Al-Assisted Early Detection System for Osteoarthritis (OA) Risk Markers in North Eastern Region (NER)

The receipts 🧾

Every red flag on this statement, in full. These are the four places it bites.

1. Exhibit A

   It is categorised Hardware — a pure app submission risks being judged as not answering the category, so you need a physical assessment rig even if the intelligence is all software

2. It gets worse

   No dataset exists that maps gait features to confirmed OA severity, so any accuracy figure you quote is against labels you generated, and an orthopaedic judge will ask exactly that

3. Still reading?

   If you fall back to X-ray classification you have contradicted the premise — the description exists precisely because these centres have no imaging

4. And the finisher

   Screening in health camps means uncontrolled lighting, uneven ground and patients in saris and lungis, all of which break pose estimation in ways your lab video will not reveal

The damage report 📉

Every score this statement earned, and what each one actually costs you.

• Feasibility

  3/5

  Buildable. Not comfortably. There is a week in here you have not planned for yet.

  Pose estimation on a phone is free and works, and public knee OA imaging sets with Kellgren-Lawrence grades exist for the imaging route, but there is no dataset anywhere linking gait video to confirmed OA grade, so the model that the whole product hinges on has no ground truth you can actually obtain.

• Innovation scope

  3/5

  Mildly interesting. The novelty will not carry the room; the build has to.

  The description names four assessment channels but hedges the imaging one with 'if applicable' and never fixes the sensing modality, so you genuinely choose the mechanism even though the workflow around it is dictated.

• Clarity

  3/5

  Clear enough to start, vague enough to drift. Write the scope down and stop reinterpreting it weekly.

  The workflow, interface and deployment context are spelled out clearly, but the clinical target is not — 'OA risk markers' is never defined against any grading scale, so you have to decide yourself what the system is claiming and that ambiguity sits at the centre of the build.

• Acceptance potential

  3/5

  Middle of the pack. This statement will not win the room for you — you will have to.

  Genuinely underserved clinical need and a hands-on demo, but it is filed as Hardware while the description reads mostly as software, and the absence of any obtainable gait-to-diagnosis ground truth means your central claim rests on a proxy you invented.

• Effort

  Heavy

  Heavy. Somebody on this team is not sleeping in week three. Pick who, on purpose.

  A sensing rig, a pose or imaging pipeline, a risk model, a health-worker record system with offline sync, multilingual UI and a patient education module is six pieces, and the portability requirement forces hardware packaging work on top.

• Demo-ability

  Medium

  Demoable, if you rehearse it. Nobody rehearses it.

  The walk-and-measure interaction is immediate and hands-on, but the number it produces cannot be verified in the room, so you are demonstrating a measurement rather than demonstrating that the measurement is right.

• Data

  None supplied

  No dataset comes with this one, so every accuracy figure you quote is a number about labels you invented.

  Nothing is provided with the statement. You are sourcing, cleaning and labelling it yourself, and that work is invisible in the demo but very visible in the questions.

The demo they will have already seen 🪞

Somewhere around 40–100 teams are heading here, and the description is doing the choosing for most of them. They will read the same brief, reach the same architecture, and build a version of the same demo you are planning. Being correct is the floor. If your five minutes could be swapped with the team before you and nobody in the room would notice, you have not picked badly — you have built predictably, which costs exactly the same and hurts more.

What survives 🛡️

The ground worth standing on when the questions start.

• The theme is filed under Space Technology, so teams filtering for MedTech will never see this and the field is thinner than the clinical need would suggest
• Gait analysis with a phone camera needs no purchased sensor, so you can build a credible Hardware-category prototype for the cost of a Raspberry Pi and a tripod
• The description explicitly frames the output as risk assessment and referral rather than diagnosis, which pre-empts the regulatory objection a medical PS usually attracts