Development of a graphical user interface for automatic separation of human voice from Doppler ultrasound audio in diving experiments.

Doppler ultrasound (DU) is used in decompression research to detect venous gas emboli in the precordium or subclavian vein, as a marker of decompression stress. This is of relevance to scuba divers, compressed air workers and astronauts to prevent decompression sickness (DCS) that can be caused by these bubbles upon or after a sudden reduction in ambient pressure. Doppler ultrasound data is graded by expert raters on the Kisman-Masurel or Spencer scales that are associated to DCS risk. Meta-analyses, as well as efforts to computer-automate DU grading, both necessitate access to large databases of well-curated and graded data. Leveraging previously collected data is especially important due to the difficulty of repeating large-scale extreme military pressure exposures that were conducted in the 70-90s in austere environments. Historically, DU data (Non-speech) were often captured on cassettes in one-channel audio with superimposed human speech describing the experiment (Speech). Digitizing and separating these audio files is currently a lengthy, manual task. In this paper, we develop a graphical user interface (GUI) to perform automatic speech recognition and aid in Non-speech and Speech separation. This constitutes the first study incorporating speech processing technology in the field of diving research. If successful, it has the potential to significantly accelerate the reuse of previously-acquired datasets. The recognition task incorporates the Google speech recognizer to detect the presence of human voice activity together with corresponding timestamps. The detected human speech is then separated from the audio Doppler ultrasound within the developed GUI. Several experiments were conducted on recently digitized audio Doppler recordings to corroborate the effectiveness of the developed GUI in recognition and separations tasks, and these are compared to manual labels for Speech timestamps. The following metrics are used to evaluate performance: the average absolute differences between the reference and detected Speech starting points, as well as the percentage of detected Speech over the total duration of the reference Speech. Results have shown the efficacy of the developed GUI in Speech/Non-speech component separation.

An open-source framework for synthetic post-dive Doppler ultrasound audio generation.

Doppler ultrasound (DU) measurements are used to detect and evaluate venous gas emboli (VGE) formed after decompression. Automated methodologies for assessing VGE presence using signal processing have been developed on varying real-world datasets of limited size and without ground truth values preventing objective evaluation. We develop and report a method to generate synthetic post-dive data using DU signals collected in both precordium and subclavian vein with varying degrees of bubbling matching field-standard grading metrics. This method is adaptable, modifiable, and reproducible, allowing for researchers to tune the produced dataset for their desired purpose. We provide the baseline Doppler recordings and code required to generate synthetic data for researchers to reproduce our work and improve upon it. We also provide a set of pre-made synthetic post-dive DU data spanning six scenarios representing the Spencer and Kisman-Masurel (KM) grading scales as well as precordial and subclavian DU recordings. By providing a method for synthetic post-dive DU data generation, we aim to improve and accelerate the development of signal processing techniques for VGE analysis in Doppler ultrasound.

Deep Learning-Based Venous Gas Emboli Grade Classification in Doppler Ultrasound Audio Recordings.

OBJECTIVE: Doppler ultrasound (DU) is used to detect venous gas emboli (VGE) post dive as a marker of decompression stress for diving physiology research as well as new decompression procedure validation to minimize decompression sickness risk. In this article, we propose the first deep learning model for VGE grading in DU audio recordings. METHODS: A database of real-world data was assembled and labeled for the purpose of developing the algorithm, totaling 274 recordings comprising both subclavian and precordial measurements. Synthetic data was also generated by acquiring baseline DU signals from human volunteers and superimposing laboratory-acquired DU signals of bubbles flowing in a tissue mimicking material. A novel squeeze-and-excitation deep learning model was designed to effectively classify recordings on the 5-class Spencer scoring system used by trained human raters. RESULTS: On the real-data test set, we show that synthetic data pretraining achieves average ordinal accuracy of 84.9% for precordial and 90.4% for subclavian DU which is a 24.6% and 26.2% increase over training with real-data and time-series augmentation only. The weighted kappa coefficients of agreement between the model and human ground truth were 0.74 and 0.69 for precordial and subclavian respectively, indicating substantial agreement similar to human inter-rater agreement for this type of data. CONCLUSION: The present work demonstrates the first application of deep-learning for DU VGE grading using a combination of synthetic and real-world data. SIGNIFICANCE: The proposed method can contribute to accelerating DU analysis for decompression research.

