholding method appears to be robust enough to successfully handle the wide variety of dermoscopy images encountered in clinical practice.

METHODS

In this article, we present an automated method for detecting lesion borders in dermoscopy images using ensembles of thres holding methods.

CONCLUSION

Experiments on a difficult set of 90 images demonstrate that the proposed method is robust, fast, and accurate when compared to nine state-of-the-art methods.

Hair detection in dermoscopic images using percolation

The automatic analysis of dermoscopy images is often impaired by artifacts such as air bubbles, specular reflections or dark hair covering the skin lesions. Consequently, an important pre-processing step includes their detection and elimination. The most common and probably the most compromising of these artifacts is the presence of hair and therefore specific algorithms are required for its detection. This paper proposes a method for the detection of hair in dermoscopy images based on an efficient percolation algorithm for image processing recently proposed in [1]. The percolation algorithm locally processes image points by taking into account the intensity and connectivity of neighboring pixels. A cluster of connected points is thus obtained and the shape of this cluster is subsequently analyzed. If the cluster has a shape that is approximately linear then the image point is classified as hair. The performance of the proposed method was investigated on real dermoscopy images and compared with the DullRazor software [2]. Our results indicate that the method provides effective hair detection outperforming the DullRazor method by more than 10%, both in terms of false positive and false negative rates.

Conditional random fields and supervised learning in automated skin lesion diagnosis

Many subproblems in automated skin lesion diagnosis (ASLD) can be unified under a single generalization of assigning a label, from an predefined set, to each pixel in an image. We first formalize this generalization and then present two probabilistic models capable of solving it. The first model is based on independent pixel labeling using maximum a-posteriori (MAP) estimation. The second model is based on conditional random fields (CRFs), where dependencies between pixels are defined using a graph structure. Furthermore, we demonstrate how supervised learning and an appropriate training set can be used to automatically determine all model parameters. We evaluate both models' ability to segment a challenging dataset consisting of 116 images and compare our results to 5 previously published methods.