Speech processing in radiology

Hundt, Walter; Stark, O.; Scharnberg, B.; Hold, M; Kohz, P.; Lienemann, Andreas; Bonél, Harald; Reiser, Maximilian

doi:10.1007/s003300050869

Cited by 19 publications

(17 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The key question is whether the inherent error rate of a speech recognition system is acceptable to the individual user, and whether significantly decreased radiologist productivity is offset adequately by the realized institutional benefits. 5 Hopefully, communication of our experience will allow radiologists to make an informed decision on whether a speech recognition system would be suitable for their practices. Knowing the problems with our system should facilitate more careful scrutiny of the specifications and functionality of prospective systems.…”

Section: Resultsmentioning

confidence: 99%

Experience with implementation of a radiology speech recognition system

Houston

Rupp

2000

J Digit Imaging

View full text Add to dashboard Cite

Recent advances in speech recognition technology have allowed development of computer systems for real-time radiologist-driven generation of reports. The transition to a speech recognition system is a technically complex process with many potential piffalls that can decrease efficiency and disrupt workflow. In our recent experience with installation of such a system in an academic radiology department, factors that have worked against optimal performance have included environmental Iogistics, hardware incompatibilities, radiology information system interface problems, lack of suitable training, and inadequate technical support. Communication of our experience is intended to allow radiologists to anticipate complications of these systems and make informed decisions regarding the feasibility of such a system in their practices. With this information, potential buyers should be able to carefully scrutinize specifications for prospective systems and, by avoiding many of the possible pitfalls, make an easier transition to a speech recognition environment. ENVIRONMENTAL LOGISTICSThere is a wide range in performance of a speech recognition system, which is indirectly proportional to the amount of noise in the local environment. We have had the greatest success in small reading rooms in which there is a single system in use anda lack of extraneous personnel and ambient

show abstract

Section: Resultsmentioning

confidence: 99%

Experience with implementation of a radiology speech recognition system

Houston

Rupp

2000

J Digit Imaging

View full text Add to dashboard Cite

show abstract

“…The time required for dictation using a voice-recognition system is longer than that required using a tape-based transcription system, but the lack of intermediary involvement by a transcriptionist and the capability of immediate final approval of the report by the staff radiologist makes the total turnaround time shorter with a voice-recognition system (7,22,23). If the workflow is designed such that the primary reader of the film images is immediately signing off on his or her interpretations rather than "batching" them to be signed later, TOVs and, hence, report turnaround times can be reduced by an even greater margin.…”

Section: Discussionmentioning

confidence: 99%

“…In addition to improving communication between physicians by producing an official report more rapidly, a reduction in turnaround times can also produce increased revenue because of the fewer days during which charges are in accounts receivable (1)(2)(3)(4)(5)(6)(7)(8). Report turnaround times are also one of the standards regulated by the American College of Radiology (Standard for Communication-Diagnostic Radiology), accreditation organizations, utilization review panels, and some managed care organizations.…”

mentioning

confidence: 99%

Impact of Pager Notification on Report Verification Times

Oğuz¹,

Yousem²,

Deluca³

et al. 2002

Academic Radiology

View full text Add to dashboard Cite

verification times of each physician were collected for 3 months in the years before and after initiation of the PNS. Radiologists enrolled in the PNS and those who were not were assigned into two groups for analysis. Mean TOVs for the two sets of 3 months and for the two groups were calculated and differences recorded. Two-tailed t tests were used to assess for statistical differences between the groups.Results. Twenty-nine of 37 radiologists voluntarily enrolled in the PNS (group 1). Mean TOV was 26.75 hours (standard deviation [SD] ϭ 17.76) for these physicians before and 14.48 hours (SD ϭ 11.86) after the PNS was employed (P Ͻ .01). For those physicians who did not enroll in the PNS, mean TOV was 11.53 hours (SD ϭ 5.55) before and 9.77 hours (SD ϭ 9.86) after the PNS was employed (P ϭ .33). Both the absolute and percentage reductions in TOVs were significantly greater for those physicians enrolled in the PNS than for those who were not (P ϭ .035). Twenty-three of 29 (79%) physicians who used the PNS showed a reduction in their report turnaround times. Conclusion.Linking the PNS with the radiology information system to notify physicians of unsigned reports was effective in reducing report verification times.Key Words. Clinical operations, radiology reporting, economics. © AUR, 2002Improving turnaround times for radiology reports has been one of the major concerns in a practice. In addition to improving communication between physicians by producing an official report more rapidly, a reduction in turnaround times can also produce increased revenue because of the fewer days during which charges are in accounts receivable (1-8). Report turnaround times are also one of the standards regulated by the American College of Radiology (Standard for Communication-Diagnostic Radiology), accreditation organizations, utilization review panels, and some managed care organizations. From a medicolegal standpoint, a single untoward event resulting from a delay in transcription and report verification could lead to a monetary settlement that far outweighs the cost of extra transcriptionists or information technology solutions.As some departments have evolved from paper production of radiology reports to a complete electronic patient record, use of electronic signatures in radiology reports has become common. Previously, transcribed paper reports could be placed at the radiologist's location for immediate signature after a transcriptionist concluded a parcel of work. Whether those reports actually were signed immediately, were edited and retyped, or were left for Acad Radiol 2002; 9:954 -959

show abstract

“…DR. WEISS: The use of SR in radiology reporting clearly has the potential for decreasing radiologists' productivity by increasing dictation time. A number of studies have documented a time penalty in the use of SR. [1][2][3] Some studies tested an earlier version of the speech engine used for conversion of spoken word to text. Most studies so far have been performed at academic centers with a heterogeneous population of users.…”

mentioning

confidence: 99%