How NIRAMAI’s AI-Powered Device Is Making Breast Cancer Screening Affordable For Women In India

Why are artificial intelligence, machine learning and neural networks drawing the attention of the most brilliant scientists, sharpest entrepreneurs and billions of dollars in funding? It’s simple, really. AI can change the world and cancer research is one of the clearest examples of just what a crucial role this technology is going to play in one of the longest medical battles in human history.

Even in India, over a million new cases were reported last year of cancer and according to data released by Global Cancer Observatory (GCO), out of over 1.15 Mn new cancer cases reported in India in 2018, more than 10% or over 162K were breast cancer cases. The number of deaths caused by breast cancer in 2018, in India, was over 87K, which shows the severity of the issue facing Indian healthcare providers in detecting and treating the rising prevalence of breast cancer cases in India.

Breast cancer is the largest cause of cancer deaths in women today. According to WHO, one in every 12 women is at risk of a breast abnormality. Indian women have only about 50% chance of survival, which is why early diagnosis is highly critical to decrease mortality rates.

The current technologies, that are not suitable for frequent screening and have portability issues, do not work on women under 45 years of age. Today 90% of the women on a global scale detect breast cancer by hand; they accidentally find a lump and then go to the doctor.  This late detection can be avoided with good technological intervention. Here is where NIRAMAI has stepped in, using technology and AI/ML algorithms to help women detect breast cancer at an early stage, giving them a better chance of surviving the disease.

From Tragedy To Saving Lives: The Story Of NIRAMAI

Established in July 2016, NIRAMAI was co-founded by Dr Geetha Manjunath and Nidhi Mathur. With a Ph.D. from the Indian Institute of Science and has over 25 years of expertise in the IT innovation at Hewlett Packard Labs and Xerox Research, Dr Manjunath was well versed in how science and technology can work together. Mathur had been a senior product manager at Xerox Research and was responsible for finding product-market fit for technology research being conducted at the company. As such, she was also updated on what technology is catching on in the market. The duo made a great fit.

Dr Manjunath also had a stint at Xerox Research as a senior director, where she researched multiple use cases of AI, and that’s when tragedy struck as she lost two young cousins to breast cancer. This heartbreaking incident became the catalyst for her to start working on technology that help intervene and save lives. In that state of mind, amidst a discussion with another colleague, she learnt about thermography and its ability to detect abnormalities early.

With a small research team, Dr Manjunath started analysing thermal images of cancer patients. Going a step further, she postulated to use this on normal women who do not have symptoms, thereby enabling early detection. With the help of the team, she deployed AI algorithms to address early detection. With early promising results, Dr Manjunath decided to do this full time and created NIRAMAI with Mathur, as well as team members Himanshu Madhu and Siva Teja Kakileti.

Simplifying Cancer Detection To Boost Survival

While it sounds like an ancient Sanskrit phrase, NIRAMAI actually stands for “Non-Invasive Risk Assessment with MAchine Intelligence”. The non-contact procedure screens for breast abnormalities. According to the company, it is the only privacy-aware automated solution that can detect early-stage breast cancer. The patient is made to cool down in a room for 10-15 minutes before the actual screening takes place. The procedure takes less than five minutes. The NIRAMAI solution is based on thermography, wherein a thermal device is placed three feet in front of the patient, which captures five thermal images. These images are then automatically analysed by NIRAMAI’s software tool, which sends the analysis report to a senior radiologist for certification.

The core of NIRAMAI solution is Thermalytix, the computer-aided diagnostic engine that is powered by artificial intelligence and machine learning models. The solution uses a high-resolution thermal sensing device and a cloud-hosted analytics solution for analysing the thermal images, checking it against its database of positive and negative reports. Its SaaS solution uses big data analytics, artificial intelligence and machine learning for breast cancer screening.

Why Thermalytix Is Better Than Mammography

According to Dr Manjunath, although mammography is the defacto standard for breast cancer screening, it has certain limitations such as low sensitivity for women under 45 years of age. Since it is X-ray based, the test cannot be performed more than once in two years. It is also not affordable for many women.

Thermalytix is radiation-free, and works on women of all age groups, on the other hand. “Given that it is portable and low-cost, the solution makes cancer screening more accessible and affordable to all, Thermalytix has been shown to detect very small tumours as small as 4 mm,” Dr Manjunath tells Inc42.

