Frequently Asked Questions
This page collects answers to the most common student questions about CACoM — ranging from project proposals and grading to collaborations, AI tools, and deliverables.
If your question isn't listed here, ask your instructors — and if it turns out to be common, we'll add it!
About the Course
What exactly is CACoM about?
CACoM (Clinical Applications of Computational Medicine) is a project-based course where students design and execute small research projects at the intersection of data science and medicine. It's not a software engineering course or a design competition — it's about understanding a clinical or biomedical question deeply, and using computational methods to answer it. You may use any tools or programming languages you like, as long as they serve a meaningful scientific goal.
Can we focus on building an app, tool, or game?
Not as the end goal.
You are welcome to build tools, interfaces, or simulations as a means to address a research question — but if your project's primary outcome is “we built a cool app,” you're in the wrong course.
CACoM is about insight, not interfaces.
Do we need medical data?
Not necessarily.
You can work with simulated, public, or synthetic data, or even design systematic reviews or surveys.
What matters is that your question is clinically relevant and your methodology is rigorous.
Can we change our topic later?
Yes — up to the topic approval deadline (see Schedule). After approval, your core topic and direction are set, but methods, metrics, and analysis can evolve as you learn more. Any major change of scope must be discussed with instructors.
Is attendance required?
In-person attendance for scheduled sessions is expected unless you have a justified reason. Some sessions may be hybrid (in-person + remote). Check the current time/room and notes on the Schedule.
Project Development
What's the difference between a “vague idea” and a “refined question”?
A vague idea is a broad area of curiosity (e.g., “Analyze gait using IMUs”).
A refined question turns that into something testable (e.g., “Can ZUPT-based odometry accurately estimate step length in daily-life IMU recordings?”).
You'll find examples in the Refinement Guide.
Can we collect our own data (e.g., IMUs, stethoscope, simulator)?
Yes. Clearly describe your setup, safety, data handling, and permissions in the proposal.
For clinical/participant data, follow privacy rules and obtain any needed approvals.
See Data Privacy.
What if our idea is outside the listed Areas of Interest?
That's fine — but you must persuade us: show clinical relevance, feasibility, access to data/devices, and a clear evaluation plan.
Strong initiative and a convincing rationale are required.
(Projects aligned with collaborator-proposed topics are also welcome.)
We missed the topic approval deadline — can we still submit?
There's a short late approval window (see Schedule) with a grade penalty. After that, new topics can't be accepted. Start early and use the in-class mini-pitch sessions to get feedback before the deadline.
What happens if more than one group wants to work on the same topic?
That depends on what you mean by “topic.”
If you mean the Area of Interest (e.g., CTG analysis, IMU-based gait modeling, or wearable sensors), then of course — yes. It's very unlikely that two groups will end up doing exactly the same thing. Topic areas in CACoM are broad and can support many different research questions, perspectives, or data types.
In fact, if two groups can set up their work synergistically — where one group's results or datasets inform or benefit another's — that's excellent and even encouraged.
However, if you mean that two groups want to pursue the same specific research question, effectively working toward the same outcome, then no — that's not allowed. Each project must define its own distinct question, dataset, or approach to ensure originality and independent evaluation.
Our collaborator stopped responding — what do we do?
Notify the CACoM instructors immediately. Do not wait for weeks. We can help you reframe or redirect your project if needed — but only if you communicate early.
What if our collaborator gives us access to proprietary data or devices?
You can absolutely use proprietary data or hardware (e.g., the fetal heartbeat simulator), but only the parts that are visible and reproducible to CACoM instructors can be graded.
If most of your work relies on inaccessible or confidential elements, document what you did — but note that those parts won't be evaluated.
See Collaboration Policy and Data Privacy Policy.
Can we contact clinicians or companies ourselves?
It depends on the situation:
-
Established CACoM collaborators (connected via Martin/teaching team):
First express interest to Martin. He will make the introduction.
After the connection is made, CC CACoM instructors on subsequent emails. -
New external partner — recommended path:
Before you reach out, discuss the idea briefly with the CACoM team. We'll help assess fit, decide who should contact whom first, and agree on whether to CC instructors (usually yes after the intro). -
New external partner — you reach out without prior discussion:
You may do so without CC'ing CACoM initially. If the contact becomes active/relevant, notify the CACoM team promptly and align on next steps (including when to start CC'ing instructors).
This approach avoids conflicting outreach with existing partners, keeps communication professional, and ensures the collaboration fits CACoM's scope and policies.
Deliverables
What exactly do we have to submit?
By the final deadline, your team must deliver:
- 🖼️ A poster summarizing your project
- 🎬 A 1-minute video teaser
- 🗣️ An oral poster presentation
- 💾 A reproducibility package (report/code/data or equivalent)
See the Presentation Package and Reproducibility Package pages for details.
How do we submit the proposal?
Email your proposal (PDF or Markdown using the template) to Prof. Martin Daumer, CC Pooja N. Annaiah, by the topic approval deadline (see Schedule).
There is no submission portal for proposals.
Where do we upload our files?
- Proposals: email Prof. Martin Daumer (CC Pooja N. Annaiah).
- Poster & 1-minute video: you'll receive a shared Google Drive link close to the deadline.
- Reproducibility package: upload to the CACoM Google Drive; code may be on GitHub (no sensitive data).
