Newage health management Dinesh Koka considers the twelve years he worked in GE Healthcare, across different P&Ls businesses like the Healthcare IT vertical, Medical Equipment, Life Sciences and Biotech, as a great learning. When he set up and ran the public-private partnership business for GE, he worked in close quarters with state governments, setting up outsourced services projects in low-resource areas. Surveying district hospitals and assessing the affordability of technology there set him off on an entrepreneurial path, and joined hands with Vikas. Vikas Ramachandra had taken the research track after his engineering degree which took him through a Ph.D. in Signal processing and then stints in the data science stream with the best in the business like Nvidia, Qualcomm, and then in healthcare AI through startups and companies like Roche. Through his work, Vikas had built an appreciation for the challenges that clinicians face everyday in making decisions related to patients and the kinds of data they consider while doing that. The idea for their startup, Onward Assist — a set of analytics tools, which are built to assist cancer pathologists, radiologists and oncologists—came while they were working on their previous startup idea. “At the time, we observed the limitations oncologists face when trying to take cancer treatment decisions,” he says. “The seed of what we are doing today was planted during that period.” The Indian healthcare market is projected to be valued at $372bn by 2022 and the time is ripe for healthcare startups to offer innovative and quality solutions for various problems plaguing the industry. Sensing this opportunity, Dinesh and Vikas co-founded Onward Assist. The startup was initially set up as a patient engagement platform that was focused on medical adherence and helping patients to manage and complete their medications, other post-discharge instructions on time. This also allowed doctors to monitor their patients remotely. “The goal was to use machine learning to collect patient behavioural data and enhance patient adherence,” says Dinesh. The pivot Around the end of 2017, the co-founders embarked on a pivot that led them to develop a clinical decision support tool for oncology. The user of the new platform is the doctor and not the patient, as was earlier the case. “Initially, we had done it for orthopaedics and cardiology. Later, our conversations with oncologists convinced us to take a turn,” explains Dinesh. “We believed this would be a better place for putting our combined capabilities to use. That is what got us here.” Oncologists while deciding on the course of treatment for patients, take inputs from radiologists and pathologists. (While radiologists study images from equipments such as Xrays, CT scanners and MRIs, pathologists study biopsy slides and other IHCs, related tests.) Today, the company’s key focus is in diagnostics assistance in computational pathology (specifically in breast cancer and cervical cancer). These tools are built to assist the pathologist, and help extend such a service even in a low resource setting. Building the tool After the business model was pivoted, Dinesh and Vikas started building prototypes for the app. Soon, some of the best cancer institutes came on board to partner with Onward Assist. After gaining the trust of the medical community, the co-founders spent over one-and-a-half years working with clinical partners and building specific tools for cancer diagnosis. Having scored high on accuracy, the company is now confident to go to market and commercialise the product within the year. The Onward Assist tools were developed after extensive research and multiple iterations of clinical tools that were shared with clients for validation and feedback. Additionally, Dinesh and Vikas consulted a number of doctors across specialities— radiology, histopathology, oncology—and across the country for the iterations. The analytics tools are being built in collaboration with oncologists—treating oncologists who are at the head of the decision tree when deciding the line of treatment for the patient. These tools would ideally be positioned for the “tumour board”, a collection of doctors from different specialties who come together to take a decision on how to treat a patient. These tools look at multiple data types about patients. For instance, one of the machine learning-based predictive tools they built for a cancer hospital takes in around 250 types of data about the patient, including clinical history, lab tests data, genetics and lifestyle, among others. The individual treatment options are arrived at based on the predicted treatment response. “We tell the doctors, ‘you can use our tools to take better decisions, but you’re still at the head of the decision table’,” says Dinesh. “That’s why the platform is called Onward Assist—it’s more about assisting doctors in their clinical decisions.” The company has built three main types of tools: How the tool works Onward Assist is a web platform that runs off the cloud. The Artificial Intelligence (AI) tools for diagnostics assistance can also be provided as an API (Application Program Interface) which can be accessed through the reporting workflow solutions. For example, the radiology AI tool can be used through an existing radiology / DICOM reporting workstation. Likewise, the pathology AI tool can be used through the online pathology reporting solution. The web platform enables the biopsy slide image to be uploaded onto the AI tool that gives the processed image, along with the respective annotations and quantifications with respect to the specific biomarkers and the types of cells, among other parameters. The Onward Assist tool looks at the biopsy slide image, carries out automated diagnostics and also extracts additional clinical insights from it. Histopathologists provide critical insights as part of their report, which are used by oncologists. Due to the sheer volume of cases, histopathologists are pressed for time to study biopsy slides that range anywhere from 20x to 40x magnifications or in some cases higher. Vikas identified this as a classical problem that Computer Vision (CV) techniques could address. Therefore, Onward Assist tools are built to reduce the amount of time histopathologists spend studying biopsy slides. They also gather additional insights about the tumour cell microenvironments that helps oncologists understand the tumour better. Further, the tools have the ability to process a huge cohort of patients, and this wisdom is presented to the doctors to enable them to make the final decision. Growth and tapping into resources Onward Assist is part of IIM-A’s CIIE (Centre for Innovation Incubation and Entrepreneurship) seed program, and a member of the Yale Sustainable Health Initiative