Computer-Aided System Diagnoses Melanoma with More Than 90% Accuracy

Computer-Aided System Diagnoses Melanoma with More Than 90% Accuracy

A new method of detecting melanoma has been developed by researchers at the Kaunas University of Technology and Lithuanian University of Health Sciences that is non-invasive and has been shown to be more than 90% accurate.

Melanoma is a common cause of cancer and accounts for around a fifth of all cancer diagnoses in Europe. If not diagnosed and treated early, melanoma is often fatal. When mortality rates are age-standardized, Northern Europe has the highest mortality rate in the region of 3.8 per 10,000 individuals and an incidence of 23.4 per 10,000 individuals.

It is essential for melanoma to be diagnosed early before the cancer metastasizes. Treatment involves excision of the melanoma; however, melanoma must first be diagnosed. Diagnosis of melanoma starts with a visual inspection; however, since many primary care physicians do not get many cases, melanomas can easily be misdiagnosed. Currently only around 65% of cases are accurately diagnosed.

The new computer-aided diagnostic test developed by the researchers combines diagnostic imaging from non-invasive imaging technologies such as optical spectrophotometry and ultrasound and helps to differentiate melanoma from a dysplastic nevus.

To develop their computer-aided system, the researchers used diagnostic images of skin lesions from 100 patients and compared and analyzed complex data on skin tumors obtained using multiple techniques. The computer-aided system they developed was shown to be capable of diagnosing melanoma with more than 90% accuracy.

The system is not intended to be used as a replacement for a trained clinical eye, but it is hoped it will help to speed up diagnosis and allow more patients to be examined, which could lead to an earlier diagnosis at a time when treatment is most likely to be effective.

The cost of any system could be a limiting factor, but the calculations of the researchers suggest the system could be affordable for even small medical practices.

You can read more about the study in the paper – Diagnostics of Melanocytic Skin Tumours by a Combination of Ultrasonic, Dermatoscopic and Spectrophotometric Image Parameters – which was recently published in the journal Diagnostics. DOI: 10.3390/diagnostics10090632

Leave a Reply