The CASH (color, architecture, symmetry, and homogeneity) algorithm for dermoscopy
ABSTRACT The color, architecture, symmetry, and homogeneity (CASH) algorithm for dermoscopy includes a feature not used in prior algorithms, namely, architecture. Architectural order/disorder is derived from current concepts regarding the biology of benign versus malignant melanocytic neoplasms.
We sought to evaluate the accuracy of the CASH algorithm.
A total CASH score (TCS) was calculated for dermoscopic images of 325 melanocytic neoplasms. Sensitivity, specificity, diagnostic accuracy, and receiver operating characteristic curve analyses were performed by comparing the TCS with the histopathologic diagnoses for all lesions.
The mean TCS was 12.28 for melanoma, 7.62 for dysplastic nevi, and 5.24 for nondysplastic nevi. These differences were statistically significant (P < .001). A TCS of 8 or more yielded a sensitivity of 98% and specificity of 68% for the diagnosis of melanoma.
This is a single-evaluator pilot study. Additional studies are needed to verify the CASH algorithm.
The CASH algorithm can distinguish melanoma from melanocytic nevi with sensitivity and specificity comparable with other algorithms. Further study is warranted to determine its intraobserver and interobserver correlations.
- SourceAvailable from: Cliff Rosendahl
[Show abstract] [Hide abstract]
- "According to the original dermatoscopic method of classic pattern analysis , the global pattern of the lesion presented here is the so-called “cobblestone” pattern (pattern of clods) described as the pattern of a congenital type nevus and this lesion had none of the specific criteria of melanoma according to that method. Likewise this lesion did not score as a melanoma according to the ABCD dermatoscopic algorithm , the 3-point checklist , the 7-point checklist , the Menzies method  or the CASH algorithm . The “Chaos & Clues” algorithm [14,15] identifies suspicion for malignancy based on the presence of chaos (defined as asymmetry of structure and/or color) plus the presence of at least one of eight clues, including the clue of “white lines,” and therefore that method would identify this lesion as suspicious but only if polarized dermatoscopy was employed. "
ABSTRACT: We report a case of a melanoma arising in a congenital-type compound nevus, which was excised because it was observed by both the patient and the treating dermatologist to have changed. Because the lesion was routinely photo-documented with both polarized and non-polarized dermatoscopy images prior to excision, these images were available for subsequent examination. Matched images are presented in what appears to be unique in the published literature: polarizing-specific white lines are identified as a compelling clue to the diagnosis of melanoma in a lesion that contains no clues apparent in the non-polarized image. Dermatopathology images reveal that the melanoma is arising in conjunction with a congenital type nevus. As expected, dermatoscopic polarizing-specific white lines are evident on the melanoma but not the nevus, and while a possible explanation is discussed, this remains speculative.01/2014; 4(1):83-7. DOI:10.5826/dpc.0401a14
[Show abstract] [Hide abstract]
- "Several methods have been developed to help clinicians interpret the structures revealed through dermoscopy . Well known algorithms include the ABCD rule of dermoscopy , the Menzies method , the Three-point checklist , the 7-point checklist , the CASH algorithm for dermoscopy , the Chaos and Clues algorithm , the BLINCK algorithm , and Pattern Analysis . However, intensive and time consuming training is required to become an expert in dermoscopy . "
ABSTRACT: Background It is often difficult to differentiate early melanomas from benign melanocytic nevi even by expert dermatologists, and the task is even more challenging for primary care physicians untrained in dermatology and dermoscopy. A computer system can provide an objective and quantitative evaluation of skin lesions, reducing subjectivity in the diagnosis. Objective Our objective is to make a low-cost computer aided diagnostic tool applicable in primary care based on a consumer grade camera with attached dermatoscope, and compare its performance to that of experienced dermatologists. Methods and Material We propose several new image-derived features computed from automatically segmented dermoscopic pictures. These are related to the asymmetry, color, border, geometry, and texture of skin lesions. The diagnostic accuracy of the system is compared with that of three dermatologists. Results With a data set of 206 skin lesions, 169 benign and 37 melanomas, the classifier was able to provide competitive sensitivity (86%) and specificity (52%) scores compared with the sensitivity (85%) and specificity (48%) of the most accurate dermatologist using only dermoscopic images. Conclusion We show that simple statistical classifiers can be trained to provide a recommendation on whether a pigmented skin lesion requires biopsy to exclude skin cancer with a performance that is comparable to and exceeds that of experienced dermatologists.Artificial Intelligence in Medicine 12/2013; 60(1). DOI:10.1016/j.artmed.2013.11.006 · 2.02 Impact Factor
[Show abstract] [Hide abstract]
- "This lightly pigmented brown lesion had a structureless pattern centrally combined with a circumferential pattern of radial lines peripherally. This was arguably a symmetrical pattern, taking into account that perfect symmetry is rare in biological material, and although radial lines are regarded as a clue to malignancy by a number of published algorithms [4,5,6,7], this lesion failed to reach the threshold for excision with respect to any algorithm which had a clearly defined flowchart method of predicting malignancy [5,6,7,8,9,10]. The treating clinician made the decision to excise this lesion because of its having a distinctly different dermatoscopic pattern to the signature nevi; he regarded it as a dermatoscopic “ugly duckling.” "
ABSTRACT: We present a case report of a 3.5 mm diameter superficial spreading melanoma on the upper back of a 27-year-old woman, signed out as Clark level 2, Breslow thickness 0.2 mm with regression to 0.45 mm. The patient, with Fitzpatrick type 1 skin and minimal actinic damage, had presented for a routine skin check with no previous history of skin cancers. At the age of 17 she had received chemotherapy and radiotherapy for Ewing's sarcoma of the right hip with pulmonary metastases. The skin lesion was assessed as dermatoscopically symmetrical and was not predicted as a melanoma by any algorithmic method. The provisional diagnosis of melanoma was made on the basis that this lesion was completely different in dermatoscopic pattern to her other nevi, a dermatoscopic "ugly duckling" lesion. We draw attention to the recently established link between defects in the STAG2 gene and Ewing's sarcoma, glioblastoma and melanoma.04/2013; 3(2):59-62. DOI:10.5826/dpc.0302a09