Testing of full face snorkel masks to examine recreational snorkeler deaths.

A recent rise in snorkeling-related deaths in Hawaii has inspired several bans on full face snorkel masks (FFSMs). However, while there are theories to explain the deaths, little physiological data exists about the way the FFSMs provide gas to an exercising subject. To evaluate the safety of the FFSM concept, this study was designed to test how use of a full face snorkel mask (FFSM) may be physiologically different than use of a conventional snorkel, and to assess if any of those differences could lead to increased risk for the snorkeler. Ten (10) volunteer human subjects were tested using a variety of commercially available FFSMs, with real-time monitoring of blood oxygen saturation (SpO2), inspired airway pressure, and inspired and expired levels of carbon dioxide and oxygen. Two of the three FFSM design types were shown not to function as advertised, but none of the masks provided physiologically problematic gas supplies to the snorkelers. While this testing yielded no conclusive "smoking gun" to explain the snorkeler deaths, some of the mask models showed patterns of increasing breathing resistance with water intrusion because of a shared design characteristic, and this increased resistance could potentially create elevated levels of respiratory distress to snorkelers during real-world use.

The Dewey Monitor: Pulse Oximetry can Warn of Hypoxia in an Immersed Rebreather Diver in Multiple Scenarios.

Divers who wish to prolong their time underwater while carrying less equipment often use devices called rebreathers, which recycle the gas expired after each breath instead of discarding it as bubbles. However, rebreathers' need to replace oxygen used by breathing creates a failure mechanism that can and frequently does lead to hypoxia, loss of consciousness, and death. The purpose of this study was to determine whether a pulse oximeter could provide a useful amount of warning time to a diver with a rebreather after failure of the oxygen addition mechanism. Twenty-eight volunteer human subjects breathed on a mixed-gas rebreather in which the oxygen addition system had been disabled. The subjects were immersed in water in four separate environmental scenarios, including cold and warm water, and monitored using pulse oximeters placed at multiple locations. Pulse oximeters placed on the forehead and clipped on the nasal ala provided a mean of 32 s (±10 s SD) of warning time to divers with falling oxygen levels, prior to risk of loss of consciousness. These devices, if configured for underwater use, could provide a practical and inexpensive alarm system to warn of impending loss of consciousness in a manner that is redundant to the rebreather.

The Dewey monitor: Pulse oximetry can independently detect hypoxia in a rebreather diver.

Rebreather diving has one of the highest fatality rates per man hour of any diving activity in the world. The leading cause of death is hypoxia, typically from equipment or procedural failures. Hypoxia causes very few symptoms prior to causing loss of consciousness. Additionally, since the electronics responsible for controlling oxygen levels in rebreathers often control their alarm systems, frequently divers do not receive any external warnings. This study investigated the use of a forehead pulse oximeter as an independent warning device in the event of rebreather failure. Ten test subjects (seven male, three female, median age 29, range 26-35) exercised at a targeted rate of 2 L/minute oxygen consumption while on a non-functional rebreather breathing loop (mean consumption achieved 2.09 ± 0.36 L/minute). Each subject was tested both at the surface and at pressurized depth of 77 fsw (starting pO2=0.7 atm). The data show that a pulse oximeter could be used to provide an Mk 16 rebreather diver with a minimum mean of 49 seconds (± 17 seconds SD) of warning time after a noticeable change in blood oxygen saturation (SpO2 ≤ 95%) but before any risk of loss of consciousness (calculated SpO2 ≤ 80%), so that the diver may take mitigating actions. No statistical difference in warning time was found between the tests at surface and at 77 fsw (P=0.46).

Air blast injuries killed the crew of the submarine H.L. Hunley.

