Wiard Jorritsma

Adaptive support for user interface customization

ObjectivesThis study aimed to evaluate the usefulness of adaptive customization support in a natural work environment: the Picture Archiving and Communication System (PACS) in radiology. MethodsAdaptive support was given in the form of customization suggestions, generated based on behavioral user data, which participants could choose to accept or ignore. Twelve radiologists worked with the standard adaptable version of the PACS for six weeks, during which their actions on the PACS interface were logged. Based on these logging data, customization support was generated for each specific participant. Half of the participants received support and half did not. After the support was given, logging continued for another six weeks. Participants? customization behavior and performance, measured as the average time needed to review a radiography study, were compared between the groups. Subjective responses to the customization support were also measured. ResultsParticipants who received support used the PACS?s customization facilities more effectively than participants who did not receive support. Participants receiving support accepted most of the customization suggestions and all participants indicated that the support was useful. We did not find an increase in efficiency due to the support, possibly because the performance measure we used was not sensitive enough. Subjective responses did show that participants perceived the support as increasing their efficiency. ConclusionAdaptive customization support would be a useful addition to the standard adaptable PACS interface, because it allows radiologists to customize their interface more effectively. Radiologists do not always customize their user interface effectively.They sometimes do not customize at all.Adaptive customization support improves radiologists? customization effectiveness.Radiologists have a positive attitude towards adaptive customization support.

Pattern mining of user interaction logs for a post-deployment usability evaluation of a radiology PACS client

OBJECTIVES To perform a post-deployment usability evaluation of a radiology Picture Archiving and Communication System (PACS) client based on pattern mining of user interaction log data, and to assess the usefulness of this approach compared to a field study. METHODS All user actions performed on the PACS client were logged for four months. A data mining technique called closed sequential pattern mining was used to automatically extract frequently occurring interaction patterns from the log data. These patterns were used to identify usability issues with the PACS. The results of this evaluation were compared to the results of a field study based usability evaluation of the same PACS client. RESULTS The interaction patterns revealed four usability issues: (1) the display protocols do not function properly, (2) the line measurement tool stays active until another tool is selected, rather than being deactivated after one use, (3) the PACSs built-in 3D functionality does not allow users to effectively perform certain 3D-related tasks, (4) users underuse the PACSs customization possibilities. All usability issues identified based on the log data were also found in the field study, which identified 48 issues in total. CONCLUSIONS Post-deployment usability evaluation based on pattern mining of user interaction log data provides useful insights into the way users interact with the radiology PACS client. However, it reveals few usability issues compared to a field study and should therefore not be used as the sole method of usability evaluation.

Post-deployment usability evaluation of a radiology workstation.

OBJECTIVES To determine the number, nature and severity of usability issues radiologists encounter while using a commercially available radiology workstation in clinical practice, and to assess how well the results of a pre-deployment usability evaluation of this workstation generalize to clinical practice. METHODS The usability evaluation consisted of semi-structured interviews and observations of twelve users using the workstation during their daily work. Usability issues and positive usability findings were documented. Each issue was given a severity rating and its root cause was determined. Results were compared to the results of a pre-deployment usability evaluation of the same workstation. RESULTS Ninety-two usability issues were identified, ranging from issues that cause minor frustration or delay, to issues that cause significant delays, prevent users from completing tasks, or even pose a potential threat to patient safety. The results of the pre-deployment usability evaluation had limited generalizability to clinical practice. CONCLUSIONS This study showed that radiologists encountered a large number and a wide variety of usability issues when using a commercially available radiology workstation in clinical practice. This underlines the need for effective usability engineering in radiology. Given the limitations of pre-deployment usability evaluation in radiology, which were confirmed by our finding that the results of a pre-deployment usability evaluation of this workstation had limited generalizability to clinical practice, it is vital that radiology workstation vendors devote significant resources to usability engineering efforts before deployment of their workstation, and to continue these efforts after the workstation is deployed in a hospital.

Comparing Four Touch-Based Interaction Techniques for an Image-Based Audience Response System

This study aimed to determine the most appropriate touch-based interaction technique for I2Vote, an image-based audience response system for radiology education in which users need to accurately mark a target on a medical image. Four plausible techniques were identified: land-on, take-off, zoom-pointing, and shift. The techniques were implemented in such a way that they could be used on any modern device. An empirical study was performed in which users marked a target on an image using all four techniques on either a smartphone or a tablet. The techniques were compared in terms of accuracy, efficiency, ease of use, intuitiveness, and compatibility with the different devices. The results showed that shift was the most accurate technique, but it was hampered by its high complexity and low intuitiveness. Land-on was the fastest technique but also the least accurate. Take-off and zoom-pointing provided the best trade-off between accuracy, efficiency, ease of use, and intuitiveness. We therefore conclude that both take-off and zoom-pointing are viable interaction techniques for I2Vote.

Medical Imaging Informatics in Nuclear Medicine

Medical imaging informatics is gaining importance in medicine both in clinical practice and in scientific research. Besides radiology, nuclear medicine is also a major stakeholder in medical imaging informatics because of the variety of available imaging modalities and the imaging-oriented operation of this specialization. The amount of data produced by imaging modalities and other diagnostic devices per exam is growing steadily together with the number of exams performed, resulting in a data explosion that provides major opportunities for data analytics and data mining in big data projects. However, it can also fuse problems in storage, distribution, review, and interpretation of the enormous amounts of clinical (image) data while maintaining high-quality patient care. To tackle these problems, standardization, structuring, and automation are crucial. As one of the major producers of medical imaging data, nuclear medicine should stay at the forefront of these developments.