Improvements of central respiratory events, Cheyne–Stokes respiration and oxygenation in patients hospitalized for acute decompensated heart failure

SLEEP-SEE-THROUGH: Explainable Deep Learning for Sleep Event Detection and Quantification From Wearable Somnography

Evidence is rapidly accumulating that multifactorial nocturnal monitoring, through the coupling of wearable devices and deep learning, may be disruptive for early diagnosis and assessment of sleep disorders. In this work, optical, differential air-pressure and acceleration signals, acquired by a chest-worn sensor, are elaborated into five somnographic-like signals, which are then used to feed a deep network. This addresses a three-fold classification problem to predict the overall signal quality (normal, corrupted), three breathing-related patterns (normal, apnea, irregular) and three sleep-related patterns (normal, snoring, noise). In order to promote explainability, the developed architecture generates additional information in the form of qualitative (saliency maps) and quantitative (confidence indices) data, which helps to improve the interpretation of the predictions. Twenty healthy subjects enrolled in this study were monitored overnight for approximately ten hours during sleep. Somnographic-like signals were manually labeled according to the three class sets to build the training dataset. Both record- and subject-wise analyses were performed to evaluate the prediction performance and the coherence of the results. The network was accurate (0.96) in distinguishing normal from corrupted signals. Breathing patterns were predicted with higher accuracy (0.93) than sleep patterns (0.76). The prediction of irregular breathing was less accurate (0.88) than that of apnea (0.97). In the sleep pattern set, the distinction between snoring (0.73) and noise events (0.61) was less effective. The confidence index associated with the prediction allowed us to elucidate ambiguous predictions better. The saliency map analysis provided useful insights to relate predictions to the input signal content. While preliminary, this work supported the recent perspective on the use of deep learning to detect particular sleep events in multiple somnographic signals, thus representing a step towards bringing the use of AI-based tools for sleep disorder detection incrementally closer to clinical translation.

Prognostic Value of Cheyne-Stokes Respiration and Nutritional Status in Acute Decompensated Heart Failure

Malnutrition frequently coexists with heart failure (HF), leading to series of negative consequences. Cheyne-Stokes respiration (CSR) is predominantly detected in patients with HF. However, the effect of CSR and malnutrition on the long-term prognosis of patients with acute decompensated HF (ADHF) remains unclear. We enrolled 162 patients with ADHF (median age, 62 years; 78.4% men). The presence of CSR was assessed using polysomnography and the controlling nutritional status score was assessed to evaluate the nutritional status. Patients were divided into four groups based on CSR and malnutrition. The primary outcome was all-cause mortality. In total, 44% of patients had CSR and 67% of patients had malnutrition. The all-cause mortality rate was 26 (16%) during the 35.9 months median follow-up period. CSR with malnutrition was associated with lower survival rates (log-rank p < 0.001). Age, hemoglobin, albumin, lymphocyte count, total cholesterol, triglyceride, low-density lipoprotein cholesterol, creatinine, estimated glomerular filtration rate, B-type natriuretic peptide, administration of loop diuretics, apnea-hypopnea index and central apnea-hypopnea index were significantly different among all groups (p < 0.05). CSR with malnutrition was independently associated with all-cause mortality. In conclusion, CSR with malnutrition is associated with a high risk of all-cause mortality in patients with ADHF.

Positionspapier „Schlafmedizin in der Kardiologie“, Update 2021

Es hat sich viel getan in der Welt der Schlafmedizin in der Kardiologie, weshalb eine vollwertige Überarbeitung des Positionspapiers „Schlafmedizin in der Kardiologie“ erforderlich wurde. In der aktuellen neuartigen Version finden sich nicht nur alle verfügbaren Studien, Literaturstellen und Updates zu Pathophysiologie, Diagnostik- und Therapieempfehlungen, sondern auch Ausblicke auf neue Entwicklungen und zukünftige Forschungserkenntnisse. Dieses überarbeitete Positionspapier gibt Empfehlungen für Diagnostik und Therapie von Patienten mit kardiovaskulären Erkrankungen mit schlafassoziierten Atmungsstörungen und erteilt darüber hinaus einen fundierten Überblick über verfügbare Therapien und Evidenzen, gibt aber ebenso Ratschläge wie mit Komorbiditäten umzugehen ist. Insbesondere enthält dieses überarbeitete Positionspapier aktualisierte Stellungnahmen zu schlafassoziierten Atmungsstörungen bei Patienten mit koronarer Herzerkrankung, Herzinsuffizienz, arterieller Hypertonie, aber auch für Patienten mit Vorhofflimmern. Darüber hinaus finden sich erstmals Empfehlungen zur Telemedizin als eigenes, neues Kapitel. Dieses Positionspapier bietet Kardiologen sowie Ärzten in der Behandlung von kardiovaskulären Patienten die Möglichkeit einer evidenzbasierten Behandlung der wachsend bedeutsamen und mit zunehmender Aufmerksamkeit behafteten Komorbidität schlafassoziierter Atmungsstörungen. Und nicht zuletzt besteht mit diesem neuen Positionspapier eine enge Verknüpfung mit dem neuen Curriculum Schlafmedizin der Deutschen Gesellschaft für Kardiologie, weshalb dieses Positionspapier eine Orientierung für die erworbenen Fähigkeiten des Curriculums im Umgang von kardiovaskulären Patienten mit schlafassoziierten Atmungsstörungen darstellt.

