2020 |
■ Ziemer, Tim, Nuchprayoon, Nuttawut & Schultheis, Holger, “Psychoacoustic Sonification as User Interface for Human-Machine Interaction” in: International Journal of Informatics Society 12(1), 2020.
Abstract
Hide Abstract, Full paper
Abstract
========
When operating a machine, the operator needs to know some spatial relations, like the relative location of the target or the nearest obstacle. Often, sensors are used to derive this spatial information, and visual displays are deployed as interfaces to communicate this information to the operator. In this paper, we present psychoacoustic sonification as an alternative interface for human-machine interaction. Instead of visualizations, an interactive sound guides the operator to the desired target location, or helps her avoid obstacles in space. By considering psychoacoustics --- i.e., the relationship between the physical and the perceptual attributes of sound --- in the audio signal processing, we can communicate precisely and unambiguously interpretable direction and distance cues along three orthogonal axes to a user. We present exemplary use cases from various application areas where users can benefit from psychoacoustic sonification.
|
■
Yin, M. S., Haddawy, P., Suebnukarn, S., Kulapichitr, F., Rhienmora, P., Jatuwat, V., & Uthaipattanacheep, N. (2020). Formative Feedback Generation in a VR-based Dental Surgical Skill Training Simulator. Journal of Biomedical Informatics, 103659.
Abstract
Hide Abstract, Full paper
Abstract
========
Fine motor skill is indispensable for a dentist. As in many other medical fields of study, the traditional surgical master-apprentice model is widely adopted in dental education. Recently, virtual reality (VR) simulators have been employed as supplementary components to the traditional skill-training curriculum, and numerous dental VR systems have been developed academically and commercially. However, the full promise of such systems has yet to be realized due to the lack of sufficient support for formative feedback. Without such a mechanism, evaluation still demands dedicated time of experts in scarce supply. To fill the gap of formative assessment using VR simulators in skill training in dentistry, we present a framework to objectively assess the surgical skill and generate formative feedback automatically. VR simulators enable collecting detailed data on relevant metrics throughout a procedure. Our approach to formative feedback is to correlate procedure metrics with the procedure outcome to identify the portions of a procedure that need to be improved. Specifically, for the errors in the outcome, the responsible portions of the procedure are identified by using the location of the error. Tutoring formative feedback is provided using the video modality. The effectiveness of the feedback system is evaluated with dental students using randomized controlled trials. The findings show the feedback mechanisms to be effective and to have the potential to be used as valuable supplemental training resources.
|
■
T. Siriapisitha, W. Kusakunnirana, P. Haddawy, Pyramid Graph Cut: Integrating Intensity and Gradient Information for Grayscale Medical Image Segmentation. Computers in Biology and Medicine, 126, November 2020.
Abstract
Hide Abstract, Full paper
Abstract
========
Segmentation of grayscale medical images is challenging because of the similarity of pixel intensities and poor gradient strength between adjacent regions. The existing image segmentation approaches based on either intensity or gradient information alone often fail to produce accurate segmentation results. Previous approaches in the literature have approached the problem by embedded or sequential integration of different information types to improve the performance of the image segmentation on specific tasks. However, an effective combination or integration of such information is difficult to implement and not sufficiently generic for closely related tasks. Integration of the two information sources in a single graph structure is a potentially more effective way to solve the problem. In this paper we introduce a novel technique for grayscale medical image segmentation called pyramid graph cut, which combines intensity and gradient sources of information in a pyramid-shaped graph structure using a single source node and multiple sink nodes. The source node, which is the top of the pyramid graph, embeds intensity information into its linked edges. The sink nodes, which are the base of the pyramid graph, embed gradient information into their linked edges. The min-cut uses intensity information and gradient information, depending on which one is more useful or has a higher influence in each cutting location of each iteration. The experimental results demonstrate the effectiveness of the proposed method over intensity-based segmentation alone (i.e. Gaussian mixture model) and gradient-based segmentation alone (i.e. distance regularized level set evolution) on grayscale medical image datasets, including the public 3DIRCADb-01 dataset. The proposed method archives excellent segmentation results on the sample CT of abdominal aortic aneurysm, MRI of liver tumor and US of liver tumor, with dice scores of 90.49±5.23%, 88.86±11.77%, 90.68±2.45%, respectively.
|
■ Prachyabrued, M., Haddawy, P., Tengputtipong, K., Yin, M. S., Bicout, D., & Laosiritaworn, Y. (2020, December). Immersive Visualization of Dengue Vector Breeding Sites Extracted from Street View Images. In 2020 IEEE International Conference on Artificial Intelligence and Virtual Reality (AIVR) (pp. 37-42). IEEE.
