Solution Overview & Team Lead Details

What is the name of your organization?


What is the name of your solution?


Provide a one-line summary of your solution.

MendelScan is a software medical device used to detect patients who may have unsuspected and undiagnosed rare diseases. It does so by identifying cases with commonly overlooked patterns of signs and symptoms in clinical settings.

Film your elevator pitch.

What specific problem are you solving?

The problem: Late diagnosis of rare diseases. 

On average, it takes 5-7 years to receive a rare disease diagnosis. For some patients, the diagnostic journey can stretch as long as 30 years or result in no diagnosis at all. 

This is significant as the collective impact of rare diseases is considerable (despite individual disease rarity): c. 3.5 million patients in the UK (around 5% prevalence), and 350m+ globally. 

Impact on patients

Negative impacts on patients are threefold: (1) limit access to timely treatments, miss opportunities for clinical trials; (2) deterioration of prognosis – potentially causing irreversible harm; (3) negative impacts on quality of life due to prolonged periods of unexplained illness, anxiety, disruption to livelihoods, and multiple invasive tests.

Impact on healthcare systems

Late diagnosis incurs avoidable costs during lengthy diagnostic journeys. Imperial College demonstrated that in the NHS pre-diagnosis, rare disease patients cost more than double non-rare disease patients and attend 55% more outpatient appointments. Additionally, late diagnosis leads to costly disease progression (e.g. Cystic Fibrosis and Batten disease, where delays ranging from 1.5 years to 20 years result in high healthcare expenses due to disease advancement). Paediatric cases of late rare disease diagnosis are particularly concerning, leading to longer hospital stays and substantially higher costs per discharge compared to common conditions.

Environmental impact

During the long “diagnostic odyssey” patients attend many avoidable appointments, undergo unnecessary diagnostic investigations and in many cases are prescribed unnecessary medications. This activity and consumption entails a significant - and avoidable - environmental burden.

Why does the problem arise?

Mendelian explains late diagnosis in rare diseases as being a two part problem:

  1. Information problem

Matching varied phenotypic presentation to diagnostic or suspicion criteria is a hard matching problem for the human brain. With 7,000+ rare diseases and ever-evolving science patients are missed due to (1) their rarity –  clinicians don’t consider rare diseases in their differential list and are unlikely to have encountered the specific rare disease; (2) diverse and varied symptoms; (3) the gradually progressive nature of many rare diseases, with key red flags slow to appear; (4) mismatch of time and tools: lack of capacity to proactively address needs of frequent flyer patients with complex, multisystemic presentation and to solve the complex “matching / information” problem of linking phenotype to disease criteria.

  1. Systems problem

Rare diseases are missed due to how patients flow through healthcare systems. To diagnose multisystemic, progressive and varied rare diseases, input from a multidisciplinary team is often necessary to piece together biomarkers and phenotypic hallmarks. Healthcare systems typically manage how patients get access to specialist input – either by requiring referral from a general practitioner, or by patients accurately self-selecting into the right specialism. In all healthcare systems this limits the accurate matching of right patient to right specialism(s), resulting in extended diagnostic delays.

What is your solution?

MendelScan is an AI case-finding tool that scans through electronic health records (EHRs) using algorithms to identify patients who lack a diagnosis but have a high probability of a rare disease. 

Mendelian partners with healthcare providers and payers to run MendelScan proactively (i.e. in the background, not as a desk-side clinical tool) on large population datasets to pick out these patients (for 40 rare diseases). Once a shortlist of patients is identified by the tool, Mendelian provides a summary to each patient’s treating physician on the possible diagnosis and the next steps in the diagnosis journey.

It solves the “information problem” by using intelligent machine algorithms at scale to piece together missed phenotypic patterns (in a way the human brain couldn’t), and solves the “systems problem” by working proactively with healthcare providers to seek-out and identify patients through targeted operational teams.

MendelScan has been run on 1.5 million patients in the NHS, growing to 10 million this year, and is the winner of the prestigious NHS AI Award. 

In retrospective studies it has shown it can find missed patients an average of 4 years earlier than they receive a diagnosis, and its findings have been published in the leading rare diseases journal OrphaNet. To date, in deployment it has picked out over 450 NHS patients with rare and ultra-rare conditions and is progressing them towards diagnosis.

