- Access throughyour institution
- Section snippets
- References (41)
- Cited by (11)
- Recommended articles (6)
Computers in Biology and Medicine
1 March 2016
, Pages 23-39
Author links open overlay panel
Differential diagnosis of squamous cell carcinoma in situ is of great importance for prognosis and decision making in the disease treatment procedure. Currently, differential diagnosis is done by pathologists based on examination of the histopathological slides under the microscope, which is time consuming and prone to inter and intra observer variability. In this paper, we have proposed an automated method for differential diagnosis of SCC in situ from actinic keratosis, which is known to be a precursor of squamous cell carcinoma. The process begins with epidermis segmentation and cornified layer removal. Then, epidermis axis is specified using the paths in its skeleton and the granular layer is removed via connected components analysis. Finally, diagnosis is done based on the classification result of intensity profiles extracted from lines perpendicular to the epidermis axis. The results of the study are in agreement with the gold standards provided by expert pathologists.
Cutaneous squamous cell carcinoma (SCC) is the second most common form of skin cancer, constituting about 20% of all reported cases of non-melanoma skin malignancies , . It arises from uncontrolled growth of squamous cells in the upper layer of the skin tissue and has the potential to metastasize to other organs of the body if left untreated. Squamous cell carcinoma in situ, also called intra-epidermal SCC or Bowen disease, is the first stage of SCC. “In situ” means that the malignant cells are still only in the epidermis and have not invaded deeper into the dermis .
Differential diagnosis is the procedure of distinguishing a particular disease from others that present similar clinical signs, e.g. differentiating between prostate cancer and benign prostatic hyperplasia. In the case of SCC in situ, there are number of skin lesions that should be considered in differential diagnosis, among which, actinic keratosis and basal cell carcinoma (BCC) are more remarkable .
Differential diagnosis of SCC in situ versus actinic keratosis is not a trivial task as both share similar clinical and histological features. Currently, differential diagnosis is done by pathologists based on examination of the histopathological slides under the microscope, which is laborious, time consuming and prone to inter and intra observer variability. Moreover, there are relatively few expert pathologists against the large number of tissue samples to be investigated. So, automated systems for analyzing the histopathological slides are desirable.
Actinic keratosis is commonly believed to be a precursor of squamous cell carcinoma in situ and is referred to as incipient SCC by some authors , because it can progress to invasive squamous cell carcinoma if left untreated. Hence, in this work, the problem of distinguishing SCC in situ from actinic keratosis is addressed.
Skin is composed of three primary layers: epidermis, dermis and hypodermis (subcutaneous adipose layer). Epidermis is divided into three layers: Malpighian (basal and squamous) layer, granular layer and cornified layer. Fig. 1a shows a histopathological image of normal skin in which the two main layers, epidermis and dermis are indicated. Cells existing in basal and squamous layer are called keratinocyte.
In pathology, SCC in situ is recognized by the presence of dysplastic keratinocytes which cover the full-thickness of epidermis and in advanced stages, invade the dermis . Dysplastic keratinocytes are immature cells with a relatively high nucleus to cytoplasm (N:C) ratio, the ratio which indicates the maturity of a cell . As a cell matures, this ratio generally decreases. Unlike healthy cells, cancerous cells reproduce very quickly and do not have a chance to mature. Other pathological cues to rule out SCC in situ are cells polymorphism, hyperkeratosis (thickened cornified layer), parakeratosis (presence of keratinocytes nucleus in cornified layer) and increase in epidermis layer thickness. Actinic keratosis is also characterized by the above mentioned cues specially dysplastic keratinocytes in epidermis layer, but it differs from SCC in situ in the way that abnormal cells present mainly in lower third of the epidermis , .
Fig. 1b and c shows, respectively, skin histopathological images of actinic keratosis and SCC in situ. In contrast to the normal case of Fig. 1a in which immature cells just cover a relatively thin layer at the bottom of the epidermis (basal layer), in actinic keratosis and SCC in situ cases, they cover the lower one-third and full thickness of the epidermis, respectively. Note also the atypia which presents among keratinocytes in the last two cases, in the sense that they are more polymorphic and of different sizes. White bracket indicates the area with atypical keratinocytes.