The submarine H.L. Hunley was the first submarine to sink an enemy ship during combat; however, the cause of its sinking has been a mystery for over 150 years. The Hunley set off a 61.2 kg (135 lb) black powder torpedo at a distance less than 5 m (16 ft) off its bow. Scaled experiments were performed that measured black powder and shock tube explosions underwater and propagation of blasts through a model ship hull. This propagation data was used in combination with archival experimental data to evaluate the risk to the crew from their own torpedo. The blast produced likely caused flexion of the ship hull to transmit the blast wave; the secondary wave transmitted inside the crew compartment was of sufficient magnitude that the calculated chances of survival were less than 16% for each crew member. The submarine drifted to its resting place after the crew died of air blast trauma within the hull.

Did a "lucky shot" sink the submarine H.L. Hunley?

The H.L. Hunley was the first submarine to be successful in combat, sinking the Union vessel Housatonic outside Charleston Harbor in 1864 during the Civil War. However, despite marking a milestone in military history, little is known about this vessel or why it sank. One popular theory is the "lucky shot" theory: the hypothesis that small arms fire from the crew of the Housatonic may have sufficiently damaged the submarine to sink it. However, ballistic experiments with cast iron samples, analysis of historical experiments firing Civil War-era projectiles at cast iron samples, and calculation of the tidal currents and sinking trajectory of the submarine indicate that this theory is not likely. Based on our results, the "lucky shot" theory does not explain the sinking of the world's first successful combat submarine.

Did the crew of the submarine H.L. Hunley suffocate?

On the evening of February 17th, 1864, the Confederate submarine H.L. Hunley attacked the Union ship USS Housatonic outside Charleston, South Carolina and became the first submarine in history to successfully sink an enemy ship in combat. One hypothesis for the sinking of the Confederate submarine H.L. Hunley is that the crew, in the enclosed vessel, suffered a lack of oxygen and suffocated. This study estimates the effects of hypoxia and hypercapnia on the crew based on submarine gas volume and crew breathing dynamics. The calculations show the crew of the Hunley had a minimum of 10 min between the onset of uncomfortable hypercapnia symptoms and danger of loss of consciousness from hypoxia. Based on this result and the location of the crew when discovered, hypoxia and hypercapnia do not explain the sinking of the world's first successful combat submarine.

Human Injury Criteria for Underwater Blasts.

Underwater blasts propagate further and injure more readily than equivalent air blasts. Development of effective personal protection and countermeasures, however, requires knowledge of the currently unknown human tolerance to underwater blast. Current guidelines for prevention of underwater blast injury are not based on any organized injury risk assessment, human data or experimental data. The goal of this study was to derive injury risk assessments for underwater blast using well-characterized human underwater blast exposures in the open literature. The human injury dataset was compiled using 34 case reports on underwater blast exposure to 475 personnel, dating as early as 1916. Using severity ratings, computational reconstructions of the blasts, and survival information from a final set of 262 human exposures, injury risk models were developed for both injury severity and risk of fatality as functions of blast impulse and blast peak overpressure. Based on these human data, we found that the 50% risk of fatality from underwater blast occurred at 302±16 kPa-ms impulse. Conservatively, there is a 20% risk of pulmonary injury at a kilometer from a 20 kg charge. From a clinical point of view, this new injury risk model emphasizes the large distances possible for potential pulmonary and gut injuries in water compared with air. This risk value is the first impulse-based fatality risk calculated from human data. The large-scale inconsistency between the blast exposures in the case reports and the guidelines available in the literature prior to this study further underscored the need for this new guideline derived from the unique dataset of actual injuries in this study.

Underwater blast injury: a review of standards.

The first cases of underwater blast injury appeared in the scientific literature in 1917, and thousands of service members and civilians were injured or killed by underwater blast during WWII. The prevalence of underwater blast injuries and occupational blasting needs led to the development of many safety standards to prevent injury or death. Most of these standards were not supported by experimental data or testing. In this review, we describe existing standards, discuss their origins, and we comprehensively compare their prescriptions across standards. Surprisingly, we found that most safety standards had little or no scientific basis, and prescriptions across standards often varied by at least an order of magnitude. Many published standards traced back to a US Navy 500 psi guideline, which was intended to provide a peak pressure at which injuries were likely to occur. This standard itself seems to have been based upon a completely unfounded assertion that has propagated throughout the literature in subsequent years. Based on the limitations of the standards discussed, we outline future directions for underwater blast injury research, such as the compilation of epidemiological data to examine actual injury risk by human beings subjected to underwater blasts.