Artificial Intelligence Device Enhances Diagnosis of Skin Cancer by NPs, PAs

Credit: Pexels

The Elastic Scattering Spectroscopy (ESS) device may allow for significantly enhanced decision-making in the diagnosis and management of skin cancer, according to recent findings, and specifically for nurse practitioners (NPs) and physician assistants (PAs) in primary care.¹

These findings resulted from a new study titled ‘Evaluating the Impact of Elastic-scattering Spectroscopy and Machine Learning on Skin Cancer Detection Among Nurse Practitioners and Physician Assistants.’ The data were presented in part by Cason Hucks, with DermaSensor, Inc., at the Maui Derm NP+PA Fall 2024 conference in Nashville, Tennessee.

The implementation of artificial intelligence (AI) in the field of dermatology, and specifically in the early detection of skin cancers, is a frequently-discussed topic of conversation in the dermatology space. Prior research has suggested that clinicians and patients differ in their opinions of AI’s use within the field.²

“As artificial intelligence (AI) broadens its scope within dermatology, leveraging AI coupled with technology such as Elastic Scattering Spectroscopy (ESS) may aid skin cancer detection for these providers,” Hucks and colleagues wrote.¹

Study Design

Across the globe, skin cancer is known to be the most frequently-diagnosed cancer. In fact, it has higher rates than all other cancers combined. Despite this fact, skin cancer also maintains one of the highest rates of survival provided that it is detected at an early enough point.

Although there is a critical role played by primary care providers in the detection of skin cancer, there are areas in which access to such clinicians is limited. In such regions, many patients will only have access to NPs and to PAs.

The investigators of this analysis noted that the implementation of artificial intelligence has expanded within the field of dermatology. They added that the combination of AI with technologies such as ESS may provide aid to detection of skin cancer by care providers, and this includes NPs and PAs.

The research team conducted their randomized, observational, web-based study, with participants of their research reviewing clinical cases shown to them through the use of an online survey. After this period, the subjects would be required to respond to diagnostic inquiries related to whether the lesion was malignant or benign.

They were also provided with inquiries related to management, specifically whether one should investigate further or not. Lastly, they were instructed to rate their confidence on a scale from 1 - 10.

Following the participants’ first evaluation, they were provided with the ESS device output. This would suggest either "Investigate Further" or "Monitor," in addition to a spectral score (1-10) for cases that would be flagged as requiring more assessment. Subjects were then asked to review the cases again and provided the same answers as they had prior.

Major Conclusions

The research team found that there were significant improvements observed in the accuracy of diagnostic and management results among PAs and NPS working in primary care and implementing the ESS device. The severity of diagnoses increased from 77.1% to 79.4% (P = .0123), and the team also found that specificity showed improvement from 45.6% to 50.0% (P < .0001).

In terms of management sensitivity, the investigators also found there was a notable increase. Specifically, it rose from 84.2% to 88.2% (P < .0001), but the team also found that management specificity, as opposed to sensitivity, slightly dipped from 34.7% to 32.5% (P = .0056). Overall, the investigators found that the ESS device led to enhancements in diagnostic and management decision-making for both NPs and PAs.

“The use of this device shows promise in assisting primary care providers in clinically assessing suspicious lesions which may ultimately lead to lower morbidity and mortality in those with skin cancer,” they concluded.

References

1. ^{Hucks C, Smith T, Gregory J, et al. Evaluating the Impact of Elastic-scattering Spectroscopy and Machine Learning on Skin Cancer Detection Among Nurse Practitioners and Physician Assistants. Presented at: Maui Derm NP+PA Fall 2024. Nashville, TN. September 15-18, 2024.}
2. ^{Smith T. Dermatologists, Patients Divided on Augmented Intelligence for Melanoma Screening. HCPLive. March 4, 2024. https://www.hcplive.com/view/dermatologists-patients-divided-on-augmented-intelligence-for-melanoma-screening. Date accessed: September 24, 2024.}