Health Canada seeks technological approaches in Deep Learning and Artificial Intelligence to predict the success of possible donor-recipient matches and transplant outcomes to support evidence-based decision-making about organ donation and transplantation.
Sponsoring department: Health Canada
Funding mechanism: Grant
Opening date: August 23, 2019
Closing date: October 18, 2019 14:00 Eastern Daylight Time
Artificial Intelligence (AI) is poised to transform the field of organ donation and transplantation in Canada. Existing clinical decision support tools include: a tool that provides a standardized and personalized assessment of donor and organ suitability to make decisions about organ retrieval; and, a similar tool that helps kidney failure patients determine if they should accept the donor kidney or wait for a better match. While these tools can be used to help improve clinical decisions in donation and transplantation, AI experts are needed to use Deep Learning to continually enhance the predictive ability of these tools towards improved circulatory determined death (DCD) donation rates and to identify the best potential donor-transplant recipient matches.
The solution will focus on both the organ donor (for all types of organs), and the transplant recipient. For the organ donor, the solution would consider the suitability of the donor and the quality and suitability of the organ by linking it to the recipient's outcome data. For transplant recipients, the tool would help recipients make a better decision about whether to accept the organ that is offered. Currently, as there is a shortage of organs, the solution would focus on kidney transplants, which make up over 60% of all transplants done in Canada. Once more hearts, lungs, and livers are available, the tool could be expanded to other organs.
Desired outcomes and considerations
The solution must:
- develop a tool that leverages technological approaches in Deep Learning and Artificial Intelligence to:
- Continually improve the predictive ability of real-time, clinical decision support tools so that it may be used by donation and transplant physicians across Canada to accurately identify candidates that are mostly likely to be successful donors;
- Personalize kidney transplantation by predicting the success of potential donor-recipient matches and choosing those with the best chances for excellent long-term transplant outcomes;
- Reduce unsuccessful DCD attempts and improve transplant outcomes, thereby enhancing family experience of donation and optimizing system costs.
- develop a tool that is able to match kidneys to transplantation recipients, but is also expandable to other organs, as needed.
Background and Context
Every year, thousands of Canadians receive life-saving, cost-effective organ transplants, while thousands still wait, and hundreds die because not enough organs are available. The practice of donation after circulatory determined death (DCD) has increased the rates of deceased donation in Canada by 40% in the last 10 years but despite this success, many potential DCD donors fail to donate their organs. DCD occurs after withdrawal of life-sustaining measures (WLSM) in consented patients with non-recoverable illness in the intensive care unit. However, prolonged exposure of organs to low oxygen during the dying process after WLSM renders them increasingly unsuitable for transplantation. It is currently very difficult to predict which potential DCD donors will die within a timeframe and a manner that permits successful donation. As a result, over 30% of DCD attempts are unsuccessful, which adds to the anguish of grieving families and unnecessarily consumes vast resources from both the donation and transplant teams. Some centres in Canada refuse to implement DCD programs because of the uncertainty of success.
On the transplant side, patients waiting for a new kidney struggle trying to decide if they should accept the donor kidney or wait on dialysis for a kidney that might provide better short- or long-term outcomes. This decision involves a complex evaluation process that is currently supported by manual calculations and algorithms, which are not informed by current and emerging evidence.
Clinical decision support tools have been developed by the Children's Hospital of Eastern Ontario Research Institute Inc. and the Centre de recherche du Centre hospitalier de l'Université de Montréal to provide real-time, standardized, personalized assessment of donor and organ suitability to make better decisions about organ retrieval and kidney transplant matches. Although a novel waveform-based variability-derived predictive model has been developed into a clinical decision support tool, healthcare practitioners need to work with AI experts and industry to build the AI interface that will take the prediction tool recently designed and turn it into a living clinical tool that will continue to make better predictions the more data that is collected. Deep Learning to help improve organ donation (prediction of who will be a better donor) has not been funded or developed and is still at the conceptual stage. Deep learning to help make better decision for recipients (should I accept this kidney or wait for the next one) is in early research phases in Montréal and a partnership with AI researchers at the Institute for Data Valorization (IVADO) has begun some work to explore how they would set this up. Innovators would build on early phase research that has been initiated in Montréal and would be expected to work with both the Ottawa and Montréal teams. The innovator would have access to existing data through multi-institution agreements for clinical testing.
Maximum grant value:
Multiple grants could result from this Challenge.
Funding of up to CAD150,000 for up to 6 months could be available for any Phase 1 grant resulting from this Challenge.
Funding of up to CAD1,000,000 for up to 2 years could be available for any Phase 2 grant resulting from this Challenge. Only eligible businesses that received Phase 1 funding could be considered for Phase 2.
This disclosure is made in good faith and does not commit Canada to award any grant for the total maximum funding value.
