MSTFA / Industries
MSTFA Industries
Navigate
Home Systems Capabilities Company Investors
More
Updates Talent Contact
MSTFA/Capabilities

Four domains. One platform.

Every MSTFA program shares one microcontroller platform, one electrode chemistry, one signal-processing methodology, and one data plane. Engineering effort compounds instead of fragmenting.

01

Pulmonary function measurement

Differential-pressure spirometry against ATS/ERS acceptability and repeatability criteria; per-user baseline modeling versus population reference equations (NHANES 2007–2012, 16,596 subjects).

02

Neural & autonomic monitoring

Microvolt-scale EEG on an ADS1299 24-bit front end, fused with heart-rate variability, electrodermal activity, and peripheral temperature into a continuous sympathetic-load index.

03

Swallowing & dysphagia detection

Surface EMG and vibroacoustic classification of swallowing events with videofluoroscopy as ground truth — building on published accuracies of 89–92%.

04

Clinical data infrastructure

HL7 FHIR R4-native ingestion, storage, and export. One canonical record per patient; LOINC coding where standardized; audit logging on every access.

05

Signal processing methodology

A common chain across programs: Nyquist-margin sampling, bandpass and notch filtering, artifact-epoch rejection, Welch power spectral density, windowed feature extraction.

06

Regulatory engineering

Explicit FDA 510(k) posture per device, ISO 13485 quality management planned for manufacture, ISO 10993 biocompatibility for skin-contact components.

Express interest

Shared Platform

Common architecture across every instrument.

ComputeESP32-S3 microcontroller with Bluetooth Low Energy, standardized across all three devices
Analog front endADS1299 24-bit biopotential AFE for microvolt-scale EEG acquisition (1 µVrms input-referred noise)
ElectrodesConductive hydrogel contact chemistry, shared between the neural and dysphagia programs
InferenceOn-device gradient-boosted models for low-latency classification; cloud-side sequence models for longitudinal prediction
InteroperabilityHL7 FHIR R4 observations via Plexus, with LOINC coding where standardized
Regulatory frameFDA 510(k) pathway for the three devices; ISO 13485 quality management planned for manufacture