. Author manuscript; available in PMC: 2022 Mar 21.

Published in final edited form as: IEEE Rev Biomed Eng. 2017 Oct 24;11:2–20. doi: 10.1109/RBME.2017.2763681

Table VI. Publicly Available Datasets Used to Assess BR Algorithms.

Dataset	Ref	No subjs	Age	ECG	PPG	Resp sigs	Type of breathing	Level of illness
Datasets containing breath annotations
BIDMC	[31]	53	adult	Y	Y	ImP	spont, vent	critical
CapnoBase	[8]	42	paed, adult	Y	Y	CO₂	spont, vent	surgery, anesthesia
Datasets without breath annotations
MIMIC-II	[234]	23,180	paed, adult	Y	Y	ImP	spont, vent	critical
MGH/MF	[235]	250	paed, adult	Y	N	ImP, CO₂	spont, vent	critical
MIMIC	[236]	72	unk	Y	Y	ImP	spont, vent	critical
VORTAL	[18]	57	adult	Y	Y	ImP, Press	spont	healthy
Fantasia	[237]	40	adult	Y	N	unk	spont	healthy
UCD Sleep Apnea	[238]	25	adult	Y	N	flow	spont	healthy, apnea
CEBS	[239]	20	adult	Y	N	piezo	spont	healthy
ECG and resp	[240]	20	adult	Y	N	flow, pleth	spont	healthy
MIT-BIH Polysomnographic	[241]	18	adult	Y	N	flow	spont, vent	healthy, apnea
Apnea-ECG	[242]	8	adult	Y	Y	InP, flow	spont	healthy, apnea
Portland State	[243]	1	paed	Y	Y	unk	unk	critical

Definitions: Age—pediatric (paed); respiratory signals (Resp Sigs)—capnometry (CO₂), piezoresistive thoracic band (piezo), oral or nasal flow (flow), oral–nasal pressure (press), inductance plethysmography (InP), impedance pneumography (ImP), body plethysmography (pleth); Breathing—spontaneous (spont), ventilated (vent); unknown (unk).