It should be mentioned that Fig. 1b and c are representative samples of actinic keratosis and SCC in situ which exhibit almost distinct appearances in texture of the epidermis. However, there are borderline cases of actinic keratosis in which dysplastic keratinocytes have covered, to some extent, the middle third part of the epidermis as well. Fig. 1d shows an example of this kind. This case is labeled as actinic keratosis by expert pathologists.
Although parakeratosis is usually less pronounced in actinic keratosis, but there are cases in which it is relatively more severe, resembling the appearance of SCC in situ. Consequently, parakeratosis cannot be used alone as a cue for differential diagnosis. Also note that from the images in Fig. 1, it can be observed that the cells in epidermis not only are overlapping and non homogenous, but also they are too cluttered to enable simple segmentation of individual cells, especially in the SCC in situ case.
The organization of this paper is as follows. A brief review of the works in the field of histopathological images analysis is presented in Section 2. The proposed technique is described in Section 3; results are demonstrated and discussed in Section 4, followed by conclusion in Section 5.
Histopathological images processing is a popular research field the importance of which has rapidly grown over the years due to increasing incidence of malignant diseases worldwide. There are many works in the literature which have covered the processing of histopathological images from various tissues. Cervical tissue is analyzed in , , ,  to diagnose and grade cervical intraepithelial neoplasia (CIN). The automated method proposed in  consists of two stages. First, the input
For conducting this study, 30 histopathological images were obtained from skin tissue samples of unknown patients with the age between 40 and 60, sixteen images of SCC in situ and fourteen images of actinic keratosis. After sectioning the tissue samples at a thickness of about 5µm and staining using Hematoxylin and Eosin (H&E), high resolution images were digitally captured by Aperio ScanScope slide scanner from histopathology slides at 60× magnifications with pixel size of 0.28µm. From these
Results and discussion
Performance of the proposed method was evaluated in two parts: classification of samples and tumor area segmentation.
Conclusion and future works
In this paper, a method for differential diagnosis of squamous cell carcinoma in situ from actinic keratosis is proposed. After epidermis segmentation and cornified layer removal, epidermis axis is specified using the paths in its skeleton and granular layer is removed via connected components analysis. Finally, diagnosis is done based on the classification result of intensity profiles extracted from lines perpendicular to the epidermis axis. Evaluation results have shown that our method is
Conflict of interest statement
The authors would like to thank Dr. Maryam Almassi, M.D., A.P.C.P, from Vanak Pathobiology Laboratory (Tehran, Iran) for insightful advices and helping us in medical aspects of this research.
- M. Musselman et al.Time–frequency distributions in the classification of epilepsy from EEG signals
J. Expert. Syst. Appl.
- N.M. Shakya et al.Discrimination of squamous cell carcinoma in situ from seborrheic keratosis by color analysis techniques requires information from scale, scale-crust and surrounding areas in dermoscopy images
Comput. Biol. Med.
- Q. Abbas et al.Unsupervised skin lesions border detection via two-dimensional image analysis
Comput. Methods Prog. Biomed.
- K. Korotkov et al.Computerized analysis of pigmented skin lesions: a review
Artif. Intell. Med.
- N. Noroozi et al.Computerized measurement of melanocytic tumor depth in skin histopathological images(Video) Histopathology Skin-- Squamous cell carcinoma
- C. Lu et al.Automated analysis and diagnosis of skin melanoma on whole slide histopathological images
J. Pattern Recognit.
- M. Mokhtari et al.Computer aided measurement of melanoma depth of invasion in microscopic images
- C. Lu et al.Detection of melanocytes in skin histopathological images using radial line scanning
- S. Rathore et al.Ensemble classification of colon biopsy images based on information rich hybrid features
Comput. Biol. Med. J.