For Phase 1 it is anticipated that up to five meetings will require the successful bidder(s) to travel to the locations identified below:
- Kick-off meeting in Ottawa, Ontario
- Progress Review Meetings: one to three design/progress review meetings by teleconference/videoconference
- Final Review Meeting in Ottawa, Ontario.
- Innovators may wish to travel to Montréal or Ottawa to work with research teams. Meetings can be held via teleconference/video conference.
Solution proposals can only be submitted by a small business that meets all of the following criteria:
- for profit
- incorporated in Canada (federally or provincially)
- 499 or fewer full-time equivalent (FTE) employeesFootnote *
- research and development activities that take place in Canada
- 50% or more of its annual wages, salaries and fees are currently paid to employees and contractors who spend the majority of their time working in CanadaFootnote *
- 50% or more of its FTE employees have Canada as their ordinary place of workFootnote *
- 50% or more of its senior executives (Vice President and above) have Canada as their principal residenceFootnote *
Part 1: Mandatory and Minimum Pass Mark Criteria
Proposals must meet all mandatory criteria (Questions 1a and 2) and achieve the minimum pass mark for Question 3 in order to be deemed responsive and proceed to Part 2.
1 a. Scope
Describe your proposed solution and how it responds to the challenge. Include in your description the scientific and technological basis upon which your solution is proposed and clearly identify how your solution meets all of the Essential Outcomes (if identified) in the Desired Outcomes and Considerations section in the Challenge Notice.
Mandatory — Pass/Fail
2. Current Technology Readiness Level (TRL)
Mandatory — Pass/Fail
Pass: The Applicant/Bidder has demonstrated that the proposed solution is currently between TRLs 1 and 4 (inclusive), and provided justification by explaining the research and development (R&D) that has taken place to bring the solution to the stated TRL.
Fail: The Applicant/Bidder has not provided sufficient evidence to demonstrate that the current TRL is between 1 to 4 (inclusive) including:
Describe the novelty of your solution and how it advances the state-of-the-art over existing technologies, including competing solutions.
Point Rated with Minimum Pass Mark
The minimum pass mark for this criteria is 4 points.
0 points/Fail: The Applicant/Bidder has not demonstrated that the proposed solution advances the state-of-the-art over existing technologies, including available competing solutions; OR
The stated advancements are described in general terms but are not substantiated with specific, measurable evidence.
Part 2: Point-Rated Criteria
Proposals that do not achieve the overall minimum score of at least 55 points out of a possible 110 points (50%) will be declared non-responsive and given no further consideration.
The overall minimum score is determined by adding the Applicant/Bidder's scores from the following questions together (1b, 3, 4-13).
Describe how your proposed solution addresses the Additional Outcomes (if identified) in the Desired Outcomes and Considerations section in the Challenge Notice. If no Additional Outcomes are identified in the Challenge Notice, text entered in this section will not be considered.
If no Additional Outcomes are identified in the Challenge Notice, Bidders/Applicants will receive 10 points
4. Phase 1 Science and Technology Risks
Identify potential scientific and/or technological risks to the successful development of the proof of concept and how they will be mitigated in Phase 1?
5. Benefits to Canada
Describe the benefits that could result from the successful development of your solution. Applicants/Bidders should consider the potential benefits using the following three categories:
6. Phase 1 Project Plan
Demonstrate a feasible Phase 1 project plan by completing the table.
Note: Phase 1 cannot exceed 6 months and TRL 4.
7. Phase 1 Project Risks
Identify potential project risks (eg. Human resources, financial, project management, etc) to the successful development of the proof of concept and how they will be mitigated?
8. Phase 1 Implementation Team
Demonstrate how your project implementation team has the required management and technological skill sets and experience to deliver the project plan for Phase 1 by completing the table. A member of the implementation team can have more than one role.
Include the labour rates and level of effort for each member. A day is defined as 7.5 hours of work, exclusive of meal breaks. The labour rates and level of effort will be reviewed as part of the evaluation for Question 10.
If your business were to receive funding from Innovative Solutions Canada, describe what actions (e.g., recruitment strategy, internships, co-op placements, etc.) might be taken in Phase 1 to support the participation of under-represented groups (e.g., women, youth, persons with disabilities, Indigenous people, visible minorities) in the research and development of the proposed solution.
Each bidder/applicant in their response to this question must focus only on describing relevant programs, policies, or initiatives that it currently has in place or would put in place to support the R&D effort in Phase 1. Do not provide any personal information of individuals employed by your company or that of your subcontractors in the response below.
10. Phase 1 Financial Proposal
Demonstrate a realistic financial proposal for the Phase 1 project plan by completing the table.
11. Phase 1 Financial Controls, Tracking and Oversight
Describe the financial controls, tracking and oversight that will be used to manage the public funds throughout Phase 1.
12. Phase 2 Strategy
Describe a realistic strategy for the prototype development if selected to participate in Phase 2.
Responses should include:
13. Commercialization Approach
Describe your overall commercialization approach for the proposed solution.
Responses should include: