Dermatology is the second-most visited specialty in Sweden, with an estimated 400,000 dermatology‑related consultations occurring every year, in a country with less than 5 dermatologists per 100,000 people.

Many of these consultations involve suspicious moles and lesions where skin cancer is suspected. With large numbers of referrals likely to be assessed as benign – specialist clinical time is often spent assessing and providing reassurance to patients with benign lesions rather than prioritising patients with genuine clinical need.

This contributes to overall waiting times for dermatology treatment that are commonly up to 12 weeks for routine appointments (sometimes longer), exceeding the Vårdgaranti standard, which sets a requirement for first specialist assessment within 90 days of referral.

For the NHS in England, the 28‑day Faster Diagnosis Standard (FDS) requires that patients with urgent suspected cancer referrals should receive a diagnosis – or have cancer ruled out – within 28 days, with current performance around 77% meeting this target. Across both systems, the greatest bottleneck lies in early triage and risk stratification: determining accurately and efficiently who needs urgent specialist evaluation and who can be managed safely without it.

This session draws on five years of experience deploying artificial intelligence as a medical device (AI) in NHS skin‑cancer pathways, combining learnings from the real-world use of AI in a clinical setting, evaluations of the AI’s performance and value in different pathways, as well as implications for transformation and change management in the implementation of AI in care pathways.

Attendees will be provided with a clear outline of the role AI can play in addressing capacity challenges and improving patient outcomes, leaving with a practical view of where AI fits most effectively in cancer care pathways and its value in optimising scarce clinical resources.

DERM is the only AI approved to make clinical decisions autonomously in the cancer space. DERM analyses dermoscopic images of skin lesions and is trained to assess, classify and triage the most common cancerous, pre-cancerous, and benign skin lesions - outperforming dermatologists. Dr Kalsi will present data on the use of the technology, including:

Real-world performance of AIaMD in assessment and triage of urgent cancer referrals

Autonomous AI has demonstrated diagnostic accuracy comparable to consultant dermatologists in ruling out melanoma, making it a reliable tool for early assessment. By safely identifying low-risk lesions, AI enables autonomous triage, allowing up to 50% of patients to be discharged without the use of specialist capacity. 

The impact of AIaMD on specialist capacity and the implications for providers

With real-world deployments indicating that up to 95% of urgent face-to-face dermatology appointments can be avoided when AI assesses low-risk cases first – this session will assess where the use of AIaMD can provide the most value and address capability and capacity gaps in primary and secondary care – allow clinicians to focus their time and expertise on complex or high-risk patients, improving both throughput and overall patient outcomes.

The impact of AI in supporting effective teledermatology rollouts

With a number of teledermatology programmes currently active in Sweden, the session will share findings from the use of AIaMD in teledermatology rollouts and the impact of the technology on specialist capacity and cost avoidance for public health systems relying on teledermatology pathways.

Insights from operational teams

Deploying AI at scale requires careful attention to workflow design, clinical governance, and regulatory compliance. By sharing learnings from NHS experience, this session will showcase how AI pathways can be effectively developed to support clinical workflows with minimal disruption, while helping to reduce waiting lists, release capacity, and improve efficiency across primary and secondary care. 

Statista. Number of dermatologists in selected European countries in 2015. 2017. Statista, https://www.statista.com/statistics/873707/number-from-dermatologists-in-europe/? srsltid=AfmB0oo765woW8MGCOsKOJzWyVaGYYEnCoVyosJMK-bj10gCNd44NkVM. AccessedDecember 2025. 

VAN RIJSINGEN, M. "Referrals by General Practitioners for Suspicious Skin Lesions: The Urgency of Training." Acta Derm Venerol, vol. 94, 2014, pp.138-141. Medical Journal Sweden, https://medicaljournalssweden.se/actadv/article/view/6295/9558#:~:text=The%2Oaim%200f%20this%20study, GPs%20in%20skin%20cancer%20care.Accessed 15 December 2025.

NHS England. October 2025. Cancer Waiting Times for August 2025 - 26 (Provisional),https://www.england.nhs.uk/statistics/st atistical-work-areas/cancer-waiting-times/monthly-data-and-summaries/2025-26-monthly-cancer-waiting-times-statistics/cancer-waiting-times-for-august-2025-26-provisional/. Accessed 15 December 2025.

Skin Analytics. Everything you need to know about DERM. Skin Analytics, https://skin-analytics.com/wp-content/uploads/2025/02/Everything-you-need-to-know-about-DERM-02.25.pdf. Accessed 15 December 2025.