Kelvin Health

Our body is killing us by constantly hiding symptoms until it’s too late.

Kelvin Health changes that by providing a widely accessible intelligent tool that uncovers hidden anomalies in the human body. 

Empowered by AI analysis of thermal images, Kelvin Health significantly reduces the risk of breast cancer undercare by providing a cost and time efficient screening and monitoring approach which leads to wider, earlier, and more relevant intervention by health care providers and pharma companies.

The application of Kelvin Health results in healthier population, better healthcare providers’ ratings and cost savings for the intermediaries.

The primary focus of the project are patients in risk groups and subject to screening and preventive care, or such that are currently subject to active therapy that needs to be monitored.

According to WHO more than 70% of people, even in developed economies, avoid timely examinations or have fear of visiting hospitals (even more now in COVID-19 times) for invasive or radiation-emitting procedures.

Besides pursuing career, “Sandwich Generation” mothers aged between 35 and 54 experience more stress than any other age group owing to the delicate and demanding balancing act of caring for their children and their aging parents, which leave little to no space for preventive care initiated by them using the existing approaches.

In remote and/or underdeveloped regions the access to healthcare and specialists is very limited due to financial and/or logistical hurdles, creating significant health inequality.

The above factors greatly increase the chances of serious health conditions, such as developing breast cancer, being left undiagnosed/unmonitored for a period of time long enough to lead to severe complications.

The International Agency for Research on Cancer reported a total of 18,078,957 cancer cases worldwide in 2018 including 2,088,849 breast cancer (11.6%) cases, making it the second most frequent type. Breast cancer was the most common cancer among women that accounted for 24.2% of the cases. The disease is the leading cause of cancer death in over 100 countries.

Kelvin Health is an intelligent tool for detection and monitoring of various abnormalities and inflammation processes in different health conditions (currently focused on breast cancer) and stages, using AI analysis applied to mobile-based digital thermal imaging.

Inflammation is best characterized by four symptoms, commonly stated in Latin: calor (heat), dolor (pain), rubor (redness), tumor (swelling). Researchers have used thermal imaging in a broad number of diseases where the increased skin temperature (calor) reflects the presence of inflammation in underlying tissues or regions where blood flow is increased or decreased due to medical irregularities.

The IR (infrared thermal) imaging does not require any physical contact, it is non-invasive, allows for unlimited throughput and there are no liquids to drink. Furthermore, it does not use ionizing radiation, making it highly preferable in screening and monitoring for people of various age and risk groups.

Combining the expertise of medical and AI professionals, we have developed a system consisting of a thermal IR camera connected to a mobile device or a desktop computer, a unique thermal image segmentation algorithm, and a ML (machine learning) based software to analyze the thermal images received.

The primary focus of the project are patients in risk groups and subject to screening and preventive care, or such that are currently subject to active therapy that needs to be monitored, for the following health conditions (visualised with real-life images captured by the Kelvin Health application): a) oncology and breast cancer (our current focus), in particular, b) autoimmune and rheumatoid arthritis, in particular, c) vascular and thrombophlebitis, in particular, and d) respiratory infections and COVID-19 in particular:

Here are statistics regarding the cancer pandemic worldwide and the significance of breast cancer in particular. The UN has forecasted that the global population will reach 8.3 billion by 2030 (United Nations 2019). The effect of population ageing and growth will be greatest in low-income and middle-income countries. These changes translate to a predicted global burden of 20.3 million new cancer cases by 2030 compared with an estimated 12.7 million cases in 2008, and a predicted 13.2 million cancer-related deaths worldwide by 2030, up from 7.6 million in 2008 (Pilleron S 2020; World Health Organization 2018) .

The International Agency for Research on Cancer reported a total of 18,078,957 cancer cases worldwide in 2018 including 2,088,849 breast cancer (11.6%) cases, making it the second most frequent type (Bray 2018). Breast cancer was the most common cancer among women that accounted for 24.2% of the cases (Bray 2012).

