Machine learning applied to multi-sensor information to reduce false alarm rate in the ICU.

Studies reveal that the false alarm rate (FAR) demonstrated by intensive care unit (ICU) vital signs monitors ranges from 0.72 to 0.99. We applied machine learning (ML) to ICU multi-sensor information to imitate a medical specialist in diagnosing patient condition. We hypothesized that applying this data-driven approach to medical monitors will help reduce the FAR even when data from sensors are missing. An expert-based rules algorithm identified and tagged in our dataset seven clinical alarm scenarios. We compared a random forest (RF) ML model trained using the tagged data, where parameters (e.g., heart rate or blood pressure) were (deliberately) removed, in detecting ICU signals with the full expert-based rules (FER), our ground truth, and partial expert-based rules (PER), missing these parameters. When all alarm scenarios were examined, RF and FER were almost identical. However, in the absence of one to three parameters, RF maintained its values of the Youden index (0.94-0.97) and positive predictive value (PPV) (0.98-0.99), whereas PER lost its value (0.54-0.8 and 0.76-0.88, respectively). While the FAR for PER with missing parameters was 0.17-0.39, it was only 0.01-0.02 for RF. When scenarios were examined separately, RF showed clear superiority in almost all combinations of scenarios and numbers of missing parameters. When sensor data are missing, specialist performance worsens with the number of missing parameters, whereas the RF model attains high accuracy and low FAR due to its ability to fuse information from available sensors, compensating for missing parameters.

Repeated administration of vanilloid receptor TRPV1 antagonists attenuates hyperthermia elicited by TRPV1 blockade.

Capsaicin, the active ingredient in some pain-relieving creams, is an agonist of a nonselective cation channel known as the transient receptor potential vanilloid type 1 (TRPV1). The pain-relieving mechanism of capsaicin includes desensitization of the channel, suggesting that TRPV1 antagonism may be a viable pain therapy approach. In agreement with the above notion, several TRPV1 antagonists have been reported to act as antihyperalgesics. Here, we report the in vitro and in vivo characterization of a novel and selective TRPV1 antagonist, N-(4-[6-(4-trifluoromethyl-phenyl)-pyrimidin-4-yloxy]-benzothiazol-2-yl)-acetamide I (AMG 517), and compare its pharmacology with that of a closely related analog, tert-butyl-2-(6-([2-(acetylamino)-1,3-benzothiazol-4-yl]oxy)pyrimidin-4-yl)-5-(trifluoromethyl)phenylcarbamate (AMG8163). Both AMG 517 and AMG8163 potently and completely antagonized capsaicin, proton, and heat activation of TRPV1 in vitro and blocked capsaicin-induced flinch in rats in vivo. To support initial clinical investigations, AMG 517 was evaluated in a comprehensive panel of toxicology studies that included in vivo assessments in rodents, dogs, and monkeys. The toxicology studies indicated that AMG 517 was generally well tolerated; however, transient increases in body temperature (hyperthermia) were observed in all species after AMG 517 dosing. To further investigate this effect, we tested and showed that the antipyretic, acetaminophen, suppressed the hyperthermia caused by TRPV1 blockade. We also showed that repeated administration of TRPV1 antagonists attenuated the hyperthermia response, whereas the efficacy in capsaicin-induced flinch model was maintained. In conclusion, these studies suggest that the transient hyperthermia elicited by TRPV1 blockade may be manageable in the development of TRPV1 antagonists as therapeutic agents. However, the impact of TRPV1 antagonist-induced hyperthermia on their clinical utility is still unknown.