Q&A: The startup aiming to improve cancer diagnoses through AI-based analysis of biopsies

Web content editor Ian Bolland caught up with Avi Veidman, CEO of pathology startup Nucleai which aims to make cancer diagnoses more accurate, accessible and efficient through AI-based analysis of biopsies.

Give us an idea of the ‘comprehensive suite of solutions that Nucleai offers?

Nucleai provides a comprehensive suite of solutions for cancer biopsies relating to gastro, breast and prostate – and soon, skin – encompassing 85% of biopsies. Because of this product range, Nucleai answers the full spectrum of needs posed by laboratories within a simple web browser instead of through purchasing multiple-point solutions. Nucleai’s solution is cloud-based and can integrate into pathologists’ workflow to help in several ways. The solution’s post analysis (‘digital quality assurance’) reduces the potential for missed or erroneous diagnoses by comparing manual and automated reports and highlighting abnormalities; pre-analysis (‘digital resident’) highlights suspected areas and writes reports for the pathologist to review and approve. Backed by over 20 million slides and supported by pathologists in the US and in Israel, Nucleai pilots have proven a 97% accuracy rate for diagnosis.

How does the AI technology used by Nucleai increase the standard of care in oncology?

Firstly, I’d like to briefly explain why AI is the lifeblood of our team and company. I spent most of my formative professional years (20 years) serving in various Technology Units in the Military Intelligence Corps of the Israeli army, eventually serving as the head of an AI department. In this role, I worked with satellite and aerial images, with the aim to augment image analysts and detect different objects within those images. After retiring from service, a personal experience led to my understanding that the standard biopsy processes needed disruption- and that my experience in AI and image analysis could be applicable to solve this issue. Our team is built of AI and healthcare leaders who have driven our unique technology, and who have made its development possible.

In short Nucleai’s pathology platform will make standard of care diagnostics much more accurate, efficient and accessible in the rural areas that doesn’t have all the subspecialties pathologists.

We provide an offering that we’ve termed “Personalised AI”. It is a two-part approach which is personal to the user, and contextual to the working environment. Our research revealed that physicians are afraid of AI – so personalising it helps gives clinicians a more comfortable feeling in understanding that they can integrate this technology into their workflow, in a tailored way that will improve efficiency and suit their needs. Nucleai will never replace physicians, instead it will adapt into their work flow in the most personal way possible.

The second aspect of ‘Personalised AI refers to the environment in which it functions. We found out that different pathologists and organisations have different ways of determining diagnosis. Whilst we already have a unique technology that is trained by thousands of slides, with an accuracy rate of 97% and higher, the decision makers in labs are free to adapt Nucleai to their methods of diagnosis in their lab.

Give us your view on the pathology bottleneck as you call it? What do you think causes it and how does Nucleai address it?

Biopsies diagnosis are the main diagnostic procedure that can conclusively diagnose cancerous tissue. Pathologists are the sole owners of biopsies diagnosis, but research shows their numbers are rapidly declining, creating a gap in supply and demand.

The number of biopsies is increasing dramatically with no signs of slowing; however, there are limited pathologists available. Beyond that, most pathologists are highly specialised, narrowing the accessibility to high quality pathology service further.

A key factor behind the specialist shortage includes the aging and retirement of senior pathologists, contributing to a predicted decline in pathologists to 3.7 per 100,000 people by 2030. Meanwhile, demographic, environmental, and lifestyle factors are sending demand ever higher for key services, including biopsies. Given this dynamic, U.S. demand for pathologists will reach 20,000 in the intermediate term, while the projected supply will be less than 15,000.

You say Nucleai isn’t only geared towards pharma and research – can you outline the kind of work you do with those sectors?

Nucleai is a clinically oriented company, so while existing solutions in the market focus on either research or on pharmaceutical orientation, Nucleai’s approach is geared to actual market use, aiming to help real patients and pathologists by seamlessly integrating into existing processes.

The market for pathology tools is only growing, with potential in a variety of related disciplines including pharmaceutical companies developing new drugs, for broader medical research, and for direct clinical use. Our long-term vision is to develop AI-powered prognosis and prediction capabilities for cancer treatment, providing an end-to-end platform for healthcare providers. Every year, hundreds of thousands of patients receive non-beneficial treatments. In addition to discomfort and frustration on the patient’s side, this also results in huge amounts of money spent on treatment that doesn’t work. Nucleai’s enhancement technology will use AI to power prediction tools to identify and suggest the most relevant treatment path for patients.

What are the main achievements of Nucleai, in your opinion so far? And what are the future plans?

Our greatest achievement has been creating an effective pipeline for rolling out products at a rapid pace; we’ve launched a new product every three months, each product with a very high performance. We are in the midst of pilots and partnerships with various labs in the US and Europe, which will give us a greater footprint with more traction.

We’re looking forward to some exciting opportunities on the horizon. We have been approached by Big Pharma players looking to gain access to the best companies creating diagnostic tools. Personalised medicine is the next big thing and for this, they want to know which drugs are suitable for each individual patient. To know that their tests could use a technology like ours inspires us and make us very excited for what the future holds.

We are also in the midst of various pilots with leading labs in the US that are using our system in their flow. Ultimately, our growth lies in our unique strength as a tech company to create a complete offering for hardware and software through exciting strategic partnerships that will impact the lives of millions of people in a deeply profound way.

Link to article