MendelScan is a regulated medical device in the UK, and builds its algorithms using machine learning on an 18 million anonymised dataset, publishing its results to the regulator before deploying in active patient settings.

Who does your solution serve, and in what ways will the solution impact their lives?

MendelScan serves (1) the undiagnosed rare disease patient population for our 40 rare diseases; (2) healthcare systems, who benefit from earlier diagnosis.


The collective prevalence of the 40 rare diseases in MendelScan’s portfolio is 1 in 500 (e.g. 660,000 in the US). By proactively seeking out the patients with those diseases who are not yet diagnosed, and accelerating them to diagnostic steps, MendelScan seeks to shorten their diagnosis journey. This provides opportunities for these patients to: access to treatment and trials; appropriate care and support (e.g. access to patient support groups); and improved quality of life.

This cohort of undiagnosed patients exhibit several core characteristics: they are typically high healthcare system utilisers (attending many medical appointments); they have high associated healthcare costs (either paid for directly, by insurance, or by public payers); and often have multi-systemic, progressive disease presentations. Additionally, there are correlations between time to diagnosis are various socioeconomic characteristics: poorer individuals, of minority ethnic groups tend to have longer diagnosis journeys.

Mendelian does a lot to engage and involve these patient communities. The company has strong links with many patient advocacy organisations (e.g. Behcets UK); has a nominated patient representative (who advises the company on relevant issues); is setting up a patient and public representative group; and presents at many disease conferences. 

Inputs from all of these sources influence how Mendelian builds its technology, and how it notifies physicians relating to particular disease-specific issues.

Healthcare systems 

Since MendelScan is a clinician-facing technology solution, Mendelian does a lot to work with doctors who are users of the tool. Mendelian partners with NHS organisations covering 1.5m patients, and regularly conducts product feedback and improvement via this channel. The company also has a clinical advisory team representing an international group of doctors ranging from primary care practitioners to a host of top international specialists. 

How are you and your team well-positioned to deliver this solution?

Mendelian is a team of doctors, data scientists and healthcare executives with extensive personal experience and connection to the rare disease community. The company exists to develop technology to directly benefit rare disease patients. 

CEO is Dr Peter Fish, a South African medical doctor with a specialism in human genetics. Peter practised medicine across South Africa, gaining experience of the diversity of medical conditions affecting a wide variety of ethnic groups, and developed a particular interest in genetics. He is closely linked into the rare disease community – having worked at the world-leading Sanger Institute in Cambridge, UK, as well as leading several healthcare technology companies specialising in precision medicine. He is a regular speaker at rare disease conferences, and a contributor to many pieces of high impact research.

Alongside Peter, Mendelian’s leadership is deeply rooted in the rare disease community. Mendelian has a broad array of medical advisers with much patient experience. Our data scientists have PhDs from Oxford, University College London and King’s College London and have done extensive work with patients and under-represented groups. 

Mendelian’s patient representative group meets once a quarter to advise the company on product development, disease selection, patient sensitivities and communication approaches. The group comprises a number of rare disease patients and general public patients who are outside the rare disease community.

MendelScan benefits from all this advisory input, and in developing and improving the product patients are at the core. The community we work with directly impacts:

  • Which diseases we focus on – relevant disease communities are essential to deeply understanding the diagnosis journey in each case, and opportunities to accelerate it.

  • How information is communicated to patients – many RDs significantly affect patient life outcomes, quality of life, reproductive health, and we take much input on how best to communicate sensitive information to patients.

Which dimension of the Challenge does your solution most closely address?

  • Improve the rare disease patient diagnostic journey – reducing the time, cost, resources, and duplicative travel and testing for patients and caregivers.

In what city, town, or region is your solution team headquartered?

London, UK

In what country is your solution team headquartered?

  • United Kingdom

What is your solution’s stage of development?

Growth: An organization with an established product, service, or business model that is rolled out in one or more communities

How many people does your solution currently serve?


Why are you applying to the Prize?

Mendelian has a strong foundation in the UK: it is working with NHS organisations covering 1.5 million people, it will expand to 10 million in the next year, and is recipient of support by the UK government (including the NHS AI Award). It has proven a pioneering population health approach to improving diagnosis of rare diseases. 

Now Mendelian seeks to expand its product to the USA. This will require funds to support: (1) product adaptation: localising the product to healthcare systems (data, regulation, technical infrastructure); (2) sales and business development to secure new partnerships with healthcare providers and payers and biopharma companies. 