- M.M. Krishnan et al.Automated oral cancer identification using histopathological images: a hybrid feature extraction paradigm
J. Am. Acad. Dermatol.
Dermatol. Clin. J.
Dermatol. Clin. J.
Clinical Hematology: Theory and Procedures
Pathology and pathobiology of actinic (solar) keratosis – an update
Br. J. Dermatol.
Assisted diagnosis of cervical intraepithelial neoplasia (CIN)
IEEE J. Sel. Top. Signal Process.
- AI-based carcinoma detection and classification using histopathological images: A systematic review
2022, Computers in Biology and Medicine
Histopathological image analysis is the gold standard to diagnose cancer. Carcinoma is a subtype of cancer that constitutes more than 80% of all cancer cases. Squamous cell carcinoma and adenocarcinoma are two major subtypes of carcinoma, diagnosed by microscopic study of biopsy slides. However, manual microscopic evaluation is a subjective and time-consuming process. Many researchers have reported methods to automate carcinoma detection and classification. The increasing use of artificial intelligence (AI) in the automation of carcinoma diagnosis also reveals a significant rise in the use of deep network models. In this systematic literature review, we present a comprehensive review of the state-of-the-art approaches reported in carcinoma diagnosis using histopathological images. Studies are selected from well-known databases with strict inclusion/exclusion criteria. We have categorized the articles and recapitulated their methods based on specific organs of carcinoma origin. Further, we have summarized pertinent literature on AI methods, highlighted critical challenges and limitations, and provided insights on future research direction in automated carcinoma diagnosis. Out of 101 articles selected, most of the studies experimented on private datasets with varied image sizes, obtaining accuracy between 63% and 100%. Overall, this review highlights the need for a generalized AI-based carcinoma diagnostic system. Additionally, it is desirable to have accountable approaches to extract microscopic features from images of multiple magnifications that should mimic pathologists′ evaluations.
- Pre-cancer risk assessment in habitual smokers from DIC images of oral exfoliative cells using active contour and SVM analysis
2017, Tissue and Cell
Citation Excerpt :
Present study explores the possibility of segmentation of DIC images of oral epithelial cells through active contour snake model to distinguish cancer cells from normal. On the other hand, different analytical methods are used for prediction of pre-malignant trend, based on sequential changes in cellular and nuclear levels among the habitual smokers (Noroozi and Zakerolhosseini, 2016). However, according to the literature very few automated systems are available for oral cancer detection based on feature extraction (Krishna et al., 2012, 2010).(Video) Pagetoid Differential: Paget disease vs Melanoma in situ vs Squamous cell carcinoma in situ (Bowen)
Habitual smokers are known to be at higher risk for developing oral cancer, which is increasing at an alarming rate globally. Conventionally, oral cancer is associated with high mortality rates, although recent reports show the improved survival outcomes by early diagnosis of disease. An effective prediction system which will enable to identify the probability of cancer development amongst the habitual smokers, is thus expected to benefit sizable number of populations. Present work describes a non-invasive, integrated method for early detection of cellular abnormalities based on analysis of different cyto-morphological features of exfoliative oral epithelial cells. Differential interference contrast (DIC) microscopy provides a potential optical tool as this mode provides a pseudo three dimensional (3-D) image with detailed morphological and textural features obtained from noninvasive, label free epithelial cells. For segmentation of DIC images, gradient vector flow snake model active contour process has been adopted. To evaluate cellular abnormalities amongst habitual smokers, the selected morphological and textural features of epithelial cells are compared with the non-smoker (−ve control group) group and clinically diagnosed pre-cancer patients (+ve control group) using support vector machine (SVM) classifier. Accuracy of the developed SVM based classification has been found to be 86% with 80% sensitivity and 89% specificity in classifying the features from the volunteers having smoking habit.