in India, run by the CoWrks Foundry, the accelerator run by CoWrks. Prior to this, the company had raised seed funding from IIIT Foundation (International Institute of Information Technology Hyderabad), where they are a member of the MedTech cohort at Ojas Incubator. The company’s marketing approach has been primarily led through a robust network of clinicians. Word-of-mouth referrals from clinicians has enabled the company to build new partnerships and get initial validations for the product. The co-founders’ aim has been to get new technology to healthcare providers through clinicians. The startup directly approaches doctors and hospitals to build its network and has not yet leveraged social media to expand its base. Onward Assist has also hired additional sales talent towards its expansion strategy. According to Dinesh, the impetus to join T-Hub’s Lab32 program around nine months into Onward Assist’s journey was to leverage the opportunity to be part of the vibrant Telangana startup community, and tap into the Corporate Innovation network that is pioneered by T-Hub. There was a conscious effort by the co-founders to connect with corporates and healthcare experts through the T-Hub platform. Since Lab32 is a structured way of touching upon everything critical for a startup, the company joined the incubator to get access to additional resources that would fuel its growth. For instance, Onward Assist had the opportunity to pitch to big healthcare players such as Novartis and Medtronic, during the incubation program. Dinesh adds that the other benefits accruing from Lab32 were the interactions with other startups and attending sessions by various experts. Teething problems Among the challenges the company has faced enroute its entrepreneurial journey has been the longer business cycle of the healthcare industry. Also, doctors don’t naturally adopt new technological knowhow since their clinical decisions are largely based on their wisdom, experience and training. Since doctors’ time is at a premium, getting quick validation from oncologists for Onward Assist’s tools was a bit of a struggle. Further, the niche technology deployed by the company required several rounds of iterations that could be successfully completed only with the support of all the participating doctors. Since the company was looking at actual clinical data, the team had to work hard to get their algorithms right. Moreover, given today’s data privacy issues, Onward Assist had to go through the necessary regulatory approvals before even starting to process patients’ data. The onward journey Despite the various challenges, Onward Assist’s tools have received validation from pathologists that affirm that the company is on the right track. The company seeks to expand the team and enhance its research and product development capabilities. This would include partnering with more clinical channels, cancer institutes and cancer diagnostic service providers. Dinesh and Vikas believe it’s important for entrepreneurs to pay it forward and mentor other healthcare startups. “The advice I usually give to healthcare startups is to come in with a clear understanding of the healthcare industry,” sums up Dinesh. “If you haven’t worked in this industry, have a doctor or a clinician or an advisor as part of your team—right from day one. There are enough examples where we build solutions based on our consumer experience, but they need not be something that solves problems for millions.” ________________________________________________________________________________________________ Innovation geared to combat COVID-19 Onward Health has been at the forefront of innovation to meet the challenges posed by the Corona virus. Given below are the key innovations the startup has built to leverage the strong capability that had in-house and built socially relevant technologies that can extend a helping hand in the current times. 1. RT-PCR workflow, powered by Telepathology: RT-PCR testing is considered as the gold standard for Covid19 diagnosis. However, technicians and microbiologists continue to bravely go through the testing workflows as well as for reporting, in spite of challenges faced, such as the requirement to report on a huge set of tests on a daily basis given the high testing load in the present times and the risk of exposure in the lab facilities. To mitigate such risks, the startup has leveraged automated analytics, combined with telepathology, to help the microbiologists to enhance their effectiveness manifold. Onward Health also provides an automated analytical tool that works along with the RT-PCR machines. The tool enables the microbiologists with the analysis of the gene plots to swiftly complete the reporting process from remote locations. The company is implementing RT-PCR testing at one of the largest lab networks in India. 2. Rad-Assist Mass testing of even a small percentage of India’s population, using RT-PCR would come at a significant cost to the exchequer. In addition, the acute shortage of RT-PCR kits calls for an alternate but effective solution. COVID-19 diagnosis using chest radiographs (X-ray) and CT scans is currently being considered for large scale deployment, especially since recent studies have shown much promise with radiology. With this in view, the Rad-Assist tool from Onward Health has been developed to assist radiologists in evaluating radiology images for detection of disease patterns. Onward Health has developed this with expert advice from expert radiologists in leading hospitals. Rad-Assist uses a couple of machine learning models. The first one is capable of classifying patient cases between normal and cases of viral pneumonia caused by the novel coronavirus. The second model is capable of highlighting the specific regions in the breast that need to be investigated further. 3. TelePathology Onward Health provides a TelePathology solution to improve access to good quality diagnostics. The solution enables hospitals, labs and pathologists to provide pathology diagnostics services like a second opinion from a doctor, to smaller labs and remote labs. Since the current times pose challenges with regard to modalities, therefore, pathology services may be affected in clinics and hospitals. This will result in longer reporting timelines or pathologists being unavailable to report as and when required. To address this issue, Onward Health came up with the TelePathology solution that enables pathologists to remotely report on the cases that are queued up while simultaneously limiting their exposure to COVID-19 in hospitals and clinics. We at T-Hub help startups gain access to better technology, talent, business networks and funding. Be a part of T-Hub Community. Join us here.
Related Post
- March 31, 2024
- March 31, 2024
- March 31, 2024
