This has been my very first Artificial Intelligence (AI) type of conference that I recently attended. AI is a broader subject and in recent days many new developments have been carried on, companies working on its application in various areas if different industries.

Since my professional focus has been mainly tailored in clinical science, particularly the healthcare, I will share most of the insight covering that area.

‘Real potential of AI in healthcare’ was to me the most interesting and relevant topic that I could not only learn but also relate to and visualise the current and coming advancements, challenges, and the forecast. The presenter was a senior member (CEO) of IQONIC, German company that specializes in hair and skin care and aesthetics using AI model in their clinic calling ‘Digital Clinic’. The clinic uses mainly natural products and treatment approaches.

The key focus was how AI can effectively assist in their area of healthcare, from diagnostics to prevention helping the clinical services and customers.

After some process of research and validation, the company is using AI in order to make their services more targeted, efficient, and cost-effective, assisting clients with various hair and skin conditions to seek as much as possible professional assistance.

The company’s CEO explained the whole journey how it brought the company to this point, approving AI application and successfully using it in their work setting.  Their team works with number of natural products for different skin and hair conditions which requires great deal of customer’s health condition understanding following the assessment. They have created a Digital Clinic model that already operates helping them to run higher-quality service meeting customers’ satisfaction and cost effectiveness. This model is composed of the following stages:

  1. Recording- by taking a picture using a device (ex. smart phone), followed by detailed assessment.
  2. Assessment of customer’s skin / hair conditions using customised solution based on wrinkles, pores, skin, type, sensitivity, age, circles, hair type & density…. The software uses 5 million images within different gender (M/F), ethnicity, age group (15-65), relevant clinical history including skin diseases, appearances of various conditions in different light setting, temperature and other environmental effects.
  3. Evaluation – AI based indication, skin disease diagnosis (if applies)
  4. Consultation– by the software and trained doctors that are available in the background when needed for more specific & advanced type of conditions, jointly covering great spectrum of skin conditions and diseases of targeted areas for suggested treatment and care. 
  5. Recommendation – more targeted set of recommendations, including a cure where applies, using more appropriate and accurate product selection.
  6. Follow up – more tailored and professionally done, as requires.

In their applied AI model, the doctors are not necessarily removed, they are just more thoroughly trained, working rather in the background, addressing more advanced and specific hair and skin conditions, especially ones that can NOT be adequately covered by the software. This allows the doctors to focus on a customer care in a greater depth & details looking into their needs in more targeted way, resembling a model of personalised medicine. It has been also noted that this delivers not only improved quality of customer care service, but it runs in more cost-effective way too. Equally, it lowers a cost for a customer. Their model has demonstrated that involved cost can by lowed as far as 75%, which is quite substantial.

It was interesting to hear how thoughtfully and diligently the company’s personnel worked to create such sophisticated software / working model. They have incorporated thousands of different images (100, 000) where ethically speaking they had to considered all kinds of elements, including the GDPR, especially while their development was still in the frame of research. It was emphasised that initially the participants were via advertising invited for voluntary participation, accordingly reimbursed for their time and willingness to participate, while using anonymised approach where their personal identity was appropriately protected. These generated GDPR compliant data that can be used as well in the other source of their work and research. Like for using in labelling new data points. Within that area it was also mentioned some challenge. Although their data was able to be anonymised in great proportion, eyes have been difficult to be anonymised which has created grey area of full anonymity and data compliance.

To make company’s approaches credible they are regularly audited, approximately twice a year.  During their audits or inspections, they have to demonstrate their data security, confidentiality, doctors’ qualifications and trainings, etc.

This mode was extremely helpful to learn and visualise how AI can be transcribed into different clinical settings including research. 

It has demonstrated that AI can improve sustainability in the healthcare in many ways: by developing more efficient formulations that are more tailored to consumers’ needs, by optimizing resources consumption in production to minimize energy and material losses, and ultimately the cost.

The company’s representative also said the AI should never be used just for the sake of having AI. AI should always be used for a specific use and cases.  Only when the necessary data is or can be available.

Throughout all the conference talks it was indicated that there are still many challenges and ongoing developments, which are gradually being tackled, implemented and improved. 

One of the bigger parts was ethical component and liability. Experts tent to doubt Ethical AI designs. Unfortunately, in AI application processes have been noted lots of cheating, misuse and poor compliance of data that has also raised its awareness.

The conference has presented also other interesting areas of AI, by discussing the application    using great technical elements that are used in different industries like aerodynamics, machineries, robotics. But also in cultural settings, like in area of different languages, especially in Africa where it is helping them leveraging a small language model. 

It was also mentioned that AI as general is lacking people with required set of skills and knowledge. Those working processes take lots of time to be developed, addressing different methods of adaptation, employees’ resistance, intangible benefit, clearer metrics, and demand for more monitoring. 

I like to thank my company for giving me the opportunity to take a part in this AI global forum which has assisted me to have a better confidence and AI understanding of current applications and the future outlook.  

Published: 30 Apr 2024

Author: Galina Fujimori Petrikova