- Computer assisted diagnosis of basal cell carcinoma using Z-transform features
2016, Journal of Visual Communication and Image Representation
Citation Excerpt :
An automated melanoma diagnosis system is presented by the authors in  based on the methods in [17,20] for melanocyte detection and epidermis segmentation. The problem of melanocytic tumor depth measurement is addressed in [21,24] and differential diagnosis of squamous cell carcinoma from actinic keratosis using time–frequency and connected component analysis is considered in . The problem of basal cell carcinoma detection is addressed in [26–30].
Detection of basal cell carcinoma tumor is of great importance for decision making in the disease treatment procedure. Visual inspection of the histopathological slides for tumor detection is laborious, time consuming and prone to inter and intra observer variability. In this paper, we have proposed an automated method for discriminating basal cell carcinoma tumor from squamous cell carcinoma tumor in skin histopathological images using Z-transform features, which were not used previously in image classification tasks. For the first time, it is shown that how two or three Fourier transform features can be combined to form one Z-transform feature. Experiments have shown that the tumor classification results obtained by our method are in reasonable agreement with the gold standards provided by expert pathologists.
Research articleSkin Immunity to Candida albicans
Trends in Immunology, Volume 37, Issue 7, 2016, pp. 440-450
Candida albicans is a dimorphic commensal fungus that colonizes healthy human skin, mucosa, and the reproductive tract. C. albicans is also a predominantly opportunistic fungal pathogen, leading to disease manifestations such as disseminated candidiasis and chronic mucocutaneous candidiasis (CMC). The differing host susceptibilities for the sites of C. albicans infection have revealed tissue compartmentalization with tailoring of immune responses based on the site of infection. Furthermore, extensive studies of host genetics in rare cases of CMC have identified conserved genetic pathways involved in immune recognition and the response to the extracellular pathogen. We focus here on human and mouse skin as a site of C. albicans infection, and we review established and newly discovered insights into the cellular pathways that promote cutaneous antifungal immunity.
Research articleComputer-aided diagnosis of psoriasis skin images with HOS, texture and color features: A first comparative study of its kind(Video) Squamous Cell Carcinoma In Situ (Bowen's Disease) with Clear Cell Change: 5-Minute Pathology Pearls
Computer Methods and Programs in Biomedicine, Volume 126, 2016, pp. 98-109
Psoriasis is an autoimmune skin disease with red and scaly plaques on skin and affecting about 125 million people worldwide. Currently, dermatologist use visual and haptic methods for diagnosis the disease severity. This does not help them in stratification and risk assessment of the lesion stage and grade. Further, current methods add complexity during monitoring and follow-up phase. The current diagnostic tools lead to subjectivity in decision making and are unreliable and laborious.
This paper presents a first comparative performance study of its kind using principal component analysis (PCA) based CADx system for psoriasis risk stratification and image classification utilizing: (i) 11 higher order spectra (HOS) features, (ii) 60 texture features, and (iii) 86 color feature sets and their seven combinations. Aggregate 540 image samples (270 healthy and 270 diseased) from 30 psoriasis patients of Indian ethnic origin are used in our database. Machine learning using PCA is used for dominant feature selection which is then fed to support vector machine classifier (SVM) to obtain optimized performance. Three different protocols are implemented using three kinds of feature sets. Reliability index of the CADx is computed.
Among all feature combinations, the CADx system shows optimal performance of 100% accuracy, 100% sensitivity and specificity, when all three sets of feature are combined. Further, our experimental result with increasing data size shows that all feature combinations yield high reliability index throughout the PCA-cutoffs except color feature set and combination of color and texture feature sets. HOS features are powerful in psoriasis disease classification and stratification. Even though, independently, all three set of features HOS, texture, and color perform competitively, but when combined, the machine learning system performs the best. The system is fully automated, reliable and accurate.
