The Harry M. Zweig Memorial Fund for Equine Research

Spectro-temporal Signature for Identification of
Upper Airway Abnormalities in Exercising Horses

Dr. Normand G. Ducharme

Upper airway problems have been recognized in horses since at least 1866. These disorders affect both racehorses and horses used for other purposes. In racehorses, the complaint is ordinarily that the horse's athletic performance is diminished. Abnormal respiratory noise is the more common problem in show horses. The two most common causes of upper airway obstruction and noise are laryngeal hemiplegia (roaring) and dorsal displacement of the soft palate (DDSP). Laryngeal hemiplegia is a common cause of diminished athletic performance and abnormal respiratory noise. Laryngeal hemiplegia affects 5-8% of racehorses. When laryngeal hemiplegia is complete, this disorder is readily diagnosed by resting endoscopy. Partial paralysis is more difficult to diagnose and often requires treadmill endoscopy. The percentage of horses affected by palate displacement is difficult to pinpoint precisely since most horses with this disorder are affected only during strenuous exercise and, hence, appear normal on resting endoscopy. Palate displacement typically leads to markedly impaired performance and abnormal respiratory noise (choking down) in racehorses. In one such study, 53% of horses presented to a sport medicine clinic with impaired performance were found, by scoping during treadmill exercise, to have an upper airway obstruction with the majority of these being palate displacement. Furthermore, if one uses the frequency of tongue-ties or the performance of the Llewellyn procedure as a measure, then the prevalence of palate problems is considerably higher.

With the advent of scoping during high-speed treadmill exercise, the identification of throat problems can now be made with confidence. In addition, treadmill endoscopy has led to the discovery of new causes of airway obstruction associated with tissue collapsing under pressure, so they're only seen during exercise. Examples of such obstructions that are evident only during high-speed work are displaced epiglottis, vocal cords and aryepiglottic folds.

As early as 1866, diagnostic methods for upper airway obstructive disorders consisted of subjective criteria such as type and characteristics of respiratory noise as well as the "grunt test." In the 1980's, objective measurements of upper airway pressure and airflow were introduced for assessment of the upper airway in horses. In 1991, we introduced a laryngeal grading system, which was based on correlation of endoscopies of the upper airway at rest and at exercise on a high-speed treadmill. Treadmill endoscopy is now considered the gold standard for diagnoses of upper airway obstruction. However, high-speed equine treadmills are available only at university hospitals or specialized private clinics. Development of a field test to screen horses for airway obstruction at the track or on the farm would be a major breakthrough in diagnosing these problems.

To this day, there are no objective criteria for evaluating upper airway noise in horses. This is important since noise usually accompanies obstructive problems leading to decreased performance in racehorses and may be the only clinical sign of concern to a show horse owner. Additionally, elimination of noise has been used as an indicator of treatment success for various airway disorders. Assessment of noise has been entirely subjective. We are no further ahead than the classical 1936 presentation of Sir Hobday where in his experience with 4,000 roarers, he felt he could render 85% of hunters useful, and "20% sufficiently sound to pass the average hunting man without comment." This is still the current standard for noise!

Joint time-frequency analysis (JTFA) is a significant mathematical advance in sound. This technology is now well established and is used in human speech analysis. Last year we adapted this technique to equine upper airway sounds recorded during exercise. We have developed a method to precisely record equine upper airway sounds; and there is a clear difference between normal horses, complete roarers, and horses with paralyzed nostrils. The upper airway sounds in these situations have a clear "signature" which enables recognition of each status. In the second half of the current funded project, we will adapt this technique for field use.

We are approaching our goal of developing a technique to enable equine practitioners to record upper airway sounds during exercise using a microphone placed in a horse's nostril and connected for sound transmission to a FM tape recorder. Following this field test, the FM tape/recording, not the horse, would be shipped for further analysis to obtain a final diagnosis. So far we have data indicating that roaring can readily be recognized from normal. However, for this technique to be widely helpful to equine practitioners, we need to map the "signature" of the various levels of severity of all airway diseases causing airway noises. We might even be able to recognize sounds created by lower airway disease associated with movement of mucous or blood in the trachea during exercise. Therefore, we propose the following three objectives:

1) Continue refining the mapping technique for sounds analysis using special speech analysis software.
2) Map the airway sounds of horses examined at our Equine Performance Testing Clinic so various naturally occurring abnormalities and their severity can be studied.
3) Investigate feasibility of using a telemetry basis unit in the field to record airway sounds for analysis.