The disease is the most frequently diagnosed female cancer in the vast majority of the countries (154 of 185) and is also the leading cause of cancer death in over 100 countries (Bray 2012).

Globocan and other authors (Bray 2018; All Cancer Globocan 2018) have predicted a 50% rise in the number of prevalence in breast cancer by 2040.

Cancer registry data clearly show that early detection of breast cancer (lower TNM stage) is associated with better survival (Ng, 2009). Detection of non-invasive breast cancer (ductal carcinoma in situ, DCIS) exhibits a relative survival of affected women of > 94%. For invasive breast cancers of < 1 cm, relative survival after 15 years ranges from 90 to 92%; for tumors of 1–2 cm, it ranges around 75% and continues to drop with increasing T stage. On average, the death rate increases by 1.3% per millimetre increase in size. With increasing tumor size at detection, a shift towards higher grading becomes visible. All these facts support our knowledge that early detection has a significant impact on prognosis (tumorregister-muenchen).

Early detection of breast cancer could increase the survivability rate up to 90% of all cases within a five-year window, thus, the need of an easy-access, cheap and trustable breast cancer screening method still latent in underdeveloped countries (Ng, 2009).

The management of patients with breast cancer has a significant impact on the national budget. It is, therefore, a public health priority to support effective prevention programs that could reduce cancer costs or provide therapeutic interventions that increase patients’ chances of survival, thus reducing the indirect costs of morbidity and mortality. The reality of the cost burden is evidenced by numerous studies; for example, Smidova 2012 showed that the productivity loss related to breast cancer in Poland amounted to 583.7 million EUR in 2010 and 699.7 million EUR in 2014. During this period, this amounts to 0.162–0.171% of the GDP. Public finance expenditure for social insurance benefits to breast cancer sufferers ranged from 50.2 million EUR (2010) to 56.6 million EUR (2014), or 0.72–0.79% of the total expenses for all diseases. Loss of opportunity in public finance revenue amounted to 173.9 million EUR in 2010 and 211.0 million EUR in 2014. 

Unar-Munguıa et al., 2017 estimated that 245 million USD in medical expenses and income lost due to breast cancer could be spared over the life of a cohort of Mexican women. Medical costs account for 80% of the economic burden; loss of income and opportunity costs for careers represent 15% and 5%, respectively.

In contract to all worrying statistics above, the results of the thermography can be correct 8–10 years before a mammogram can detect a mass in the body of the patient. In addition to this, a shortage of qualified radiologists has caused an urgent need for the development of computer technologies like Computerized Thermal Imaging (CTI), which analyzes and interprets a series of infrared images using algorithms that correlate infrared data regarding the breast being examined with infrared models associated with benign or malignant breast tissue. It does not require any physical contact, and there are no liquids to drink. Furthermore, it does not use ionizing radiation, making it highly preferable in diagnostic procedures for pregnant women and younger women (Mambou S 2018 and Mambou S 2018).

Literature:

All Cancer Globocan 2018. International Agency for Research on Cancer WHO. [cited 2019 Mar 3] Available from: http://gco.iarc.fr/today/data/... cancers/39-All-cancers-fact-sheet.pdf.

Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: Globocan estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: Cancer J Clin. 2018;68(6):394–424.

Bray F, Jemal A, Grey N, et al. Global cancer transitions according to the human development index (2008–2030): a population-based study. Lancet Oncol. 2012;13(8):790–801.

Diakides N.A, M. Diakides , J.C. Lupo , J.L. Paul , R. Balcerak , Advances in medical infrared imaging, in: A Diakides Nicholas, D Bronzino Joseph (Eds.), Medical Infrared Imaging, Taylor and Francis, CRC Press, Boca Raton, 2007.

Heywang-Köbrunner, Hacker, Sedlacek. Advantages and Disadvantages of Mammography Screening. Breast Care 2011;6:199–207

Global Oncology Trends 2019: https://www.iqvia.com/-/media/iqvia/pdfs/institute-reports/global-oncology-trends-2019.pdf

Global Oncology Trends 2019 – IQVIA: www.iqvia.com › insights › the-iqvia-institute › reports

Kakileti S.T., G. Manjunath , H. Madhu , H.V. Ramprakash , Advances in Breast Thermography, In New Perspectives in Breast Imaging, InTech, 2017.

