Amid the COVID-19 pandemic, Maya Nathani faced a deeply personal health crisis—a cancer diagnosis. This experience gave her profound insight into the struggles of vulnerable populations, particularly seniors. As she recovered, Maya observed firsthand the challenges older adults face daily: complex health conditions, limited mobility, and restricted access to care. Inspired to make a difference, she founded Universal Health Hub (UHH). “We have been serving over a thousand seniors a month … and we realized that …seniors could have avoided so many foot conditions if they could have diagnosed these cases sooner [because] our service is actually called upon very late once the condition has worsened.” Recognizing this accessibility issue, her team saw an opportunity to develop a predictive and easy digital system and WIMTACH was the right partner to tackle the challenge. WIMTACH worked closely with UHH’s in-house IT talent to train the Aging Feet Condition Detection with Artificial Intelligence (AFCDAI) machine learning proprietary model using five years of anonymized patient data.

Maya entrusted the WIMTACH research team with developing an Artificial Intelligence model, designed to assist with identifying early signs of foot complications. “Once we connected with WIMTACH, we saw the passion in them to bring in innovation into the healthcare [industry],” she said. “It’s the same passion that … Universal Health Hub has in solving these problems for this vulnerable population in the community. So, there was a synergy.”

Together, the two teams are developing an Artificial Intelligence model for a feet condition image classification system which will be integrated into a UHH mobile application using a machine learning model. When a picture of feet is submitted to the app, the AI integration will detect Corn disease, Calluses, Long Nails, and Fungus to signal when to consult a healthcare professional. The mobile application will also enable patients to find a health care provider near their home and assist with scheduling an appointment for a visit to get their concerns addressed. The goal of this model is to improve proactive and timely detection and management of aging foot conditions that can be indicators of more severe health complications like diabetes. In fact, a CDC report estimated that 60% of diabetic foot ulcers result in hospitalizations due to infections. Studies suggest that early foot condition detection can impact broader health outcomes including reducing the likelihood of falls, preventing and lessening foot deformities such as bunions, hammertoes, and malalignment. The UHH model is designed to be used during regular foot examinations to alert users when changes in foot health suggest substantial health risks, giving patients and caregivers a proactive approach to tailoring care.

The technical teams are using empirical studies of AI models to design the feet condition classification system using data that has been provided by UHH. In this initial phase of the project, the research team is investigating how the data and model perform without any model optimization or hyperparameter tuning to gain a clear evaluation of the raw capabilities of the data and model combination for future enhancements.

To this end, the team is analyzing the UHH’s dataset and investigating the best approach to generating synthetic data to effectively train the initial model and address any missing information. Student Researcher Rodolfo Cota Borbon is heavily involved in the process of cleaning, preprocessing, and analyzing of this dataset. With this process, he is able to identify key features and patterns of the dataset to understand the relationships between the data points and their relevance to the classification task. “We kind of realized very [early] that WIMTACH had the talent and the technology to kind of help and support us with our idea and innovation,” said Maya.

The partnership has been very smooth and Maya has been pleased with the development process. “The collaboration is great, we love the way we are communicating with one another,” she said. She explained that the WIMTACH team was able to quickly adapt to the needs of the project and develop the system quickly, while ensuring not to compromise on the quality of work produced. Her team was also impressed by the ideas offered by the WIMTACH research team in shaping the end product of the system. “Great, great contributions by the [research and development leads and students] at WIMTACH has helped us expedite this project and work on solving this problem,” she said.

With this AI model, Universal Health Hub hopes to innovate the senior care industry and transform the care of vulnerable populations. They are aiming to use this classification model to collaborate with the ministry of health, long term care homes, and retirement homes to reduce the number of emergencies resulting in emergency admissions, fall incidents, and worsening foot symptoms of serious conditions. As a result, the software will enable facility operators to also include diagnosis reports in their care record management systems with the features to upload and send reports to family doctors or pharmacies.

“Our main focus is to improve the lives of seniors and their health … and avoid complications,” she said. Her team has plans to continue the partnership with WIMTACH as they transition to become a technology first health care company. “WIMTACH has exceeded the expectations we had for them on the target that we wanted and we had in mind to achieve,” she said. “The kind of experience we’ve had working with WIMTACH has only left us with a positive note and positive experience to collaborate again with WIMTACH on something similar because the collaboration has only helped us with going faster and getting the latest and greatest of technology and skillsets from the industry.”

Learn more about Universal Health Hub:
500-7030 Woodbine Ave, Markham, ON L3R 6G2
Ph: 416-848-7032