Abstract
Hide Abstract, Full paper
Abstract
========
Dengue is considered one of the most serious global health burdens. The primary vector of dengue is the Aedes aegypti mosquito, which has adapted to human habitats and breeds primarily in artificial containers that can contain water. Control of dengue relies on effective mosquito vector control, for which detection and mapping of potential breeding sites is essential. The two traditional approaches to this have been to use satellite images, which do not provide sufficient resolution to detect a large proportion of the breeding sites, and manual counting, which is too labor intensive to be used on a routine basis over large areas. Our recent work has addressed this problem by applying convolutional neural nets to detect outdoor containers representing potential breeding sites in Google street view images. The challenge is now not a paucity of data, but rather transforming the large volumes of data produced into meaningful information. In this paper, we present the design of an immersive visualization using a tiled-display wall that supports an early but crucial stage of dengue investigation, by enabling researchers to interactively explore and discover patterns in the datasets, which can help in forming hypotheses that can drive quantitative analyses. The tool is also useful in uncovering patterns that may be too sparse to be discovered by correlational analyses and in identifying outliers that may justify further study. We demonstrate the usefulness of our approach with two usage scenarios that lead to insights into the relationship between dengue incidence and container counts.
|
■
Suppawong Tuarob, Sung Woo Kang, Poom Wettayakorn, Chanatip Pornprasit, Tanakitti Sachati, Saeed-Ul Hassan, Peter Haddawy,
Automatic Classification of Algorithm Citation Functions in Scientific Literature, IEEE Transactions on Knowledge and Data Engineering, 32(10), pp 1881 – 1896, October 2020.
Abstract
Hide Abstract, Full paper
Abstract
========
Computer sciences and related disciplines evolve around developing, evaluating, and applying algorithms. Typically, an algorithm is not developed from scratch, but uses and builds upon existing ones, which often are proposed and published in scholarly articles. The ability to capture this evolution relationship among these algorithms in scientific literature would not only allow us to understand how a particular algorithm is composed, but also shed light on large-scale analysis of algorithmic evolution through different temporal spans and thematic scales. We propose to capture such evolution relationship between two algorithms by investigating the knowledge represented in citation contexts, where authors explain how cited algorithms are used in their works. A set of heterogeneous ensemble machine-learning methods is proposed, where the combination of two base classifiers trained with heterogeneous feature types is used to automatically identify the algorithm usage relationship. The proposed heterogeneous ensemble methods achieve the best average F1 of 0.749 and 0.905 for fine-grained and binary algorithm citation function classification, respectively. The success of this study will allow us to generate a large-scale algorithm citation network from a collection of scholarly documents representing multiple time spans, venues, and fields of study. Such a network will be used as an instrument not only to answer critical questions in algorithm search, such as identifying the most influential and generalizable algorithms, but also to study the evolution of algorithmic development and trends over time.
|
■ Myat Su Yin, Mihai Pomarlan, Peter Haddawy, Muhammad Rauf Tabassam, Chitpol Chaimanakarn, Natchalee Srimaneekarn, Saeed-Ul Hassan, Automated Extraction of Causal Relations from Text for Teaching Surgical Concepts, Proc. 8th IEEE Int’l Conf. on Healthcare Informatics, Oldenburg, 30 Nov. – 3 Dec., 2020.