Though Thermalytix can be applied to several types of other diseases, NIRAMAI’s focus is on breast health at this stage. In the future, the company will go beyond breast cancer screening to general breast health and diagnosis and prognosis of other breast diseases. Earlier this year, the FDA in the US issued a warning about the usage of thermography as a substitute for mammography. So how is Thermalytix different from thermography?

Thermography is a technique wherein heat radiated from the body is captured. Due to the high metabolic activity of cancer cells, greater heat is radiated from the affected region. This can be visualised as an abnormality in thermographic images.

Dr Manjunath explains that the use of Thermalytix for breast cancer screening is actually FDA-approved as an adjunct modality and has been used by doctors around the world. The recent FDA warning was to dissuade misuse of thermography by some clinics that claimed it as a standalone modality. Manual interpretation of thermal images is very complex as the radiologist needs to look at millions of colour data points to decide if a subject is abnormal or not.

Thermalytix, on the other hand, uses machine analysis, instead of manual interpretation, to generate a quantitative thermal analysis report. In addition to thermal images, NIRAMAI’s Thermalytix has a unique algorithm to detect and extract blood vessel structures from the images that help in determining deep-seated tumours, that are otherwise missed in thermography.

“We also use many other demographic data from the patient for our prediction. We have clinical publications that show good accuracy of the technique, comparable or even better than the current standard of care. Still, Thermalytix report will be reviewed by a doctor to determine whether a person is malignant and requires further followup. It is an aid to the doctor,” adds Dr Manjunath.

NIRAMAI is currently marketing the Thermalytix-based cancer screening solution only in India and has the regulatory clearances for it. The thermal imaging device is already approved by the US FDA as well. The company is working on an FDA application to submit its clinical evaluation results and get a clearance for Thermalytix software for the US and other markets with similar regulations.

Moreover, Thermalytix is not limited to breast cancer, which is why NIRAMAI is banking on the success of this application before taking it to other diseases. As an experimental research project, NIRAMAI is also working on using Thermalytix technology to combat the spread of river blindness. It is an infection caused by a parasite called parasitic worm Onchocerca volvulus transmitted by the bite of an infected Simulium blackfly, causing skin diseases and even blindness.

The live adult worms of Onchocerciasis are detected by using Thermalytix. As it is a non-invasive method, it is expected to help assess the efficacy of new drugs being developed to control river blindness by killing the adult worms. The Bill & Melinda Gates Foundation has provided funding support to NIRAMAI in this new endeavour and also access to medical experts in Ghana who understand this disease well.

The other major tool that NIRAMAI works with is SMILE (short for Software with Machine Intelligence for Life Enhancement). It is a web interface for the NIRAMAI-certified technician to upload demographic information about the patient along with the subject’s thermal images. This is processed to analyse the subject’s breast health condition. Their solution automatically generates a report listing certain unique parameters obtained from patient thermal images and also recommends a breast health score.

Dr Manjunath says NIRAMAI currently imports its thermal hardware from Sweden. “However, our solution can work with any good thermal imaging device that satisfies our imaging specifications. There are some vendors in India too with whom we are working to see if they can be our future partners.”

Clinical Trial By Fire

As India has one of the most stringent clinical research regulatory environments in the world, running the necessary clinical trials would have been a time-consuming and arduous process for NIRAMAI. Dr Manjunath agrees that it most certainly was.

“In India, like all processes, just starting up a clinical trial takes a lot of time as there are strict procedures and processes involved.”

She explains that is imperative to obtain a clearance from the Independent Ethics Committee, which is a group of experts from different fields who review the proposed protocol of the trial and ensure the validity of the procedure to test the claims made, as well as ensure that no harm will be caused to the patients in the process and so on. Patient recruitment also takes time as often women are hesitant to partake in a trial, even though it’s a harmless and private process.

“Conducting clinical trials on an AI-based product is a continual process. Since we keep evolving the AI model and continue to improve the accuracy or applicability of the solution to many other subcategories of diseases or other diseases, clinical trials are very much as the organizational need for AI-based health tech companies,” she adds.

NIRAMAI conducted trials at Narayana Hrudayalaya and HCG Oncology. It has also employed an independent expert CRO, who ensured proper clinical guidelines for the trial. Other hospitals have shown interest in the trials too, Dr Manjunath tells us and NIRAMAI is in different stages of the process in multiple other locations. “In parallel, we do retrospective trials, and as and when we get good results we release the model to production and move it to our commercial pipeline.”

