01 / Applied AI / 2025–2026
Cough-based TB screening.
TB-Turing / HeAR
01 / SCOPE
Problem and
scope.
This project evaluates whether cough audio contains useful screening signal and how symptom metadata changes the result.
My work focused on the HeAR pipeline within a wider team research project. The Wav2Vec2 programme was shared team work.
02 / IMPLEMENTATION
Implementation
details.
- 01Cough windows
- 02HeAR embeddings
- 03Patient bags
- 04Attention MIL
- 05Patient-level evaluation
Use patient-level evaluation.
Group multiple cough clips into patient bags and split at patient level. This keeps clips from the same person from appearing in both training and validation folds.
Aggregate variable-length patient bags.
Use gated-attention multiple-instance learning over frozen HeAR embeddings to aggregate variable-length patient bags.
Separate audio from context.
Compare audio-only, basic-metadata and symptom-enhanced configurations. Analyse sensitivity-targeted operating points and bootstrap uncertainty alongside AUROC.
03 / RESULTS
Results.
Pooled retrospective internal cross-validation. Symptom-enhanced AUROC: 0.829575; 95% bootstrap interval 0.801032–0.858900. The separate DS1 audio-only result was 0.850330; DS1 and DS2 are not interchangeable.
Context & limitations
These are research results, not clinical validation. Symptom availability, dataset shift and threshold selection limit how the numbers can be interpreted. Patient data and private prediction files are not published here.
04 / SUMMARY
The result depends on the patient-level split, the input features and the uncertainty around the reported estimates.
OpenMP and CUDA implementations.
Count Gliders / Histogram / Emboss