Kandlikar SG, Perez-Raya I, Raghupathi PA, et al. Infrared imaging technology for breast cancer detection– current status, protocols and new directions. Int J Heat Mass Transf. 2017;108:2303–2320.

Kosus N., Kosus A., Duran M., Simavlı S., Turhan N. Comparison of standard mammography with Digital mammography and Digital infrared thermal imaging for breast cancer screening. J. Turk. Ger. Gynecol. Assoc. 2010;11:152–157.

Mambou S., Maresova P., Krejcar O., Selamat A., Kuca K. Breast Cancer Detection Using Modern Visual IT Techniques. In: Andrzej S., Adrianna K., Manuel N., Quang Thuy H., editors. Modern Approaches for Intelligent Information and Database Systems. Volume 769. Springer; Berlin, Germany: 2018. pp. 397–407.

Mambou Sebastien Jean, Petra Maresova, Ondrej Krejcar,1 Ali Selamat, and Kamil Kuca1 Breast Cancer Detection Using Infrared Thermal Imaging and a Deep Learning Model, Sensors (Basel). 2018 Sep; 18(9): 2799. Published online 2018 Aug 25. doi: 10.3390/s18092799

Ng E.Y.K., N.M. Sudarshan, Numerical computation as a tool to aid thermographic Interpretation, Journal of Medical Engineering & Technology 25 (2001) 53–60.

Pilleron Sophie,  Enrique Soto-Perez-de-Celis,  Jerome Vignat, Jacques Ferlay, Isabelle Soerjomataram, Freddie Bray, Diana Sarfati, Estimated global cancer incidence in the oldest adults in 2018 and projections to 2050, https://doi.org/10.1002/ijc.33...

Šmídová I. Alcohol and Breast Cancer-Economic Costs. Hygiena. 2012;51:17–21

tumorregister-muenchen: www.tumorregister-muenchen.de/....

Unar-Munguıa M., Meza R., Colchero M.A., Torres-Mejía G., de Cosío T.G. Economic and disease burden of breast cancer associated with suboptimal breastfeeding practices in Mexico. Cancer Causes Control. 2017;28:1381–1391. 

United Nations 2019 Revision of World Population Prospects: https://population.un.org/wpp/

World Health Organization, Latest global cancer data: Cancer burden rises to 18.1 million new cases and 9.6 million cancer deaths in 2018, PRESS RELEASE N° 263, 12 September 2018 

Yszczarz B.L., Nojszewska E. Productivity losses and public finance burden attributable to breast cancer in Poland, 2010–2014. BMC Cancer. 2017;17.

Kelvin Health's approach will increase significantly the level of care for breast cancer patients, especially in underdeveloped and remote regions.

The best ways to improve the care of an oncology disease patients are the prevention and the early detection of the disease. Our proposed solution responds to the second one. As we have already seen early detection of breast cancer (lower TNM stage) is associated with better survival and a shift of the therapy towards adjuvant setting.

Kelvin Health is a step toward precision medicine, providing equal access to quality care even in the lack of expensive equipment and specialists.

We are currently in research collaborations with several hospitals in Europe  (including 1st Genealogical Hospital "St. Sofia", Bulgaria and Oncology Centre Maribor, Slovenia) and Novartis Oncology Europe (and their innovation department BIOME).

The focus of the collaborations is the collection of enough clinically validated data for breast cancer patients, as well as healthy patients screened for breast cancer, in combination with thermal images captured with our specific thermal palette and segmentation, in order to build a robust machine learning model for early detection of the disease.

The hardware and software infrastructure are already implemented. As we progress collecting data we are constantly updating the models.

Once statistically significant real-world evidence is collected we are ready to proceed with certifications and piloting the solution with our existing partners, this time in a commercial setting.

