Developed by Nurithm Labs, the app addresses the problems of access and accuracy in clinical diagnosis for dermatological disorders/diseases, including skin cancers and oral cancers
Skin tech has undoubtedly earned its place in the sun — its popularity fuelled by the transformative potential of AI solutions and telemedicine services. With the deep penetration of smartphones and internet services, it was only a matter of time before apps gained centrality in the practice and delivery of dermatologic care.
The efficacy of Derma AidTM, a skin diseases diagnostic mobile application, reiterates that increasing availability of healthcare data and rapid development of smart analytics can revolutionise our healthcare system. “We have developed a machine-learning AI-driven algorithm, encapsulated in a mobile app, so that even a basic smartphone with 1MP camera can become a potent tool in skincare. Derma Aid will solve the problem of access and accuracy in clinical diagnosis for dermatological disorders/diseases,” says Sharad Kumar, co-founder and adviser of the Delhi-based start-up Nurithm Labs with 20-plus years of experience in leading cutting-edge technology development, academic research collaborations, and driven technology roadmaps.
The other co-founders are Vikas Chouhan, Merwin Rodrigues and Dr Jyoti Mathur. Vikas is an IIT alumnus with 10-plus years of experience in embedded systems, software design/architecture and applied machine learning. Dr Jyoti, with a PhD from Tufts University, comes with more than a decade’s experience in preclinical and clinical healthcare research. Merwin has spent over 20 years in new market development and go-to-market strategy.
Decoding Derma Aid
The technology behind the app is deceptively simple: A doctor takes photographs of the lesions on the patient’s body and sends them to the cloud server. Within 15 seconds, the app comes back with possible disease conditions based on the images. The actual work of diagnosis happens at the backend. The AI-ML algorithms hosted on the secure server provide dermatologist-level classification of skin diseases and treatment options. The spectrum includes precancerous, malignant, viral, bacterial and fungal conditions.
Though there are thousands of skin diseases, Derma Aid can identify only 50 of them as of now, with plans to get to 60-plus diseases by the end of this year. These 50 conditions are the most common ailments encountered by a GP/dermatologist. The app can pick up acne, psoriasis, vitiligo, tinea, eczema, and male alopecia in addition to basal cell carcinoma, a form of oral cancer, and melanoma, a type of skin cancer, with 80 percent accuracy.
A couple of measures can enhance the level of accuracy by an additional 5 percent, such as the use of multiple photos of the lesions and the doctor’s response to the questions posed by the algorithm on the non-visible conditions of the disease,” says Merwin.
“This is the future where AI will be increasingly applied within the healthcare field, especially for the tasks of diagnosis and treatment recommendations, given the acute shortage of qualified healthcare professionals and resources,” he adds.
Skin diseases are the fourth leading cause of non-fatal disease burden at the global level. According to Union health ministry’s estimates, there are only 12.5 lakh of allopathic practitioners in the country, of which only 3.71 lakh have specialist/post-graduate qualifications. So, the number of skin specialists in India would be a fraction of the total number of specialists.
Derma Aid is a clinical decision support tool for general physicians who can augment their capability and understanding of skin conditions, says Dr Jyoti. “This is particularly relevant since studies have revealed that the level of accuracy among GPs vis-à-vis skin diagnoses is 40-50 percent,” she adds.
The founders feel that the app can be handy in resource-constrained settings, especially in rural India where GPs are in short supply and healthcare workers must address the immediate needs of rural and marginalised communities. Derma Aid, says Dr Jyoti, can be deployed to screen oral cancer, widely prevalent among Indian men, where 60-80 percent of the patients are diagnosed in advanced stages when the condition has already become acute.
Idea to Application
The idea for the app was conceived three years back when the founders thought of using AI to meet healthcare needs. Skin diseases became an obvious choice because of their wide prevalence, insufficient diagnostic accuracy levels, not-so-stringent regulatory approvals and ML-amenable solutions.
But what firmed up their decision was a prompt, encouraging response from Dr (Prof) Somesh Gupta at the Department of Dermatology and Venereology, AIIMS, New Delhi, when Sharad reached out to him for advice. “Two days after that exchange on LinkedIn, I visited him at AIIMs. A month later, in March 2018, we signed an MoU with AIIMS,” says Sharad, adding that with Dr Somesh coming on board, access to data from AIIMS became easy. The company was formed two months later.
In October 2018, Derma Aid had a pilot run at AIIMS with just 10 diseases. In April 2019, the app became public with 20-odd conditions and by October that year, the list had grown to include 40 diseases. By March 2020, the app had 50 diseases in its bank.
Business Model and Scale-up
Nurithm has tremendous growth potential, given that there are hardly any apps for serious skin disease diagnoses in India. The ones currently doing the rounds are “skincare coaches” for cosmetic treatment. While there is no definitive study on the prevalence of skin diseases in India, a 2018 study pegged the then market size of the dermatology segments in India at Rs 1950 crore, growing rapidly at about 11 percent.
The start-up is in talks with two pharma giants who manufacture and market several derma products recommended by skincare specialists and GPs. “We are exploring the possibility of bulk licensing of the app with these companies. They would like GPs and specialists to avail of this tool to increase their accuracy and efficiency,” says Merwin. Nurithm has also teamed up with University Hospital of Basel in Switzerland as part of a grant from Foundation Botnar, a Swiss foundation, for automated diagnosis of skin diseases in Africa. The data sets from the continent will increase the startup’s disease portfolio and eventually help find untapped markets in African countries. Nurithm is also optimistic about its association with the Swiss university opening work opportunities in Europe. The possibilities of tie-ups with telemedicine portals are also being explored.
Social Alpha’s role
The founders began interacting with Social Alpha in 2019. “It was the first organisation to believe in the social and commercial merit of the solution that Nurithm Labs was building. Social Alpha could foresee the role Derma Aid would play in reducing inequity in healthcare delivery – either as an augmentation tool for physicians or as part of an integrated solution in telemedicine,” says Merwin.
Early on, he was impressed by the “knowledge and dedication” of the Social Alpha team. He even told the other founding members that Social Alpha would be an ideal partner for this venture because of “likemindedness” – a shared value system, perhaps. “Nurithm has benefitted from Social Alpha’s timely advice – particularly from Dr Kshama Kothari with whom we regularly brainstormed. She also guided us on business models, potential partners, and ecosystem connection,” he says.
Nurithm Labs fits Social Alpha’s investment thesis that looks for digital health solutions focussed on improving primary healthcare. “By augmenting the capability of general physicians and non-dermatologists, the product is providing dermatology services in places where expert services are unavailable,” says Dr Kshama Kothari, Clinical Director, Medtech, Social Alpha. Nurithm’s training data is built, keeping in mind the ethnic diversity of the Indian population. It means that the colour of the skin is not a barrier to the product’s accuracy.
It is also addressing the training gap and creating online access for assisted diagnosis. A simple tool like this, which can be hosted on a smartphone in the hands of a last-mile community healthcare worker, can improve the rate of early diagnosis of various skin diseases, including skin cancers and oral cancers, and guide towards accurate treatment or referral to a higher centre, concludes Dr Kshama.