The issues related to artificial intelligence cover other areas in medicine. It turns out that they can also help couples trying to conceive. This is done by MIM Solutions, which is dynamically developing on the start-up market in Poland. Urszula Sankowska, COO of MIM Solutions, will talk about the solutions she introduces in an interview with Diana Żochowska.
1. Hearing about the use of artificial intelligence in medicine, most people imagine other areas than the treatment of infertility. Where did the idea to deal with this aspect come from?
Urszula Sankowska, COO MIM Solutions: That’s true. Artificial intelligence in medicine may now be associated with other areas of health care, although it seems to me that this is due to the fact that the problem of infertility is less publicized in the media in Poland. It is being talked about more and more in the world, because the percentage of people constantly or periodically affected by the problem of infertility is constantly growing, and today the number of couples struggling with this disease reaches 300 million globally. This is very disturbing. It is also worth mentioning that the World Health Organization has entered infertility on the list of civilization diseases. The problem is serious. This is one of the reasons why we decided to act in this area.
At this point, someone might say – okay, but there is already a solution for these people, such as in vitro fertilization (also known as IVF). I have to write here that despite the fact that the effectiveness of IVF has increased significantly in recent years, this procedure is still only about 30% effective, it is expensive and can be mentally and physically burdening for patients. This is proof that there are still many challenges and limitations to overcome. It is in several such challenges that we try to help doctors with artificial intelligence (AI). We believe that AI-based systems can increase the chances of success, reduce errors and lead to faster, cheaper and more accessible treatment processes.
I must admit that the idea was also born out of personal experience. For many years I was a patient of IVF clinics, my path to motherhood was long and winding. I know what patients go through and I know the handicaps of treatment. I want to help women fulfill their dreams of the longed-for child. If we manage to revolutionize infertility treatment, I will be able to say with a clear conscience that my life dream has come true.
2. How can the solutions implemented by MIM Solutions help couples trying to conceive? And can they really help doctors?
U.S.: First of all, we want to increase the effectiveness of the in vitro method and lead to a reduction in treatment costs. We are currently working on three technologies that promise to meet these goals.
The first technology is EMBRYOAID – an application that will support even the most experienced embryologist in choosing the most promising embryo for implantation. Choosing the right embryo to administer to a woman is extremely important as it increases the chance of success, minimizes complications and keeps the baby healthy. The challenge here is the general lack of objective methods for the precise assessment of the quality of the embryos, or – more specifically – the lack of predictive systems that, based on information about the development of the embryo, are able to predict its chances of implantation. Currently available tools to support this decision are limited, highly subjective, usually time-consuming and often very expensive. As a result, embryologists often rely on their experience and observational skills to select embryos for implantation. In particular, the lack of objective indicators can lead to high variability in the process itself. An artificial intelligence (AI) system that will contain such indicators and help doctors make decisions could therefore improve IVF success rates. Our EMBRYOAID system learns how embryos develop over time and then uses that information to identify the best embryos for implantation. By understanding the entire development process, the system is able to identify the appropriate embryos, even on the basis of just one image. This is a cheaper alternative to the current analytical tools that are only available at the most expensive IVF clinics. We believe that EMBRYOAID will enable clinicians to choose the best embryo, thereby reducing the number of IVF cycles needed to achieve a successful pregnancy, improving the success rate and minimizing the risk of multiple pregnancies.
The next system we create is called FOLLISCAN (from the English word follicles i scanning) – this is software designed to support the doctor in the ultrasound examination (USG) of the woman’s ovarian reserve. This is a key test because it allows you to determine a woman’s fertility in a given cycle. This test is performed several times during the IVF process itself and allows you to determine the timing of its individual stages. Thanks to FOLLISCAN, the gynecologist and medical staff will have access to highly specialized medical knowledge, until now reserved mainly for a narrow group of specialists. In addition, FOLLISCAN will automate a large part of the activities that currently have to be performed by a human. This will significantly facilitate the diagnosis in terms of the assessment of the ovarian reserve, as well as the development of antral follicles. We believe that FOLLISCAN will improve the diagnosis of female fertility, and thus contribute to the use of treatments better suited to the patient’s needs and increase their effectiveness.
Another project in which we are involved is to create a system that will provide the doctor with support in the process of making diagnostic and therapeutic decisions. Together with the INVICTA clinic, we are working on a solution that will suggest to the doctor what clinical treatment will be the most optimal for a given patient in the process of treating fertility disorders. Advanced machine learning algorithms will analyze hundreds of thousands of patient histories to create the most effective and cost-effective pathways for diagnosis and treatment. The SI platform we are building will base its suggestions on dozens of parameters, individually assessing the situation of each person. Thanks to this, the treatment will become more personalized and thus more effective. This software will also facilitate the daily work of the doctor. One of its elements is the appropriate presentation of the patient’s treatment history in order to first enable the physician to familiarize himself with the most important information at a given stage of the diagnosis and treatment process. In addition, the platform will prompt the doctor what other important medical parameters to verify in order to properly assess the patient’s situation.
4. How wide is the availability of your solutions? Is there a chance that really every couple in Poland could take advantage of them in the next few years?
U.S.: We would like as many couples as possible to use our technologies in Poland and in the world. Therefore, MIM Solutions plans to go very broadly when it comes to commercialization of created technologies. The target markets in the first line are Poland and Europe. Unfortunately, the problem of infertility is growing year by year and technologies such as EMBRYOAID and FOLLISCAN meet the needs of the market today and will be further developed and improved in the future to meet the demand for this type of product. It seems that there is no turning back from the use of AI in medicine and this trend will only intensify. Thanks to EMBRYOAID and FOLLISCAN, we want to become pioneers in the field of fertility disorders treatment. Currently, there are still very few companies in the world that deal with similar topics, although we know that this trend will change rapidly.
5. On your website you can read that in terms of MedTech, you run projects not only in the field of infertility treatment. In what other areas do you plan to implement AI solutions?
U.S.: MedTech is actually a bottomless well when it comes to AI applications. Recently, we have been running a project in cooperation with TelDoc, aimed at developing a system that allows for the automatic generation of summaries based on electronic medical records and a medical interview conducted with the use of a voicebot using models based on deep neural networks. In the first stage, the system will be dedicated to three specializations, such as general surgery, neurosurgery, and orthopedics and traumatology of the musculoskeletal system.
6. If you were to name three main areas in which artificial intelligence in medicine will develop the most in the coming years. What would it be?
U.S.: The development of AI applications in medicine will probably largely depend on the availability of data fed by the models. We observe very different levels of access to medical data around the world. It seems to me that AI-based solutions will develop the most in places where the availability of huge data sets will be the greatest – and these areas are currently remote diagnostics.
However, I would not expect intelligent technologies to completely replace doctors. As the history of medicine shows, new techniques enable doctors to focus on even more advanced aspects of our health and lead to even greater specialization. And this process alone will lead to more effective treatment of even more diseases.