Machine learning predicts lung recruitment in acute respiratory distress syndrome using single lung CT scan

Pennati, Francesca; Aliverti, Andrea; Pozzi, Tommaso; Gattarello, Simone; Lombardo, Fabio; Coppola, Silvia; Chiumello, Davide

doi:10.1186/s13613-023-01154-5

Annals of Intensive Care

Table 1 Baseline characteristics at 5 cmH₂O of PEEP in patients divided according to lung potential recruitment (LPR)

From: Machine learning predicts lung recruitment in acute respiratory distress syndrome using single lung CT scan

Variables	Recruiters (LPR > 15%)	Non-recruiters (LPR ≤ 15%)	p-value
Demographic
Number	110	111
Age (years)	62 (47, 74)	61 (48, 72)	0.317
Male sex (%)	44	56	0.739
Weight (kg)	75 (61, 85)	75 (61, 89)	0.718
BMI (kg/m2)	25 (22, 28)	25 (22, 29)	0.894
SAPS II	43 ± 17	42 ± 13	0.406
Origin of ARDS			0.001
Pulmonary (%)	73	51
Extrapulmonary (%)	27	49
ARDS severity			< 0.001
Mild (%)	61	27
Moderate (%)	28	42
Severe (%)	11	31
Ventilatory parameters
Tidal volume (TV, ml)	500 (425, 560)	522 (461, 600)	0.013
TV per ideal body weight (mL/kg)	7.9 (6.7, 8.5)	8.0 (7.3, 9.2)	0.017
Respiratory rate (breath per minute)	16 (14, 20)	16 (14, 18)	0.541
Minute ventilation (L/min)	8.3 (6.9, 9.9)	8.4 (7.6, 10.0)	0.252
Peak inspiratory pressure (cmH₂O)	28 (23, 34)	26 (23, 29)	0.029
Plateau pressure (cmH₂O)	19 (16, 22)	18 (15, 20)	0.030
Physiological dead space	0.66 ± 0.13	0.57 ± 0.12	< 0.001
Respiratory mechanics
Driving pressure (cmH₂O)	13 (10, 16)	12 (10, 15)	0.112
Respiratory system elastance (cmH₂O/L)	27 (21, 35)	24 (19, 30)	0.007
Mechanical power (J/(min kg))	15 (11, 20)	15 (11, 18)	0.890
Respiratory partitioned mechanics
Lung elastance (cmH₂O/L)	21(14, 28)	18 (13, 24)	0.048
Chest wall elastance (cmH₂O/L)	6 (4, 9)	6 (4, 8)	0.995
Gas exchange
PaCO₂ (mmHg)	47 (41, 53)	42 (38, 49)	< 0.001
PaO₂ (mmHg)	68 (60, 75)	76 (65, 91)	< 0.001
PaO₂/FiO₂	113 (84, 144)	163 (119, 207)	< 0.001
CT parameters
Total lung volume (mL)	2377 (2073, 2798)	2701 (2056, 3201)	0.022
Total lung weight (g)	1598 (1278, 1968)	1319 (1116, 1472)	< 0.001
Total lung gas volume (mL)	736 (504, 979)	1233 (912,1798)	< 0.001
Not aerated lung tissue (%)	52 ± 14	35 ± 14	< 0.001
Poorly aerated lung tissue (%)	31 (21, 39)	28 (22, 39)	0.652
Well aerated lung tissue (%)	14 (10, 22)	34 (26, 42)	< 0.001
Overaerated lung tissue (%)	0.00 (0.00, 0.04)	0.03 (0.00, 0.29)	< 0.001

Continuous data are expressed as mean ± SD or median (interquartile range), while categorical data are expressed as %. Student’s t test or Mann–Whitney rank-sum tests and Chi-square test were used as appropriate

Back to article page