Characteristics and circadian distribution of cardiac arrhythmias in patients with heart failure and sleep-disordered breathing

BACKGROUND

Cardiac arrhythmias and sleep-disordered breathing (SDB) are common comorbidities in heart failure with reduced ejection fraction (HFrEF). However, understanding of the association between arrhythmias and SDB is poor. This study assessed the occurrence and circadian distribution of ventricular arrhythmias in HFrEF patients with and without SDB.

METHODS

This retrospective analysis included HFrEF patients admitted for unattended overnight cardiorespiratory polygraphy and 24-h Holter-ECG recording. Holter-ECG data (events/h) were categorized by time of day: morning, 06:00-13:59; afternoon, 14:00-21:59; nighttime, 22:00-05:59. Respiratory events were expressed using the apnea-hypopnea index (AHI) and an AHI ≥ 15/h was categorized as moderate to severe SDB.

RESULTS

167 patients were included (82% male, age 65 ± 10.4 years, left ventricular ejection fraction 30.9 ± 7.9%); SDB was predominantly central sleep apnea (CSA) in 45.5%, obstructive sleep apnea (OSA) in 23.9% or none/mild (nmSDB) in 17.4%. Morning premature ventricular contractions (PVCs) were detected significantly more frequently in CSA versus nmSDB patients (44.4/h versus 1.8/h; p = 0.02). Non-sustained VT was more frequent in patients with CSA versus versus OSA or nmSDB (17.9 versus 3.2 or 3.2%/h; p = 0.003 and p = 0.005, respectively). There was no significant variation in VT occurrence by time of day in HFrEF patients with CSA (p = 0.3). CSA was an independent predictor of VT occurrence in HFrEF in multivariate logistic regression analysis (odds ratio 4.1, 95% confidence interval 1.5-11.4, p = 0.007).

CONCLUSION

CSA was associated with VT occurrence irrespective of sleep/wake status in HFrEF patients, and independently predicted the occurrence of VT. This association may contribute to chances by which CSA increases sudden death risk in HFrEF patients.

ResCSRF: Algorithm to Automatically Extract Cheyne-Stokes Respiration Features From Respiratory Signals

OBJECTIVE

Cheyne-Stokes respiration (CSR) related features are significantly associated with cardiac dysfunction. Scoring of these features is labor intensive and time-consuming. To automate the scoring process, an algorithm (ResCSRF) has been developed to extract these features from nocturnal measurement of respiratory signals.

METHODS

ResCSRF takes four signals (nasal flow, thorax, abdomen, and finger oxygen saturation) as input. It first detects CSR cycles and then calculates the respiratory features (cycle length, lung-to-periphery circulation time, and time to peak flow). It outputs nightly statistics (mean, median, standard deviation, and percentiles) of these features. It was developed and blindly tested on a group of 49 chronic heart failure patients undergoing overnight in-home unattended respiratory polygraphy recordings.

RESULTS

The performance of ResCSRF was evaluated against manual expert scoring (ES) (consensus between two independent sleep scorers). In terms of percentage of CSR per recording, the mean difference [reproducibility coefficient (RPC)] between ResCSRF and ES was 0.5(6.4) and 0.5(8.1) for development and test set, respectively. The nightly statistics of CSR-related features output by ResCSRF showed high correlation with ES on the blind test set with the mean difference of less than 3 s and RPC of less than 7 s.

CONCLUSIONS

These results indicate that ResCSRF is capable of automating the scoring of CSR-related features and could potentially be implemented into a remote monitoring system to regularly monitor patients' cardiac function.

