Grossbard BP, Loughin CA, Marino DJ, et al.
OBJECTIVE: To: (1) determine the success of medical infrared imaging (MII) in identifying dogs with TLIVDD, (2) compare MII localization with magnetic resonance imaging (MRI) results and surgical findings, and (3) determine if the MII pattern returns to that of normal dogs 10 weeks after decompression surgery. STUDY DESIGN: Prospective case series. ANIMALS: Chondrodystrophic dogs (n = 58) with Type I TLIVDD and 14 chondrodystrophic dogs with no evidence of TLIVDD. METHODS: Complete neurologic examination, MII, and MRI studies were performed on all dogs. Dogs with type I TLIVDD had decompressive surgery and follow-up MII was performed at 10 weeks. Pattern analysis software was used to differentiate between clinical and control dogs, and statistical analysis using anatomic regions of interest on the dorsal views were used to determine lesion location. Recheck MII results were compared with control and pre-surgical images. RESULTS: Computer recognition pattern analysis was 90% successful in differentiating normal dogs from dogs affected by TLIVDD and 97% successful in identifying the abnormal intervertebral disc space in dogs with TLIVDD. Statistical comparisons of the ROI mean temperature were unable to determine the location of the disc herniation. Recheck MII patterns did not normalize and more closely resembled the clinical group. CONCLUSIONS: MII was 90% successful differentiating between normal dogs and 97% successful in identifying the abnormal intervertebral disc space in dogs with TLIVDD. Abnormal intervertebral disc space localization using ROI mean temperature analysis was not successful. MII patterns 10 weeks after surgery do not normalize.