A Decade of Research

Detecting Elevated Blood Pressure from a Sensor-Embedded Toilet Seat: A Mixture-of-Experts Classifier Evaluated in Home and Clinical Settings

Review Pending · Not yet peer reviewed

Wehbe H

Hypertension affects about half of U.S. adults and only one in five has it controlled, largely because every established measurement modality requires the patient to decide to measure. 1,101 Smart Seat recordings from 96 participants, collected under IRB-approved protocols with informed consent, were analyzed: 14 participants used a seat at home for 45 days with a supplied cuff (433 recordings) and 82 contributed 668 clinic recordings with simultaneous auscultation. Three models trained at different decision boundaries were combined by a fixed gating rule (a mixture of experts), with calls aggregated by vote against the 130/80 mmHg ACC/AHA boundary. Per participant, clinic sensitivity was 0.75 at specificity 0.94 (PPV 0.92, n = 32); at home, sensitivity was 0.42 with no false alarms across 40 evaluable participant-weeks. The results support use as a passive trigger for confirmatory cuff measurement, not a substitute for it: a flag is informative, but a quiet result is not reassuring.

Estimating Sit-to-Stand Power from a Single Smart Seat Recording: Validation Against the Supervised Chair-Stand Test in 122 Adults

Review Pending · Not yet peer reviewed

Mattis N

Lower-limb muscle power predicts mobility loss in older adults better than strength or mass, but it is measured with a supervised, stopwatch-timed chair-stand test, so it mostly goes unmeasured. 896 Smart Seat recordings from 124 adults aged 24 to 89, across two IRB-approved collections, were analyzed against a supervised chair-stand reference standard using only the seat's four corner load cells. Estimated relative power correlated with the chair-test reference at a Spearman rho of 0.54 (95% CI 0.39 to 0.67, n = 122), median absolute error 20.0%, and kept ranking people within a single age decade where age alone did not.

Passive Toilet Seat-Based Monitoring Versus Traditional Blood Pressure Telemonitoring in Older Adults: A Crossover Study

Roberts M, Sappington E, Yalcin A, Dobbs D, Peterson L, Tremblay R, VandeWeerd C

In a 12-week crossover study, 49 community-dwelling adults aged 55 to 85 with cardiovascular disease (69.4% with a history of hypertension) each used a passive toilet seat-based monitoring platform and traditional telemonitoring with a blood pressure cuff and scale for six weeks apiece. Adherence was significantly higher with the passive seat (93% vs 73%, p < .001), a pattern that held among participants with self-reported high blood pressure and elevated baseline readings. Seat adherence stayed stable across its phase, while cuff and scale adherence declined by the final study week.

In-Home Cardiovascular Monitoring System for Heart Failure: Comparative Study

Conn NJ, Schwarz KQ, Borkholder DA

A toilet seat–based cardiovascular monitoring system integrating ECG, ballistocardiogram, and photoplethysmogram demonstrated clinical-grade measurements of systolic and diastolic blood pressure, stroke volume, and peripheral blood oxygenation. Diastolic BP accuracy: 1.2 (SD 6.0) mm Hg; systolic BP: –2.7 (SD 6.6) mm Hg; stroke volume: –2.5 (SD 15.5) mL compared to echocardiogram gold standard; SpO₂ RMS error: 2.3%.

Nontraditional Electrocardiogram and Algorithms for Inconspicuous In-Home Monitoring: Comparative Study

Conn NJ, Schwarz KQ, Borkholder DA

Validated a toilet seat-based buttocks ECG against a simultaneous 12-lead clinical ECG in 25 normative and 29 heart failure subjects. Achieved clinical-grade accuracy for heart rate, heart rate variability, QRS duration, and corrected QT interval. A modified Pan–Tompkins algorithm yielded >99.82% sensitivity and positive predictive value across 900,059 beats on standard databases.