Military saves thousands monthly in transcriptionAfter spending thousands of dollars on custom-made toilet seats and wrenches, the military may finally have gotten it right with a speech recognition system for radiology.An
Military saves thousands monthly in transcription
After spending thousands of dollars on custom-made toilet seats and wrenches, the military may finally have gotten it right with a speech recognition system for radiology.
An educational exhibit at the RSNA infoRAD in November demonstrated a Department of Defense-built speech-recognition dictation system that could save millions of dollars annually. Custom software developed by the military is linked to Dragon Naturally Speaking Professional (version 5) as part of a "home-grown" system developed at Wright-Patterson Air Force Base in Dayton, OH.
A shortage of transcriptionists and report turnaround times of up to 14 days at Wright-Patterson led to development of the system, which is tailored for integration with the RIS and PACS operating at the base. The initial configuration required "cutting and pasting" dictation, a cumbersome and time-consuming process for radiologists that actually slowed overall productivity. But military radiologist Dr. Cliff Sweet came up with a solution. Sweet developed an integration link now known as the Voice Patch.
The financial savings have been substantial. When the system was initiated in December 1998, speech recognition was used for about 600 of 4200 reports, and the facility realized savings of about $1000 per month. In March 2001, when approximately 4250 exams of 5200 were dictated using speech recognition, savings totaled $10,000.
To achieve this, developers had to work within the space constraints imposed by Wright-Patterson facilities. The initial four-monitor PACS diagnostic workstations took up substantial desk space. So the radiology department adopted a 12-inch flat-screen monitor to display text. A Linksys KVM switch enables a single keyboard to be used for both the PACS and the speech recognition system. These two enhancements eliminated neck strain caused by use of an awkwardly placed full-sized monitor and also resolved the problem of insufficient room for two keyboards in a workstation area.
Six of the eight radiologists at Wright-Patterson use the system extensively, said William C. Gibson, a radiologic technologist at the base. Reports dictated with the speech recognition system are turned around in less than a day, compared with an average four-day turnaround time for reports dictated and transcribed in the traditional way.
Productivity gains, and the prospect of saving thousands of dollars monthly, have attracted a lot of attention within the DoD. Several hundred military medical facilities have now adopted Wright-Patterson's speech recognition system, according to Gibson.
MRI-Based AI Radiomics Model Offers 'Robust' Prediction of Perineural Invasion in Prostate Cancer
July 26th 2024A model that combines MRI-based deep learning radiomics and clinical factors demonstrated an 84.8 percent ROC AUC and a 92.6 percent precision-recall AUC for predicting perineural invasion in prostate cancer cases.
Breast MRI Study Examines Common Factors with False Negatives and False Positives
July 24th 2024The absence of ipsilateral breast hypervascularity is three times more likely to be associated with false-negative findings on breast MRI and non-mass enhancement lesions have a 4.5-fold likelihood of being linked to false-positive results, according to new research.
Can Polyenergetic Reconstruction Help Resolve Streak Artifacts in Photon Counting CT?
July 22nd 2024New research looking at photon-counting computed tomography (PCCT) demonstrated significantly reduced variation and tracheal air density attenuation with polyenergetic reconstruction in contrast to monoenergetic reconstruction on chest CT.