Our long term vision is to make Kelvin Health accessible to primary health care providers and consumers.

Being a solution that is non-invasive, non-damaging, and with an unbound frequency of use, Kelvin Health allows regular short-interval monitoring of the conditions of patients in outpatient and even domestic (if the patients have access to thermal cameras compatible with the solution) setting. The associated cost of the Kelvin Health analysis would be an order of magnitude lower compared to other imaging approaches which allows for both cost-savings and more regular screening and monitoring, especially crucial in underdeveloped regions, as .

The continuously growing utilization of AI in the process makes it even more efficient and precise (both in terms of sensitivity and specificity) over time, learning form permanent flow of patient and clinical data.

The applications of Kelvin Health include a) regular checks of the target risk groups with screening purpose for preventive care, b) monitoring in the context of outpatient and domestic care when already diagnosed with a particular condition and c) therapy monitoring for patients on specific and/or experimental treatment plans in outpatient, inpatient, and clinical trial settings.

The current number of people we are serving is non-existent, as the project is in active R&D collaboration stage currently. Still, the potential impact it's going to have in the mid-long term is what is driving us to keep pushing the frontier on a daily basis.

The number we’ll be serving in one year is also hard to define, as Kelvin Health will still be in the process of getting formal approvals, but we would have already touched the lives of at least 3,000 patients involved in our observations.

The number we’ll be serving in five years, when the project will have achieved global reach, can be more than 8,000,000. This is the lower bound of the expected unfortunate number of people that will be diagnosed with breast cancer for 4 consecutive years (mid 2022 to mid 2026). However, if we have managed to help them detect it as early as possible, we would have contributed significantly to saving their lives.

We are monitoring and measuring the following KPI's:

• Аmount of ground-truth data (thermal images + medical annotations) collected
• Machine learning models precision (specificity and sensitivity)
• Reduced amount of cases with secondary complications as a result of the rapid and wide screening
• Cost savings from the timely treatment.

Full-time staff: 7 people

Part-time staff: 5 people

Advisory board: 5 people

Georgi Kadrev, CEO

• Pioneered the image recognition API concept model in 2011
• Brought Imagga to the global innovator IDC spot in 2016
• Invited international speaker in AI, technology innovation and entrepreneurship.

Georgi Kostadinov, CTO

• Architected the largest commercial image recognition model with 100K+ categories, 150M sample images and 97% top-5 accuracy
• Forbes Bulgaria 30-under-30 for 2020.

Tanya Raynova, VP of Regulatory Affairs

Served as Head of Public Relations of the Supreme Administrative Court of Bulgaria for 8+ years.

Evgeni Kosev, VP of Business Development, Europe

Attracted and managed a total of $195M highly competitive public funding for innovation activities in the last 8 years.

Chris Georgiev, VP of Business Development, Asia-Pacific

Firestarter of the biggest startup events in Bulgaria and established AI networking events in Seoul.

Pavel Andreev, VP of Product

Product manager of the first global automated image recognition AI platform.

Ivaylo Mihaylov, Medical Data Coordinator 

Served as Manager of High Sport Achievement evaluations in the National Sport Academy for 3 years.

Maria Sharkova, Legal Adviser

Medical law specialization in the US.

Miena Stoycheva, Business Adviser

• Chair of the Digital Health and Innovation Cluster, Bulgaria
• CEO of JA Bulgaria.

Mladen Mladenov, M.D., Medical Adviser

Renowned breast cancer specialist in 1st Gynaecological Hospital “St. Sofia”.

Evgenia Alexandrova, Patient Affairs Adviser

• Chairman of the Association of Cancer Patients (APOZ)
• Patient advocate and consultant certified by ECCO, ESMO, ECPC, UICC, EUROPA DONNA
• National Representative of the European Breast Cancer Coalition EUROPA DONNA
• Contact person, CANCER MISSION, HORIZON EUROPE Framework Program.

Our team has always abided to the values of professional encouragement of people regardless of their gender, sexual orientation and ethnicity.

