Photo Credit: Ninja Artist
Researchers compared the sensitivity of an elastic-scattering spectroscopy device versus dermatologist examinations for detecting skin lesions.
Studies have revealed an alarming trend regarding the frequency of total body skin examinations (TBSE)—less than 25% of patients at high risk of skin cancer are receiving recommended preventative TBSEs. A contributing factor is that many primary care facilities lack the training and equipment to implement this diagnostic procedure.
Advanced technology, including AI, may help to alleviate this limitation in care.
“An alternative, noninvasive, handheld device using elastic-scattering spectroscopy and artificial intelligence may augment lesion risk stratification in the primary care setting and increase health care accessibility in resource-constrained settings,” Danielle Manolakos, DO, MPH, and colleagues explained in a study.
Diagnostic Options
The study team developed a prospective, multicenter clinical validation study to determine the safety and effectiveness of elastic-scattering spectroscopy (ESS) in combination with AI for evaluating lesions.
Researchers enrolled 394 patients from four study sites. More than half (54.33%) of the participants were male, and 77.40% were older than 60. One of the original participants was removed from the study due to incomplete evaluations resulting from an uncharged battery in the ESS device.
Physicians identified and analyzed 614 lesions. The testing cohort consisted of 333 randomly identified lesions from 208 participants; researchers evaluated these lesions using ESS with AI. The cross-validation cohort consisted of 281 lesions from 175 patients, examined through clinical dermatoscopic evaluation by expert dermatologists.
Dermatoscopic Evaluation Versus ESS
The researchers observed no statistically significant difference in overall sensitivity between the application of the ESS device with AI and dermatologists’ evaluations (97.04% vs. 96.45%, respectively; P=0.8203). However, the overall specificity of the ESS device was lower (26.22%) than the overall specificity of the dermatologists (56.10%) (P<0.001). The researchers noted that this result assumes that all unbiopsied lesions evaluated by the dermatologists were benign. Investigators measured diagnostic performance with area under the receiver operating characteristic (AUROC) curve; the AUROC was 0.785 for the dermatologists and 0.773 for the ESS.
In terms of effectiveness, the ESS was validated against dermatologists; diagnosis with no statistical difference in sensitivity (P=0.8203). However, dermatologists’ specificity was higher than the ESS device (37.39% vs. 21.74%) (P=0.0096). For biopsied lesions, the AUROC was 0.734 for the ESS cohort versus 0.699 for the dermatologist-examined cohort.
Subgroup analysis focusing on specific skin cancer types showed that the ESS device displayed a sensitivity for melanoma at 96.67%. This included the detection of severely atypical melanocytic nevi and abnormal melanocytic proliferations. The ESS device also displayed a sensitivity of 97.22% and 97.01% for basal cell carcinoma and squamous cell carcinoma, respectively.
“A limitation of this study was that the sample size was for all cancer lesions and not specifically for melanoma. Therefore, the sample size for melanoma in this study (n=30) must be considered in context of the results reported,” the team noted. “A subsequent study examined 440 lesions from 328 enrolled patients to address the use of the device as an adjunctive tool for the evaluation of pigmented lesions suspicious for melanoma, and the device’s sensitivity for melanoma was found to be 95.5% (n=88).”
The negative predictive value (NPV) for all lesions was 89.58%, and for biopsied lesions only, NPV was 83.33% compared with 87.76% for dermatologist-examined results. The positive predictive value observed for the ESS device was 57.54% for all lesions examined and 64.57% for biopsied lesions only, at a prevalence of 60%. (Table)
Dr. Manolakos and colleagues concluded that “[The ESS device’s] rapid clinical analysis of lesions allows for easy integration into clinical practice infrastructures…Proper use of this device may aid in the reduction of morbidity and mortality associated with skin cancer through expedited and enhanced detection and intervention.”
