• General Dermatology
  • Eczema
  • Alopecia
  • Aesthetics
  • Vitiligo
  • COVID-19
  • Actinic Keratosis
  • Precision Medicine and Biologics
  • Rare Disease
  • Wound Care
  • Rosacea
  • Psoriasis
  • Psoriatic Arthritis
  • Atopic Dermatitis
  • Melasma
  • NP and PA
  • Skin Cancer
  • Hidradenitis Suppurativa
  • Drug Watch
  • Pigmentary Disorders
  • Acne
  • Pediatric Dermatology
  • Practice Management

AI-Enabled Devices Equip PCPs to Provide Specialty Care


The authorization of DermaSensor marks a milestone in AI-enabled devices, highlighting opportunities and challenges in their integration into clinical practice.

Physician using tablet | Image Credit: © Tom - stock.adobe.com

Image Credit: © Tom - stock.adobe.com

Earlier this year, the US Food and Drug Administration (FDA) authorized DermaSensor, the first artificial intelligence (AI)-enabled medical device for skin cancer detection in primary care.1 The device, which was already approved for use in the European Union and Australia, establishes a new regulatory precedent for FDA authorization of medical devices incorporating AI and machine learning (ML) technologies within dermatology. A recent publication examined the clinical evidence and regulatory implications of AI-enabled technologies in health care, noting the approval of DermaSensor as a milestone in the industry.2


Typically, clinical diagnosis in dermatology relies heavily on visual assessment. Researchers noted performing this visual analysis requires years of training and practice, challenging non-specialists and providing an opportunity for innovation through AI/ML pattern recognition. Oftentimes, the study found primary care physicians (PCP) are screening patients prior to referring to a dermatologist,3 who then complete a thorough assessment and dermatoscopy before diagnosis via biopsy or excision. This is where the researchers noted DermaSensor provides an opportunity to impact the broader health care system by bringing specialty care into the primary care office.

Evolution of Regulatory Framework

The FDA's regulatory framework for medical devices, originally designed for hardware-based technologies, is adapting to accommodate the integration of software technologies such as AI and ML. Researchers stated these devices can be categorized into "Software in a Medical Device" (SiMD), where AI/ML enhances the functionality of physical devices like DermaSensor, and "Software as a Medical Device" (SaMD), which are primarily software-based like Digital Diagnostic’s diabetic retinopathy software.4

Researchers found the FDA's 2021 Action Plan affirmed its approach to regulate medical devices incorporating software, intending to use existing pathways like premarket clearance (510(k)), De Novo classification, and premarket approval (PMA) while developing new regulations specific to software issues (e.g., managing algorithmic bias). Devices are categorized by risk level, determining the extent of pre- and post-market evidence required.

The PMA pathway is reserved for high-risk devices and requires clinical evidence of safety and efficacy. MelaFind, the first dermatology device to receive FDA authorization, was authorized through the PMA pathway and only indicated for use by dermatologists and melanoma detection. However, the device was discontinued due to low specificity (10%) leading to unnecessary biopsies and limited clinical utility.5 Nevisense, another AI device authorized in 2017 under PMA for dermatologists, remains on the market with better sensitivity (96%) but lower specificity (34%) for melanoma detection.6 Unlike these devices, DermaSensor was reviewed under the De Novo pathway, designed for novel devices of low to moderate risk.

DermaSensor Authorization

The FDA's authorization of DermaSensor marks a breakthrough in clinical practice, allowing non-specialists to use the device for assessing skin lesions concerning melanoma, basal cell carcinoma, and squamous cell carcinoma in patients aged 40 years and older. Researchers found the decision was supported by evidence from 3 key studies: a pivotal trial (derm-success), a supplemental validation study for melanoma (derm-assess), and a clinical utility study.7

Derm-success involved 1579 lesions from 1005 patients across 22 primary care centers, demonstrating DermaSensor's sensitivity of 95.5% compared to 83% for primary care physicians, with a high negative predictive value (96.6%) and non-inferior sensitivity to dermatologists (90%). However, researchers noted specificity was lower at 20.7%. The supplemental study, derm-assess, involved 311 patients and showed the device could assist in deciding lesion management, potentially reducing unnecessary referrals.

The clinical utility study involving 108 PCPs and over 10,000 lesions indicated that DermaSensor improved sensitivity in lesion management (91.4% vs. 82.0%) and diagnostic sensitivity (81.7% vs. 71.1%), while decreasing false negative referrals (8.6% from 18%). However, there was a significant decrease in specificity for referrals (44.2% to 32.4%), a challenge researchers have previously seen in other AI-enabled dermatological devices such as MelaFind and Nevisense.

Clinical implications and the Path Forward

DermaSensor represents to researchers a significant advancement in dermatologic care by enhancing PCP’s ability to diagnose skin cancer, addressing access limitations, and long wait times for dermatology referrals. Unlike previous AI-enabled devices, DermaSensor enables PCPs to extend their diagnostic capabilities, similar to advancements in other medical disciplines like diabetic retinopathy screening.8 Researchers found the device's authorization raised questions about the evolving role of PCPs in specialist diagnosis and the potential to defer dermatologist evaluations. They noted concerns remain about its effectiveness across diverse patient populations, particularly those historically underrepresented in clinical trials.

The study stated the FDA's regulatory conditions emphasize the need for ongoing post-market research to ensure equitable performance and monitor diagnostic accuracy to avoid unnecessary treatments. Overall, researchers found DermaSensor's approval marks a milestone in AI-enabled medical devices, highlighting both opportunities and challenges in their integration into clinical practice.


  1. Office of the Commissioner. FDA roundup: January 16, 2024. News Release. US Food and Drug Administration. Jan. 16, 2024. Accessed June 19, 2024. https://www.fda.gov/news-events/press-announcements/fda-roundup-january-16-2024.
  2. Venkatesh KP, Kadakia KT, Gilbert S. Learnings from the first AI-enabled skin cancer device for primary care authorized by FDA. NPJ Digit Med. 2024;7(1):156. Published 2024 Jun 15. doi:10.1038/s41746-024-01161-1
  3. Chen SC, Pennie ML, Kolm P, et al. Diagnosing and managing cutaneous pigmented lesions: Primary care physicians versus dermatologists. J Gen Intern Med 21, 678–682 (2006). https://doi.org/10.1111/j.1525-1497.2006.00462.x
  4. Center for Devices and Radiological Health. Software as a medical device (SaMD). US Food and Drug Administration. Accessed June 19, 2024. https://www.fda.gov/medical-devices/digital-health-center-excellence/software-medical-device-samd.
  5. Cukras AR. On the comparison of diagnosis and management of melanoma between dermatologists and MelaFind. JAMA Dermatol. 2013;149(5):622–623. doi:10.1001/jamadermatol.2013.3405
  6. Ollmar S, Grant S. Nevisense: Improving the accuracy of diagnosing melanoma. Melanoma Management, 3(2), 93–96. https://doi.org/10.2217/mmt-2015-0004
  7. Clinical studies. DermaSensor. January 23, 2024. Accessed June 19, 2024. https://www.dermasensor.com/clinical-studies/.
  8. Office of the Commissioner. FDA permits marketing of artificial intelligence-based device to detect certain diabetes-related eye problems. US Food and Drug Administration. April 11, 2018. Accessed June 19, 2024. https://www.fda.gov/news-events/press-announcements/fda-permits-marketing-artificial-intelligence-based-device-detect-certain-diabetes-related-eye.
Recent Videos
© 2024 MJH Life Sciences

All rights reserved.