Abstract
Hide Abstract
Abstract
========
Effective teaching of surgical decision making requires providing students with a deep understanding of the domain so that they have the ability to make decisions in novel situations. This means providing them with a thorough understanding of causal relations between actions and their possible effects in the context of various states of the patient as well as previous actions. Intelligent tutoring systems to teach surgical decision making thus require such domain knowledge, but there are currently no medical ontologies that encompass it. While it is possible to engineer the needed ontologies by hand, this
requires a large effort for every new domain to be covered. In this paper we explore the possibility of automatically extracting causal relations from textbooks on surgery. Specifically, we adapt the
spaCy NLP tool for this task and apply it to a collection of fifteen textbooks on endodontic root canal treatment, which is one of the most challenging areas of dental surgery. Since the main purpose
is to extract knowledge for teaching, we focus on actions that can lead to surgical mishaps. We evaluate the precision and recall of the extracted relations using a gold standard prepared by a pair of dental surgeons.
|
■ Myat Su Yin, Peter Haddawy, Benedikt Hosp, Paphon Sa-ngasoongsong, Ratthapoom Watcharopas, Thanwarat Tanprathumwong, Madereen Sayo, Supawit Yangyuenpradorn, Akara Supratak, Enkelejda Kasneci, A Study of Expert/Novice Perception in Arthroscopic Shoulder Surgery, Proc. 4th Int’l Conf. on Medical and Health Informatics (ICMHI 2020), Kamakura, Japan, August 14-16, 2020.
Abstract
Hide Abstract, Full paper
Abstract
========
Arthroscopic shoulder surgery is an advanced orthopedic surgical procedure, which is particularly challenging due to the complex anatomy of the shoulder, and tight spaces for navigation, which also limits the view from the arthroscope. In carrying out arthroscopy, the ability to quickly and effectively navigate through the joint to reach a desired location is essential. Novices often experience confusion in trying to triangulate the information from arthroscopy output with the background knowledge of anatomy while orienting and navigating the instruments. In this paper, we report on the results of the first cadaveric eye-tracking study of arthroscopic surgery in which we investigate differences in perception between experts and novices. Novices' perception is analyzed with cognitive load analysis throughout the procedure and specifically, during the portions of the procedure in which subjects are observed to be confused. In investigating such portions, the gaze data analysis is supplemented with head rotations and acceleration information from gyroscope and accelerometer sensors from the eye tracker. We also use the gathered eye tracking metrics to construct a model to classify subjects into expert/novice. We find statistically significant relations between head movement as well as pupil diameter and periods of confusion. We identify a subset of the metrics that we use to build a simple classifier that is able to distinguish between novices and experts with accuracy of 84%.
|
■ Chaitawat Sa-ngamuang, Thomas Barkowsky, Patiwat Sa-angchai, Peter Haddawy, Saranath Lawpoolsri, A study of individual human mobility patterns related to malaria transmission along the Thai-Myanmar border, Proc. 4th Int’l Conf. on Medical and Health Informatics (ICMHI 2020), Kamakura, Japan, August 14-16, 2020.
Abstract
Hide Abstract, Full paper
Abstract
========
Malaria elimination remains a major challenge worldwide largely because human mobility can result in importing cases from areas of high incidence to areas of low incidence. Thus, understanding the role of human mobility in malaria transmission is essential. In this study, we collect mobility data from 88 participants over ten months using a smartphone application. Our study area is in northern Thailand along the border with Myanmar, from which malaria may be imported. We analyze amount of time spent in Thailand/Myanmar in areas of various land cover types, spatial distribution of movement, and network patterns of movement. We find significant differences between villages in amounts of time spent in forest areas and in Myanmar, with most travel to Myanmar occurring from two villages. We find significantly higher spatial distribution of movement in the dry season than the wet season. Our results provide important insight to help target surveillance and intervention.
|
■ Vajsbaher, Tina, Ziemer, Tim & Schultheis, Holger, “A multi-modal approach to cognitive training and assistance in minimally invasive surgery”, Cognitive Systems Research 64, pp. 57–72, https://doi.org/10.1016/j.cogsys.2020.07.005, 2020.