The MIT Solve prize offers Mendelian a prestigious avenue to secure vital funds to support these avenues, and opportunities to connect with the rare disease community in the US.

Who is the Team Lead for your solution?

Peter Fish

How is your Team Lead connected to the community or communities in which your project is based?

Dr Peter Fish is a South African medical doctor with a specialism in human genetics. Peter practised medicine across South Africa, gaining experience of the diversity of medical conditions affecting a wide variety of ethnic groups, and developed a particular interest in genetics. He is closely linked into the rare disease community – having worked at the world-leading Sanger Institute in Cambridge, UK, as well as leading several healthcare technology companies specialising in precision medicine. He is a regular speaker at rare disease conferences, and a contributor to many pieces of high impact research.

More About Your Solution

What makes your solution innovative and sustainable?


MendelScan’s innovative approach is twofold: (1) the core logic underpinning the technology; (2) Mendelian’s operational approach.


MendelScan uses algorithms to detect overlooked rare disease patients from large electronic health record databases. While many common diseases have clear diagnostic “red flags” and hallmarks, rare diseases present distinct challenges: rarity - picking out cases from large population datasets carries many challenges for statistical reliability (minimising ‘noise’); lack of criteria - diseases lack well-defined diagnosis or suspicion criteria, and where these do exist they depend on idealised information about a patient; complex conditions many of the diseases are multi-systemic, progressive and heterogenous.

Due to these challenges, Mendelian developed MendelScan using a combination of three important sources: (1) existing literature on diagnostic or suspicion criteria – encoding into algorithms the bedrock criteria that are used for diagnosis; (2) specialist input – top clinicians in each rare disease who provide much insight as to the typical presentation and progression of disease; (3) very large real world research datasets (e.g. an 18 million NHS patient dataset, used to train machine learning models).

Each disease algorithm that Mendelian develops combines these three inputs, that is trained and tested on research data. Mendelian evaluates statistical measures for each (e.g. its sensitivity, specificity and positive and negative predictive values) and includes in its tool only those that meet satisfactory criteria.

MendelScan is adaptable, and is constantly incorporating new disease information from other advancements in the market (e.g. from Human Phenotype Ontology, OrphaNet).

Typically, other algorithmic approaches in the market rely on simpler rule-based approaches that are too noisy to be meaningful or reliable to clinicians. By combining three different approaches to build effective algorithms MendelScan is: interpretable (all the clinical factors that contribute to a patient being identified and transparent and clinically useful); based on real world data (by training on real world EHR data, Mendelian picks out biomarkers that are relevant to how real world patients present to healthcare systems, not based on idealised disease information); sophisticated (the machine learning algorithms trained on 18 million people are able to pick out patients with high discrimination.

Operational approach 

Mendelian took the active decision to develop MendelScan as an ‘in-the background’ proactive population tool, rather than a deskside clinical decision support system. The latter have poor evidence of efficacy in healthcare systems, and are not set up for the realities of rare diseases (which crop up very infrequently). 

Instead, where MendelScan is deployed, a dedicated, proactive operational clinical team is set up to handle the patients that are flagged by the system, and drive them to diagnosis. 


Mendelians combination of innovative technology and operational model sets it up to drive a paradigm shift in rare disease diagnosis. This is already happening in the UK, where Mendelian’s approach has swung diagnosis to being a proactive, population health mission. By eliminating unnecessary “diagnosis odysseys” for patients Mendelian will massively decrease the environmental impact of rare disease.

What are your impact goals for the next year and the next five years, and how will you achieve them?

Increase Rare Disease Diagnosis - Many Rare Diseases are significantly underdiagnosed.  People who never receive a proper diagnosis face significant challenges in accessing appropriate treatments, participating in clinical trials, and connecting with support communities tailored to their specific condition. 

Shorten Time to Diagnosis - Many Rare Diseases are degenerative, and treatments work best when the disease is caught early - before the damage has been done. Research conducted by the National Organization for Rare Disorders (NORD) indicates that it takes an average of 5-7 years to receive an accurate diagnosis for a rare disease. With newer gene therapies, patients with advanced disease will not be eligible. By focusing on shortening the time to diagnosis, we can improve patient outcomes and increase the likelihood of successful treatment interventions.