Further, NIRAMAI has regulatory clearance from the Ministry of Health and Family Welfare, India, Drug Control General of India (DCGI), and Central Drugs Standard Control Organisation (CDSCO) for the overall Thermalytix solution. It also has permission to conduct screening tests in government hospitals in Bengaluru as well as a few other states.

NIRAMAI: Breaking Gridlocks And Glass Ceilings

Every new beginning faces some challenges, and so did NIRAMAI. The first and biggest challenge faced was raising funds to buy high-resolution thermal sensors and start serious experimentation.

NIRAMAI found a partner in pi Ventures, an AI-focussed fund, which believed in the technology and the team and decided to lead the seed round, with co-investment from Axilor Ventures and Ankur Capital.

“The next challenge was to convince hospitals and doctors to try the solution. Though the hospitals and the doctors saw that the technology had advantages, NIRAMAI had to prove that their solutions actually worked. They set up randomised clinical trials to compare their results with the current standard of care,” says Dr Manjunath.

Talking about the current challenges that the healthtech startup is still struggling with, she states that getting a structure in place so as to enhance its scalability quotient and reach out to more hospitals is still a bit of a challenge. The key component of this challenge is the time involved in convincing the doctors. Although NIRAMAI currently has six international publications showcasing the workability of its technique, every reputed doctor wants to conduct a trial before usage.

Mindset is also a key issue that NIRAMAI faces. Although breast cancer awareness is steadily increasing in India, the number of women who come forward for preventative screening is minimal.

“Change in the mindset to wellness verses sickness is the need of the hour.”

Currently, NIRAMAI has over 30 installations at hospitals and diagnostic centres across 10 Indian cities.  It has conducted screenings for over 12K women and organised over 50 employee wellness camps. It also partners with NGOs and cancer societies to conduct free screening camps for the underprivileged.

Focus On Driving Acceptance, Not Revenue

Apart from the publications and patents, NIRAMAI has also won several national and international awards. It has received support and recognition from Biotechnology Industry Research Assistance Council (BIRAC), Startup Karnataka, Accenture, Philips, Google and Amazon as one of the top startups in its domain.

NIRAMAI has also won the Best Preventive Insurance Idea award in Milan and the Gold prize in Hack Osaka 2019 International competition held in Japan. It is also the only Indian company listed in the 2019 COHORT of AI 100 startups in the world by global business data intelligence platform CB Insights. Last year, NIRAMAI was recognised as one of the 42 most innovative Indian startups of 2018 by Inc42’s 42Next list.

Talking about the company’s revenue streams, Dr Manjunath states that its primary revenue stream is from its managed breast health solution for hospitals and diagnostic centres. It also has a screening operations wing where it conducts corporate camps with per-day lease charges.

And when asked about revenue projection for FY 2019-20, Manjunath said the focus is on creating more awareness for AI solution in disease detection. “Having proven early revenues, our current focus is to drive acceptance in the medical community and get international regulatory clearances – not so much on the revenue at least for the next few quarters.”

Road Ahead For NIRAMAI And Healthtech

Starting with $1 Mn in seed funds raised through Pi Ventures, Ankur Capital, Axilor Ventures, 500 Startups, and Flipkart cofounder Binny Bansal in 2017, NIRAMAI moved forward to raise $6 Mn in Series A funding earlier this year from Dream Incubator, Beenext and existing investors. At the moment the company’s focus is on business expansion, which it has already successfully kicked off over the past year.

From only one installation in Bengaluru last year, NIRAMAI has grown to have a presence in ten cities across India. This year, it plans to expand to more cities. It is also looking to partner with various corporates, diagnostic centres, hospitals, NGOs, government hospitals and more for installations and conducting screenings.

In order to ease its sales pipeline, the company is also looking to partner with an established medical device distributor. “We now have a bright team of 20+ innovators and an able leadership team.  We are currently focussing on hiring the top management so that they can build their team for expansion. We are working on hiring an Engineering Director and Chief Medical Officer,” Dr Manjunath adds.

In India, NIRAMAI competes with a number of cancer-centric health-tech startups such as Zumutor Biologics INC, Mitra Biotech, OncoStem Diagnostics, Sascan, Exocan, and Onward Health, among others. However, Dr Manjunath believes NIRAMAI stands out from the competition — most devices detect lumps but all lumps are not cancerous, thereby resulting in a lot of false positives.

“Our solution is privacy-aware too [besides being accurate] – where no one sees or touches the person, making it very comfortable for the lady.”

Note: We at Inc42 take our ethics very seriously. More information about it can be found here.