Do not upload sensitive or proprietary data to GitHub or public clouds. See the Data Privacy Policy.
Can we use AI tools like ChatGPT or Copilot?
Yes — CACoM actively supports responsible AI use. You can use AI for writing help, brainstorming, or code assistance, as long as you understand and verify everything you submit. See the AI Tools & Academic Integrity policy for guidance and examples.
What happens if we use AI to write everything?
It'll show. Fluent but shallow text with no real understanding will be graded as minimal effort. You'll lose points for depth and reflection. We don't forbid AI — but we reward insight, not autopilot.
Grading & Evaluation
What influences our grade the most?
Your depth of understanding, clarity of presentation, and scientific rigor. Slick visuals or flashy tools don't compensate for poor reasoning. See the Grading & Rubric page for the detailed breakdown.
Are mini-pitches graded?
They are formative (not separately graded) but required.
Mini-pitches are crucial for feedback and typically precede proposal approval. Skipping them risks delays or rejection.
How do grading “priorities” work?
CACoM uses priorities (not exact percentages) to signal what matters most. Priority 1 items weigh more heavily than lower priorities, but we don't publish a point formula. See Grading & Rubric for the priority layout and expectations.
How is collaboration performance graded?
Outstanding collaboration can earn a bonus, while unprofessional conduct (e.g., slow/unreliable communication, ignoring partner input, causing partners to disengage) can lead to a penalty.
See the Collaboration Policy.
How do you evaluate effort and independence?
Effort isn't about hours — it's about initiative, consistency, and ownership. Teams that plan proactively, solve problems independently, and communicate early score higher. Needing frequent reminders or relying heavily on instructors lowers your score under Professionalism & Independence.
Can poor presentation affect an otherwise good project?
Yes. Even though Presentation & Performance during discussion is a separate priority, clear communication is non-negotiable.
If you can't explain what you did, why you did it, and what it means, your evaluation will drop substantially.
Will the poster session be evaluative?
Yes. The poster session includes live discussion with instructors and attendees. You should be prepared to defend your choices, methods, and conclusions. This interaction can influence your overall evaluation.
What happens if we accidentally plagiarize?
There's no “accidental plagiarism.” You are responsible for verifying your sources and citations. Plagiarism is an automatic failure (grade = 5.0). See Plagiarism & Citation Policy.
Can we use public data (e.g., from Kaggle)?
Yes, but you must check the dataset's license and cite it properly. If it includes any real patient data, treat it as sensitive — never upload or share it publicly. See Data Privacy Policy.
Miscellaneous
How big should our group be?
Teams typically have 1–7 members. Smaller teams should pick narrower topics; larger ones must demonstrate stronger division of tasks and coordination.
Do I need to wait for the full group to form, or can I already start?
Short answer: You don’t need to wait for a full group — but you must get your topic approved first.
Why: Without approval, you risk investing time in a project that might be rejected or require major changes.
How it works:
- Topic approval comes first. Send your proposal to Prof. Martin Daumer (CC Pooja N. Annaiah) for approval and discuss during the mini-pitch sessions.
- After approval you can start immediately. A group lead is designated (usually the student who pitched the topic) and may begin work right away (reading, data access requests, pilot experiments, etc.).
- Be open to joiners. We encourage teams to welcome additional students who are genuinely interested and add value.
- Form your team autonomously. We trust you to organize roles and resolve minor disagreements; we will not micromanage group formation.
- If there’s a major issue, tell us. For significant conflicts or unclear membership, contact the instructors and we’ll help on a case-by-case basis.
What tools or programming languages can we use?
Anything — Julia, Python, R, MATLAB, TensorFlow, PyTorch, or custom setups — as long as your methods are transparent and reproducible. Tool choice doesn't matter; clarity and validity do.
Can we continue or build upon last year's project?
Yes, as long as you clearly indicate which parts are reused and which are new. See Plagiarism & Citation Policy.
Can we use a project from another course or thesis?
No — projects cannot be double-counted or graded twice. You may continue work that started elsewhere (e.g., a thesis or another module or some external project), but you must clearly specify what part of the work is new and done within CACoM and not subtmitted anywhere else. Only that part will be evaluated and graded. See Project Proposal Guide.
Can we publish our code on GitHub?
Yes — code and documentation are welcome on GitHub. Do not upload real clinical/proprietary data. Provide synthetic examples and put real data in the CACoM Google Drive. See Data Privacy and Reproducibility.
What if something unexpected happens — data lost, group conflict, illness, etc.?
Contact the instructors as soon as possible. We can help you adapt timelines or adjust scope — but only if we know early.
Can we write to collaborators in German?
Yes — while English is the default for course communication, it's fine to use German when corresponding with external partners who prefer it. Just keep CACoM instructors in CC, regardless of language.
What should we put in email subjects?
Use:
[CACoM Winter25/26][Group #][Short Project Title] — Topic/Question
(or replace “Winter25/26” with “Summer26” for the next run).
Example:
[CACoM Winter25/26][Group 4][CTG Beat Detection] — Proposal Submission
Including the semester, group number, and short title helps instructors track messages quickly — especially since multiple CACoM cohorts run each year.
Still Have Questions?
If your issue isn't covered here, send an email to the CACoM instructors — and include your team name, project title, and a short, specific question.