One of our co-founders is openly gay, our chairman and VP of regulatory affairs is a female, as well as 3 out of the 5 core members of our advisory board.

Our mother company Imagga has actively contributed and donated to Single Step, a local NGO for mental support of the LGBT community, and we have guest-lectured at multiple women empowerment events in our homeland of Bulgaria in the last three years.

We believe in the ability of technology to change lives for the better. For us MIT Solve is an example of an initiative that is committed to supporting this process.

We are grateful to have the opportunity to collaborate with Solve and its network on:

• Establishing research & clinical data collection partnerships
• Regulatory approval support
• Identifying and acting on funding opportunities, both public/governmental grants and private VC/corporate venture funding.

All of the above are important aspects of the development of Kelvin Health as a sustainable deep health tech venture.

Our key priority is collaborating with healthcare providers, typically university hospitals, on data collection, combining existing means of medical examination and diagnosis, cross-referencing them with Kelvin Health’s thermal imaging approach, to advance science.

As a result - empower our partners to extend their means of additionally examining, screening, and monitoring patients in order to get a better insight into their health status and take appropriate measures, when needed, in a timely manner.

According to WHO more than 70% of people in the US avoid timely examinations or have fear of visiting hospitals (even more now in COVID-19 times) for invasive or radiation-emitting procedures. In remote and/or underdeveloped regions the access to healthcare and specialists is very limited due to financial and/or logistical hurdles.

The above factors greatly increase the chances of serious health conditions being left undiagnosed/unmonitored for a period of time long enough to lead to severe complications. This undercare, that the patients implicitly opt for, saves some costs in the short term but leads to significant costs incurred in the long run when more severe stages of their conditions should be treated. Kelvin Health aims to significantly reduce this risk by providing a cost- and time-efficient screening and monitoring approach which leads to healthier population of patients and thus - better rating of the health care provider and cost savings for the healthcare system.

Refugees have very limited access to healthcare. With our accessible tool for early monitoring of health anomalies, currently focused on breast cancer, we assist in the rapid early detection of potential health issue, leading to adequate triage and allocation of specialist to provide quality care to the needy within a potentially massive group of people.

Besides pursuing career, “Sandwich Generation” mothers aged between 35 and 54 experience more stress than any other age group owing to the delicate and demanding balancing act of caring for their children and their ageing parents, which leave little to no space for preventive care initiated by them using the existing approaches.

The most used methodology for breast cancer screening currently is mammography. It is an X-ray image of the breast, and physicians use it to look for early signs of breast cancer. Unfortunately, many women avoid or delay recommended examination over the age of 40 years for reasons such as X-ray radiation, painful procedures, high breast density, unavailability in many medical centers, or uncovered costs in some countries. Moreover, recent studies outline that mammography screening does not play a direct part in the reductions in breast cancer fatalities.

Kelvin Health is an intelligent solution for screening and monitoring of breast cancer based on AI analysis of thermal images. Our goal is to provide women in risk groups with an entirely remote, safe, non-painful, and non-invasive solution to spot breast thermal abnormalities caused by pre-cancer conditions as early as possible and in the comfort of their home. Our main focus and goal is making the home preventive medicine the new cornerstone of the modern and personalized healthcare.

The anamnesis provided by patients as well as their existing EHR (Electronic Health Record), if any, are oftentimes incomplete and inconsistent. At the same time procedures like medical imaging are typically considered as overcare, based on the fact that most of them are expensive and radiation-emitting. In contrast, the minimal effort required to apply Kelvin Health leads to a better primary care assessment of newcoming patients that results in minimizing the chances of both under- and overcare.

Being a solution that is non-invasive, non-damaging, and with an unbound frequency of use, Kelvin Health allows regular short-interval monitoring of the conditions of patients in outpatient and even domestic (if the patients have access to thermal cameras compatible with the solution) setting. The associated cost of the Kelvin Health analysis would be an order of magnitude lower compared to other imaging approaches which allows for both cost-savings and more regular screening and monitoring. The continuously growing utilization of AI in process makes it even more efficient over time.