Which questions can we work on together?
Examples of studies to design together. The professor sets the question and the comparison basis; AetherHeal builds the tool that tests it under the same conditions.
- 01
Where do two reviewers read the same image differently?
A research screen that places approved images and annotations side by side and records review position, rationale and reasons for withholding judgment.
- Evaluated together:
- inter-reviewer agreement, missing annotations, reasons for deferral.
- 02
At which time points and fields are follow-up data missing?
A data-organization tool that collects protocol metrics from an approved cohort and reviews gaps by observation time point.
- Evaluated together:
- data completeness, gaps by time point, consistency of repeated extraction.
- 03
Can the person receiving a handoff understand what to do next?
A research form that structures problems, evidence, open tasks and owners in the department's handoff order for comparison with the current method.
- Evaluated together:
- preservation of required fields, reviewer understanding, identification of open tasks.
Depending on the question and the data, we choose from retrieval, annotation support, a workflow that combines rules and models, or separate training. Versions and conditions are fixed so results can be checked again.
The research team sets the standard; the engineering team implements it.
The professor and the research team lead the question, the evaluation criteria and the interpretation. AetherHeal provides research screens, data processing and a reproducible execution environment. Specific roles are set in the protocol and the agreement. We do not grant data access on anyone's behalf.
The role table by item (question and design · data · annotation and evaluation · engineering · reproducibility and change) and the scope of custom development: see the proposal PDF.
Settled in writing before the work starts.
Our default proposal keeps raw data within the research institution's control. Who has access, who is responsible for de-identification, which institution's IRB reviews, the ban on reuse for other purposes, and what happens after the study ends are all set before the start.
Academic leadership and the principle of publishing results do not change in exchange for technical support. We do not block publication of negative results. Authorship follows real academic contribution; providing a tool or technical support does not by itself create a claim to authorship.
The six data-management questions (default proposal and pre-start agreements) and the detailed terms on ownership of results and co-authorship: see the proposal PDF.
- 01
A one-page research question
The clinical question, the workflow currently done by hand, the type and access status of available data, and the output the tool should produce.
- 02
Scope fit review
Whether the question can be built as a research tool, whether the required data and evaluation conditions can be assembled, and whether the scope is separate from hospital-wide operational adoption.
- 03
Protocol and agreement
Roles, metrics, data responsibility, custom development scope, cost, rights to results, authorship criteria and end conditions, fixed in writing.
- 04
Institutional review and environment
Data access and environment setup begin only after the required IRB and data reviews, contracts and security review are complete.
We are currently recruiting research partners. Institutional collaborations are introduced only within signed agreements and agreed public disclosure.
Send a one-page research question and we will review together whether it can become an approvable research scope and a buildable tool. drjeeju@aetherheal.com
- Author
- AetherHeal
- Document status
- Research collaboration proposal