Improve Equity of Care - Studies have shown that diagnoses of Rare Diseases are most likely to happen if a person lives near a treatment center or Academic Hospital, which is attributed to an attuned medical community and lower access barriers to specialists. By improving equity of care, we aim to ensure that individuals with rare diseases, regardless of their geographical location, have equal access to timely and accurate diagnoses. 

MendelScan will reach these goals through deployment at scale. By leveraging artificial intelligence and machine learning algorithms, MendelScan can analyze large datasets of clinical data to aid in rare disease diagnosis. This technology will act as a backstop to the gaps in care and capacity constraints that are failing people affected by Rare Diseases and ensure those people are getting the proactive care they deserve.

How are you measuring your progress toward your impact goals?

New Diagnoses - In prospective deployments we are tracking the number of patients impacted, the clinical value of the tool, and ultimately the number of new diagnoses enabled.  

Reduced Time to Diagnosi- In large data sets, we are tracking how, if implemented, the net reduction of months from first recorded disease feature to recording of diagnosis.  Following prospective deployment of the tool, we will examine the real world time to diagnose across a portfolio of diseases in the area with MendelScan vs. an area without.  

We are analysing both of these metrics by UK post code which can be linked to measures the social determinants of health and access to health care.

Describe in simple terms how and why you expect your solution to have an impact on the problem.

As described above, Mendelian explains late diagnosis in rare diseases as having two key causes: (1) “information problem” (piecing together rare patterns of symptoms is hard for any human brain); and (2) “systems problem” (patients are overlooked in different part of the healthcare systems as non-experts guide them to referrals). 

By combining the MendelScan technology with dedicated operational rare disease teams, Mendelian solves both of these problems to have impact.

Mendelian has strong evidence that its approach effectively solves (1) the information problem by using the power of machine computation, encoded into specific algorithms, to match phenotypic patterns and find the “right” overlooked patients; (2) the systems problem by ensuring those patients are actually progressed onward to diagnosis:

  • Retrospective studies: Mendelian runs validation of its algorithms on a large retrospective dataset of 18 million patients, to show their power in picking up patients who have subsequently received a diagnosis. The company has done many studies and found e.g. that for some diseases patients are picked up 4 years earlier than diagnosis.  

  • Real world impact: in the NHS, Mendelian has run its tool across 1.5 million patients.  From the latest pilot of 213,000 EHR searching for 32 Rare Diseases, there are 82 patients on an investigation pathway.  

  • NHS AI Award: Mendelian is halfway through a major NHS award, which involves evaluating its tool at scale across 10 million patients to show the system-level impact. This will bolster existing research and lead to major publications.

If your solution is tech-based, describe the core technology that powers your solution.

Our technology platform allows for: 

  • The cleaning and ingestion of di-identified, structured electronic health record (EHR) data. 

  • The encoding of criteria to identify patients who present signs and symptoms of a particular disease (features).

    • The features are identified by SNOMED codes that might have been coded as events in the patient’s primary care EHR.

    • The criteria can also evaluate numeric values that might accompany a SNOMED code in case of laboratory results or observations entries.

    • The age of the patient at the onset of the event can be taken into account.

    • The automatic scanning process of EHRs based on the criteria can focus on patients of a certain age range and/or sex and it can consider events that have been recorded within a certain timeframe.

    • The criteria rules can incorporate the automatic exclusion of a case if certain features are identified.

    • The highlighting secondary features that are not necessarily useful to identify a patient as a possible case, but that can be helpful in the evaluation of the clinical condition.

    • It is possible to assign a weight to each feature (points), so that the list of highlighted patients is sorted based on a scoring system.

  • Running of a core matching logic that interprets multiple disease level criteria against ingested health records 

  • Visualisation of flagged records for quality control 

  • Automatic generation and delivery of a patient case report to a specified clinician securing within the NHS secure environment. 

Which of the following categories best describes your solution?

A new business model or process that relies on technology to be successful

Please select the technologies currently used in your solution:

  • Artificial Intelligence / Machine Learning
  • Big Data
  • Software and Mobile Applications

If your solution has a website, app, or social media handle, provide the link(s) here:

In which countries do you currently operate?


In which countries will you be operating within the next year?

UK, USA, Canada

Your Team

What type of organization is your solution team?

For-profit, including B-Corp or similar models

How many people work on your solution team?

- 9 full time employees

- 7 staff on contract, supporting the company on clinical matters

How long have you been working on your solution?

MendelScan has been under development for 5 years. 

What is your approach to incorporating diversity, equity, and inclusivity into your work?

Three core channels:

  • Mendelian has a patient and public representative group who meet quarterly to scrutinise Mendelian's approach to diversity and inclusion, both in the company's own practices, but also in the application of its technology.
  • The company ensures to hire from a diverse array of nationalities, professional backgrounds, healthcare experiences, and its staff are diverse in ethnicities and socioeconomics.
  • In building our technology, we have a dedicated diversity and inclusion protocol that ensures the data is appropriately representative, and we do our utmost to reduce any biases in the algorithms.
Your Business Model & Funding

What is your business model?

Mendelian has a two-sided business model:

  1. Healthcare payers and providers

  2. Biopharma companies

Healthcare payers and providers

Payers and providers pay Mendelian to deploy MendelScan to achieve two things: reducing unnecessary/excess costs associated with avoidable parts of the diagnostic journey; to diagnose patients earlier, and improve prognosis and outcomes. 

To date, Mendelian has had much success with its value proposition in the NHS: partnerships with NHS primary care providers covering 1.5 million patients; awarded the NHS AI Award, which will facilitate growth to 10 million patients over the next 12 months; NHS Genomic Medicines Service – two major transformation programmes covering 200,000 patients.

Internationally, Mendelian has nascent partnerships with healthcare systems in Canada and the US, but will seek to grow those over the next 12 months.

Biopharma companies 

Mendelian partners with pharma companies in order to help them grow their markets of accurately diagnosed patients, at an earlier stage. This increases their market for therapeutics and clinical trials enrollment. Mendelian has partnered with 8 of the top global biopharma companies including Novartis, Sanofi, Alexion/AstraZeneca and Ipsen. 

Do you primarily provide products or services directly to individuals, to other organizations, or to the government?

Organizations (B2B)

What is your plan for becoming financially sustainable?


Mendelian is already generating some revenue from healthcare payers and providers and biopharma partners. We will continue to grow revenue from these two channels:

  1. Healthcare payers and providers: Further NHS growth, firstly at a regional and then national levels with associated revenues. We are planning a number of international pilots that will pave our way to commercial expansion, capitalising on nascent partnerships in the US, Canada and Europe.

  2. Biopharma partners: We currently have existing partnerships with 8 biopharma companies, with whom we are in the process of negotiating expansion agreements for the diseases we have focused on to date and into new disease areas. Beyond this, we are in discussions with a number of additional companies and are in the process of securing new R&D deals.

R&D grants 

  1. The recently announced NHS AI Award funding will be coming in over the next 12 months and will support Mendelian to research its impact in the NHS.

  2. Government R&D: We have secured a number of innovation grants to date and we will continue to apply (and hopefully secure) grants from the UK’s innovation grant funding agency, Innovate UK, to expand MendelScan into new rare diseases areas.


Mendelian is exploring private investment from several avenues, which could fund its expansion in new markets and territories allowing access to additional revenue sources.

Share some examples of how your plan to achieve financial sustainability has been successful so far.

  1. Government R&D grants. Mendelian has been the beneficiary of multiple R&D grants from Innovate UK, totalling in excess of £1m. These have funded the development of specific disease modules on MendelScan which the company has subsequently monetised through commercial partnerships.

  2. NHS Awards. Mendelian was awarded £1.4m for the NHS AI Award, in recognition of its impactful innovation.

  3. Industry R&D support. Mendelian has received funding from several healthcare funders, in the form of grants and loans – including e.g. from LifeArc.

  4. Repeatable revenue from healthcare partners. Previously we were contracted by two of the NHS’s Genomic Medicine Service Alliances (GMSAs) to facilitate genomic medicine transformation in frontline clinical practice at a primary care level. These projects were funded by the GMSAs. The pilots have been well received and are leading to multi-year deals with organisations covering 1.5 million patients in the NHS, growing to 10 million over the next 24 to 36 months.

  5. Biopharma revenue. Over £1m of revenue has been generated to date from our partnerships with 8 global biopharma companies (Novartis, Alexion/AstraZeneca, Kyowa Kirin, Orchard Therapeutics, IPSEN, PTC Therapeutics, HRA Pharma and another that has not yet been publicly announced).

Solution Team

to Top