Fractal analysis of visual search activity for mass detection during mammographic screening.

PURPOSE: The objective of this study was to assess the complexity of human visual search activity during mammographic screening using fractal analysis and to investigate its relationship with case and reader characteristics. METHODS: The study was performed for the task of mammographic screening with simultaneous viewing of four coordinated breast views as typically done in clinical practice. Eye-tracking data and diagnostic decisions collected for 100 mammographic cases (25 normal, 25 benign, 50 malignant) from 10 readers (three board certified radiologists and seven Radiology residents), formed the corpus for this study. The fractal dimension of the readers' visual scanning pattern was computed with the Minkowski-Bouligand box-counting method and used as a measure of gaze complexity. Individual factor and group-based interaction ANOVA analysis was performed to study the association between fractal dimension, case pathology, breast density, and reader experience level. The consistency of the observed trends depending on gaze data representation was also examined. RESULTS: Case pathology, breast density, reader experience level, and individual reader differences are all independent predictors of the complexity of visual scanning pattern when screening for breast cancer. No higher order effects were found to be significant. CONCLUSIONS: Fractal characterization of visual search behavior during mammographic screening is dependent on case properties and image reader characteristics.

Simulation-based educational curriculum for fluoroscopically guided lumbar puncture improves operator confidence and reduces patient dose.

RATIONALE AND OBJECTIVES: Fluoroscopically guided lumbar puncture (FGLP) is a commonly performed procedure with increased success rates relative to bedside technique. However, FGLP also exposes both patient and staff to ionizing radiation. The purpose of this study was to determine if the use of a simulation-based FGLP training program using an original, inexpensive lumbar spine phantom could improve operator confidence and efficiency, while also reducing patient dose. MATERIALS AND METHODS: A didactic and simulation-based FGLP curriculum was designed, including a 1-hour lecture and hands-on training with a lumbar spine phantom prototype developed at our institution. Six incoming post-graduate year 2 (PGY-2) radiology residents completed a short survey before taking the course, and each resident practiced 20 simulated FGLPs using the phantom before their first clinical procedure. Data from the 114 lumbar punctures (LPs) performed by the six trained residents (prospective cohort) were compared to data from 514 LPs performed by 17 residents who did not receive simulation-based training (retrospective cohort). Fluoroscopy time (FT), FGLP success rate, and indication were compared. RESULTS: There was a statistically significant reduction in average FT for the 114 procedures performed by the prospective study cohort compared to the 514 procedures performed by the retrospective cohort. This held true for all procedures in aggregate, LPs for myelography, and all procedures performed for a diagnostic indication. Aggregate FT for the prospective group (0.87 ± 0.68 minutes) was significantly lower compared to the retrospective group (1.09 ± 0.65 minutes) and resulted in a 25% reduction in average FT (P = .002). There was no statistically significant difference in the number of failed FGLPs between the two groups. CONCLUSIONS: Our simulation-based FGLP curriculum resulted in improved operator confidence and reduced FT. These changes suggest that resident procedure efficiency was improved, whereas patient dose was reduced. The FGLP training program was implemented by radiology residents and required a minimal investment of time and resources. The LP spine phantom used during training was inexpensive, durable, and effective. In addition, the phantom is compatible with multiple modalities including fluoroscopy, computed tomography, and ultrasound and could be easily adapted to other applications such as facet injections or joint arthrograms.

Predicting diagnostic error in radiology via eye-tracking and image analytics: preliminary investigation in mammography.

PURPOSE: The primary aim of the present study was to test the feasibility of predicting diagnostic errors in mammography by merging radiologists' gaze behavior and image characteristics. A secondary aim was to investigate group-based and personalized predictive models for radiologists of variable experience levels. METHODS: The study was performed for the clinical task of assessing the likelihood of malignancy of mammographic masses. Eye-tracking data and diagnostic decisions for 40 cases were acquired from four Radiology residents and two breast imaging experts as part of an IRB-approved pilot study. Gaze behavior features were extracted from the eye-tracking data. Computer-generated and BIRADS images features were extracted from the images. Finally, machine learning algorithms were used to merge gaze and image features for predicting human error. Feature selection was thoroughly explored to determine the relative contribution of the various features. Group-based and personalized user modeling was also investigated. RESULTS: Machine learning can be used to predict diagnostic error by merging gaze behavior characteristics from the radiologist and textural characteristics from the image under review. Leveraging data collected from multiple readers produced a reasonable group model [area under the ROC curve (AUC) = 0.792 ± 0.030]. Personalized user modeling was far more accurate for the more experienced readers (AUC = 0.837 ± 0.029) than for the less experienced ones (AUC = 0.667 ± 0.099). The best performing group-based and personalized predictive models involved combinations of both gaze and image features. CONCLUSIONS: Diagnostic errors in mammography can be predicted to a good extent by leveraging the radiologists' gaze behavior and image content.

The effect of multiple wire localization in breast conservation.

Variability exists regarding the surgical technique in breast conservation therapy. The purpose of this project was to determine differences between single (SH) or flanking (FH) hooked needle localization wires used for nonpalpable breast lesions. We retrospectively reviewed 201 female patients at a single institution from 2004 to 2008. All patients had biopsy-proven ductal carcinoma in situ or invasive disease. Comparisons were made in regard to margin status, reoperation, completion mastectomy, size of lesion, and breast specimen volume. SH was placed in 122 patients (61%) and FH in 79 patients (39%). In SH, 23 patients (18%) had positive margins and 31 patients (25%) had reoperations as compared with 31 patients (25%) with positive margin and 36 patients (44%) in the FH cohort (P = 0.039 and 0.0037). Average lesion size and volume resected was 1.5 cm and 137 cm(3) in SH and 2.85 cm and 188 cm(3) in FH, respectively (P = 0.0001 and 0.006). Positive margins were associated with lesion size and not volume of tissue excised. The FH technique was associated with more positive margins, reoperation, and completion mastectomy.