Impact of sleep-disordered breathing in patients with acute myocardial infarction: a retrospective analysis

Sleep-disordered breathing (SDB) is associated with an increased risk of cardiovascular events. Previous studies showed that severe SDB has a negative impact on myocardial salvage and progression of left ventricular dysfunction after acute myocardial infarction (AMI). This study investigated the frequency of SDB and the effects of SDB on left ventricular function after AMI. This retrospective study enrolled all patients with AMI who had undergone cardiorespiratory polygraphy for SDB diagnosis. The apnea-hypopnea index was used as a standard metric of SDB severity. SDB was classified as mild (apnea-hypopnea index >5 to <15 per h), moderate (≥15 to <30 per h) or severe (apnea-hypopnea index ≥30 per h). According to the majority of events, SDB was classified as predominant obstructive sleep apnea, central sleep apnea or mixed sleep apnea (mixed SDB). A total of 223 patients with AMI (112 with ST elevation and 111 without ST elevation; 63.2 ± 11.2 years, 82% male, left ventricular ejection fraction 49 ± 12%) were enrolled. SDB was present in 85.6%, and was moderate-to-severe in 63.2%; 40.8% had obstructive sleep apnea, 41.7% had central sleep apnea and 3.1% had mixed SDB. Left ventricular ejection fraction was lower in patients with AMI with severe SDB (45 ± 14%) versus those without SDB (57 ± 7%; P < 0.005). In addition, lower left ventricular ejection fraction (≤45%) was associated with increased frequency (apnea-hypopnea index ≥5 per h in 96%) and severity (apnea-hypopnea index ≥30 per h in 48%) of SDB in general and a higher percentage of central sleep apnea (57%) in particular. SDB is highly frequent in patients with AMI. SDB severity appeared to be linked to impaired left ventricular function, especially in patients with central sleep apnea.

Impairment of pulmonary diffusion correlates with hypoxemic burden in central sleep apnea heart failure patients

PURPOSE

Central sleep apnea (CSA) and Cheyne-Stokes respiration (CSR) are highly prevalent in heart failure (HF) and are linked to increased mortality. Impaired pulmonary diffusion capacity [DLCO] and [KCO]) have been suggested to play a key role in CSA-CSR pathophysiology. This study investigated the relationship between HF, CSR, DLCO and KCO in well-characterized HF patients.

METHODS

This prospective study included HF patients with CSR, all patients underwent full overnight polysomnography (PSG) and lung function testing.

RESULTS

A total of 100 patients were included (age 70.7±9.7years, 95% male, body mass index 28.9±5.3kg/m², left ventricular ejection fraction 33.5±7.7%, New York Heart Association class III 65%. DLCO and oxygenation were significantly correlated with hypoxemic burden (p<0.05). Mean oxygen saturation, oxygen desaturation, C-reactive protein level and pH were significantly associated with CSA-CSR severity (p<0.05).

CONCLUSION

The finding that lung diffusion capacity is significantly associated with hypoxemic burden in HF patients with CSA-CSR highlights the important of lung function in HF patients.

Long-Term Experience with First-Generation Implantable Neurostimulation Device in Central Sleep Apnea Treatment

BACKGROUND

Sleep-disordered breathing (SDB) and Cheyne-Stokes respiration (CSR) are associated with shorter survival in patients with heart failure. A novel treatment method for this patient group is unilateral phrenic nerve stimulation by the remedē® system (Respicardia Inc., Minnetonka, MN, USA), a transvenously implantable neurostimulation device, which has recently been studied in a large randomized, controlled trial. Previous literature has shown efficacy and safety of the treatment with this first-generation device, but hardly any data are available on long-term clinical parameters, the remedē® device's battery lifetime, device exchangeability, lead position stability, surgical accessibility, and manageability.

METHODS

We performed remedē® device replacements in consecutive patients for battery depletion, and documented clinical parameters, longevity, operation procedure, complications, and difficulties.

RESULTS

All patients were on neurostimulation treatment by phrenic nerve neurostimulation when device replacement became necessary. Apnea-hypopnea index (from 45 ± 4/h to 9 ± 4/h), oxygen-desaturation index (from 35 ± 7/h to 7 ± 6/h), and time spent with oxygen saturation of <90% (T < 90% from 5 ± 7% to 0 ± 0%) were improved and improvements remained constant throughout the 4-year follow-up. Mean battery life was 4.2 ± 0.2 years and mean replacement procedure time was 25 ± 5.1 minutes. Apart from conventional X-ray documentation of stable lead positions in a long-term setting, no radiation or contrast dye usage was needed and no major complications occurred. In addition, clinical exercise capacity and sleepiness symptoms improved.

CONCLUSIONS

Novel remedē® device shows sustained therapy efficacy and safety in terms of stable lead positions over 4 years. Long-term phrenic nerve neurostimulation therapy for central SDB/CSR appears feasible in a clinical routine setting.