Using a machine learning algorithm to detect depressed ejection fraction from a single-lead ECG

Background

Multiple studies demonstrate the benefit of intervention for left ventricular ejection fraction (LVEF) below 40%, so the development of a low ejection fraction algorithm to detect LVEF below 40% can aid in early screening of initial asymptomatic Heart Failure with reduced Ejection Fraction (HFrEF).

Objective

To demonstrate the performance of a low ejection fraction algorithm using single-lead ECG data to detect LVEF below 40%.

Methods

We collected 1325 single-lead ECG recordings (15s duration) at various chest positions using an electronic stethoscope from 197 patients. We analyzed these ECG recordings using a deep neural network model trained on individual leads extracted from a 12-lead ECG to discriminate left ventricular ejection fractions (EFs) above or below different thresholds. We compared the model output to ejection fraction measured using echocardiograms.

Results

Across all recordings from all patients, we obtained an AUROC of 0.89, with a sensitivity of 88% and specificity of 74% using a model output threshold of 0.35 (Figure 1). The AUROC of recordings taken at different orientations and stances ranged from 0.85 to 0.92 (Table 1), with a sensitivity of at least 78% and specificity of at least 66% at any orientation.

Conclusion

Using a single lead ECG measured by an electronic stethoscope and a deep neural network model, we were able to detect depressed ejection fraction (≤40%) with a sensitivity of 88% and specificity of 74%. This work demonstrates the utility of a low-cost electronic stethoscope and machine learning for early screening and detection of depressed left ventricular ejection fraction.

Funding Acknowledgement

Type of funding sources: Private company. Main funding source(s): Eko Health

Curcumin loaded polycaprolactone-/polyvinyl alcohol-silk fibroin based electrospun nanofibrous mat for rapid healing of diabetic wound: An in-vitro and in-vivo studies

Electrospinning is emerging as a versatile technique nanofibers fabrication because due to their unique properties such as large surface area to volume ratio, porosity and maintaining moist wound environment, the nanofibers are able to deliver sustained drug release and oxygen to the wound for rapid healing of diabetic wound. The present work was aimed to prepare and evaluate silk fibroin-curcumin based nanofiber in combination with polycaprolactone (PCL) and polyvinyl alcohol (PVA) which helped to strengthen the wound healing properties of nanofiber. Silk fibroin is a naturally occurring polymer was selected one polymer for making nanofibrous mat due to its unique properties such as biodegradability, permeability, oxygen supply and maintain moisture content in the wound. SEM results showed diameters of fibers varied in the range between 200 and 350 nm and their tensile strength ranged from 12.41 to 16.80 MP. The nanofibers were causing sustained release of curcumin for many hours. The in-vivo wound healing studies in streptozotocin-induced diabetic mice showed rapid wound healing efficacy as compared to conventional formulations. Furthermore, the histopathological studies evidenced its ability to restore the normal skin structure and histological conditions of tissues. The silk fibroin-based nanofiber wound dressing, therefore appears to be an ideal preparation, in combination with curcumin, because it blends the anti-oxidant, anti-inflammatory properties of curcumin. Therefore, it was concluded that the silk fibroin-based nanofiber loaded with curcumin has great healing potential in diabetic wound.

Hospital Based Study of Poisoning Among Children, 1 to 18 Years of Age in Eastern Nepal

IntroductionThe poisoning among children had been a major cause of hospital admission and also a significant global health problem throughout the world including Nepal. There are various studies suggestive of incidence rate of poisoning in children ranging from 0.74% to 3%. In older children, poisoning is usually intentional where as in infants and younger children it is accidental in nature There are few studies done in Nepal and they are mainly from Kathmandu valley, this study will add more information from eastern region of Nepal.ObjectiveThis study aimed to find out the type of poisoning among children and adolescents, manner of poisoning and its outcomes.MethodologyThis was a retrospective study carried out at Nobel Medical College Teaching Hospital over a period of one year starting from  1st  August 2014 to 31 July 2015. This study included all the children and adolescents of 1 to 18 years of age admitted to pediatric ward and ICU for the management of poisoning. The information including name of poisons, manner of poisoning, socio-economic and demographic factors related to poisoning, geographical distribution of sample population, durations of hospital stay and outcome of treatment were retrieved from the records. The data was analyzed using SPSS. Susceptibility, urinary tract infection, uropathogens.ResultsA total of 64 cases were studied. It included almost the same number of males (n=31) and females (n=33). The age group of 13-18 years had the largest prevalence of poisoning (n=22). Organo-phosphorus (n=25) was the commonest poison. As the study was conducted in a tertiary care center of eastern Nepal located at Biratnagar, Morang district, a majority of the cases (39) were from Morang. A total of 41% children required hospital stay for more than 7 days where as 37% of them required 5-7 days of hospital stay and only 22% children stayed for 1-4 days in the hospital. No mortality was noted.ConclusionThe pediatric poisoning in eastern Nepal is a significant public health issue and nature of poisoning was accidental followed by suicidal. The commonest age group involved is adolescence, which highlights the need for education and implementation of suicidal prevention programming this region. The higher prevalence of accidental poisoning in first two year of life necessitates the need of child safety and injury prevention program in the community.  Birat Journal of Health Sciences Vol.2/No.1/Issue 2/ Jan - April 2017, Page: 138-141

Routine upper GI series after gastric bypass does not reliably identify anastomotic leaks or predict stricture formation

BACKGROUND

Many surgeons who perform Roux-en-Y gastric bypass (RYGB) for morbid obesity routinely obtain an upper gastrointestinal (GI) series in the early postoperative period to search for anastomotic leaks and signs of stricture formation at the gastrojejunostomy. We hypothesized that this practice is unreliable.

METHODS

We analyzed 654 consecutive RYGBs, of which 63% were completed laparoscopically. An upper GI series was obtained in 634 (97%) patients. The radiographic findings (leak or delayed emptying) were compared with clinical outcomes (leak or stricture formation) to calculate the sensitivity and specificity. Univariate analysis identified risk factors for leaks or stricture formation; events were too few for multivariate analysis.

RESULTS

Of 634 routine upper GI series, anastomotic leaks at the gastrojejunostomy were diagnosed in 5 (0.8%); 2 of these 5 were later reinterpreted as artifacts. Four leaks were not seen on the initial upper GI series, yielding an overall sensitivity of 43% and a positive predictive value (PPV) of 60%. Univariate analysis showed that cases done early (odds ratio [OR] 5.4 for the first 100 cases, p = 0.02) and prolonged operating time (OR 7.8 for cases >or= 300 min, p = 0.01) were associated with leaks. Emptying into the Roux-en-Y limb was delayed in 127 (20%) of the upper GI series. Strictures requiring dilatation developed in 16 (2.4%) patients. The PPV of delayed emptying for stricture formation was 6%. Risk factors for stricture formation included stapled anastomosis (OR 7.8, p = 0.002), surgeon inexperience (OR 2.9 for first 50 cases, p = 0.04), and delayed emptying (OR 3.3; p = 0.02).

CONCLUSIONS

Because the incidence of anastomotic complications and the sensitivity of upper GI series were both low, routine upper GI series did not reliably identify leaks or predict stricture formation. A selective approach, whereby imaging is reserved for patients with clinical evidence of a leak or stricture, may be more appropriate.