Research articleEndovascular Repair as a Bridge to Open Repair of a Ruptured Descending Thoracic Aspergillus Aortitis
The Annals of Thoracic Surgery, Volume 104, Issue 6, 2017, pp. e425-e428
Invasive aspergillosis rarely involves the thoracic aorta and is associated with a poor prognosis. A 56-year-old heart transplant recipient presented with invasive aspergillosis, primary Aspergillus aortitis, and a ruptured thoracic aorta pseudoaneurysm. Open surgical repair was not possible because of severe sepsis. Therefore, a sequential surgical strategy was planned, including emergency thoracic endovascular aortic repair, followed by antifungal treatment and definitive open repair with explantation of the endograft and placement of a cryopreserved arterial allograft under extracorporeal membrane oxygenator support. The infection did not reoccur during follow-up, and the patient remained alive and well 13 months after the operation.
Research articleInferring patterns in mitochondrial DNA sequences through hypercube independent spanning trees
Computers in Biology and Medicine, Volume 70, 2016, pp. 51-57
Given a graph G, a set of spanning trees rooted at a vertex r of G is said vertex/edge independent if, for each vertex v of G, , the paths of r to v in any pair of trees are vertex/edge disjoint. Independent spanning trees (ISTs) provide a number of advantages in data broadcasting due to their fault tolerant properties. For this reason, some studies have addressed the issue by providing mechanisms for constructing independent spanning trees efficiently. In this work, we investigate how to construct independent spanning trees on hypercubes, which are generated based upon spanning binomial trees, and how to use them to predict mitochondrial DNA sequence parts through paths on the hypercube. The prediction works both for inferring mitochondrial DNA sequences comprised of six bases as well as infer anomalies that probably should not belong to the mitochondrial DNA standard.
Research articleNeither global nor local: A hierarchical robust subspace clustering for image data
Information Sciences, Volume 514, 2020, pp. 333-353
In this study, we consider the problem of subspace clustering in the presence of spatially contiguous noise, occlusion, and disguise. We argue that self-expressive representation of data, which is a key characteristic of current state-of-the-art approaches, is severely sensitive to occlusions and complex real-world noises. To alleviate this problem, we highlight the importance of previously neglected local representations in improving robustness and propose a hierarchical framework that combines the robustness of local-patch-based representations and the discriminative property of global representations. This approach consists of two main steps: 1) A top-down stage, in which the input data are subject to repeated division to smaller patches and 2) a bottom-up stage, in which the low rank embedding of representation matrices of local patches in the field of view of a corresponding patch in the upper level are merged on a Grassmann manifold. This unified approach provides two key pieces of information for neighborhood graph of the corresponding patch on the upper level: cannot-links and recommended-links. This supplies a robust summary of local representations which is further employed for computing self-expressive representations using a novel weighted sparse group lasso optimization problem. Numerical results for several data sets confirm the efficiency of our approach.
Research articleSegmentation and classification of melanoma and benign skin lesions
Optik, Volume 140, 2017, pp. 749-761
The incidence ofmalignant melanoma has been increasing worldwide. An efficient non-invasive computer-aided diagnosis (CAD) is seen as a solution to make identification process faster, and accessible to a large population. Such automated system relies on three things: reliable lesion segmentation, pertinent features’ extraction and good lesion classifier. In this paper, we propose an automated system that uses an Ant colony based segmentation algorithm, takes into consideration three types of features to describe malignant lesion:geometrical properties, textureand relative colors from which pertinent ones are selected, and uses two classifiers K-Nearest Neighbor (KNN) and Artificial Neural Network (ANN). The objective of this paper is to test the efficiency of the proposed segmentation algorithm, extract most pertinent features that describe melanomas and compare the two classifiers. Our automated system is tested on 172 dermoscopic images where 88 are malignant melanomas and 84 benign lesions. The results of the proposed segmentation algorithm are encouraging as they gave promising results. 12 features seem to be sufficient to detect malignant melanoma. Moreover, ANN gives better results than KNN.(Video) Squamous Cell Carcinoma & Actinic Keratosis 101...Dermpath Basics & Beyond
Copyright © 2016 Elsevier Ltd. All rights reserved.