Research within AI
Innovation by partnership and collaboration – the development of AI features is a natural part of the improving efficiency, quality, and equality in healthcare.
R&D partnership on AI
Research and partnerships are crucial to the creation of innovative solutions in healthcare. With AI, there are great opportunities to increase the diagnostic reliability of teledermatology to assess pigmented skin lesions and skin changes.
In 2018, a collaboration, funded by Sweden’s Vinnova Fund, with a focus on artificial intelligence was initiated between Gnosco, KTH Royal Institute of Technology in Stockholm and Karolinska University Hospital. Results were presented at Vitalis 2019, the largest eHealth event in Scandinavia.
The aim of the project was to develop AI support in Dermicus’ platform in an established care process for increased knowledge and early detection of malignant melanoma. The first versions of AI features in the platform were developed and tested. What made this project unique was that the project team was the first in the world to implement an advanced algorithm in an existing care flow.
“The project by Vinnova in 2018 was the beginning of our investment in AI. That was the foundation of our continued development in AI features for our Dermicus teledermatology platform and for developing collaborations for further AI studies.”
Daniel Eliasson, CEO of Gnosco
Ongoing research within AI
2020-2025: Teledermoscopy and artificial intelligence: implementation in clinical practice, Start: 2020/11/16 End: 2025/11/30, Lead: Lunds Universitet, Åsa Ingvar with colleagues and Gnosco partner.
Completed projects within AI
2017-2019: Algorithm for image recognition as decision support for early detection of malignant melanoma, funding from Vinnova (Dnr 2017-04646), Gnosco coordinator, collaboration with Karolinska University Hospital and KTH
Gnosco highlighted in reports
In 2018 SwedenBIO launched a major new report “Precision Medicine – The Swedish Industry Guide” representing the first effort to map the Swedish precision medicine industry landscape and showcase precision medicine companies in Sweden to the global life science community. Gnosco was one of the companies in that report.
In May 2020, BioStock.se published this article “Overview of Swedish life science companies using AI” and in the section “Imaging in the era of precision medicine” referring to the SwedenBIO report, and Gnosco was highlighted as one of the companies to develop platforms for precision medicine purposes.
In 2021 the platform has expanded, including several new AI features for image quality assurance, such as Dermicus AI Focus and Dermicus AI Light. These features have been developed by years of experience from over 50 000 patient cases and users input.
“Artificial intelligence is a priority area, and we work collaboratively with leading dermatologists on innovative research projects. A current example is our collaboration is working with dermatologists, to evaluate AI functionality in clinical practice, including AI applying different filters to images and how that will support the consultants’ diagnosis.”
Daniel Eliasson, CEO of Gnosco
Work in close collaboration with healthcare on the development of AI features continues and ongoing development is underway to integrate various functionalities in the platform. Among other things, possible functions are evaluated, such as helping doctors find similar patient cases more easily. Another interesting part we work with here is to develop machine learning algorithms to be able to automate diagnoses in the future. We continue to invest in AI and develop partnerships for research and innovation.
Other research within AI
- Gillstedt M et al, Discrimination between Invasive and In situ Melanomas Using a Convolutional Neural Network, 2021 February, Journal of the American Academy of Dermatology, https://www.gu.se/en/news/algorithm-that-performs-as-accurately-as-dermatologists, svensk summering, https://www.gu.se/nyheter/traffsaker-algoritm-presterade-i-niva-med-hudlakare
- Tschandl P et al, Comparison of the accuracy of human readers versus machine-learning algorithms for pigmented skin lesion classification: an open, web-based, international, diagnostic study, The Lancet Oncology, 2019;20(7), 938-947.
- Massi D, Machine versus man in skin cancer diagnosis, The Lancet Oncology, 2019;20(7), 891-892.
- Tschandl P et al, Jama Dermatology, Expert-Level Diagnosis of Nonpigmented Skin Cancer by Combined Convolutional Neural Networks, 2019;155(1):58-65.
- Esteva A et al, Dermatologist-level classification of skin cancer with deep neural networks, 25 January 2017, Nature.