Abstract
Hide Abstract, Full paper
Abstract
========
Minimally-invasive surgery (MIS) offers many benefits to patients, but is considerably more difficult to learn and perform than is open surgery. One main reason for the observed difficulty is attributable to the visuo-spatial challenges that arise in MIS, taxing the surgeons’ cognitive skills. In this contribution, we present a new approach that combines training and assistance as well as the visual and the auditory modality to help surgeons to overcome these challenges. To achieve this, our approach assumes two main components: An adaptive, individualized training component as well as a component that conveys spatial information through sound. The training component (a) specifically targets the visuo-spatial processes crucial for successful MIS performance and (b) trains surgeons in the use of the sound component. The second component is an auditory display based on a psychoacoustic sonification, which reduces and avoids some of the commonly experienced MIS challenges. Implementations of both components are described and their integration is discussed. Our approach and both of its components go beyond the current state of the art in important ways. The training component has been explicitly designed to target MIS-specific visuo-spatial skills and to allow for adaptive testing, promoting individualized learning. The auditory display is conveying spatial information in 3-D space. Our approach is the first that encompasses both training for improved mastery and reduction of cognitive challenges in MIS. This promises better tailoring of surgical skills and assistance to the needs and the capabilities of the surgeons and, thus, ultimately, increased patient safety and health.
|
■ Tina Vajsbaher, Holger Schultheis, Paphon Sa-ngasoongsong, Ratthapoom Watcharopas, Myat Su Yin, Peter Haddawy, The Role of Spatial Cognition in Surgical Navigation in Arthroscopic Surgery, Spatial Cognition X: 12th Int’l Conf., Aug. 2020.
|
2019 |
■ Haddawy P, Wettayakorn P, Nonthaleerak B, Su Yin M, Wiratsudakul A, Schöning J, et al. Large scale detailed mapping of dengue vector breeding sites using street view images. PLoS Neglected Tropical Diseases 13(7): e0007555, July 29, 2019. https://doi.org/10.1371/journal.pntd.0007555
Abstract
Hide Abstract, Full paper
Abstract
========
Targeted environmental and ecosystem management remain crucial in control of dengue. However, providing detailed environmental information on a large scale to effectively target dengue control efforts remains a challenge. An important piece of such information is the extent of the presence of potential dengue vector breeding sites, which consist primarily of open containers such as ceramic jars, buckets, old tires, and flowerpots. In this paper we present the design and implementation of a pipeline to detect outdoor open containers which constitute potential dengue vector breeding sites from geotagged images and to create highly detailed container density maps at unprecedented scale. We implement the approach using Google Street View images which have the advantage of broad coverage and of often being two to three years old which allows correlation analyses of container counts against historical data from manual surveys. Containers comprising eight of the most common breeding sites are detected in the images using convolutional neural network transfer learning. Over a test set of images the object recognition algorithm has an accuracy of 0.91 in terms of F-score. Container density counts are generated and displayed on a decision support dashboard. Analyses of the approach are carried out over three provinces in Thailand. The container counts obtained agree well with container counts from available manual surveys. Multi-variate linear regression relating densities of the eight container types to larval survey data shows good prediction of larval index values with an R-squared of 0.674. To delineate conditions under which the container density counts are indicative of larval counts, a number of factors affecting correlation with larval survey data are analyzed. We conclude that creation of container density maps from geotagged images is a promising approach to providing detailed risk maps at large scale.
|
■ Ziemer, Tim & Schultheis, Holger, “Psychoakustische Sonifikation zur Navigation in Bildgeführter Chirurgie”, in: Rebecca Grotjahn; Nina Jaeschke (Eds.): Freie Beiträge zur Jahrestagung der Gesellschaft für Musikforschung 2019 – 1, http://doi.org/10.25366/2020.42, 2020, pp. 347–358.
Abstract
Hide Abstract, Full paper
Abstract
========
Sonification is the systematic transformation of data to sound. Sonification is a means to communicate information, to support navigation, or to explore data. In a multi-disciplinary research project we develop a psychoacoustically-motivated sonification that supports surgeons’ orientation in image-guided interventions. One drawback of image-guidance is that neither the location of monitors nor the displayed view on the patient’s anatomy coincides with the actual viewpoint of the surgeon. Surgeons need to mentally scale, rotate, and translate the displayed graphics. These operations are cognitively demanding. Sonification can reduce cognitive load and relieve the visual channel to improve the ergonomic situation in the operating room by communicating the location of the target, relative to the tool tip, like the center of a tumor relative to the ablation needle. By means of psychoacoustic sonification we ensure that the sounds are readily interpretable in terms of orthogonal and linear dimensions with a high resolution.
|
■ Schwarz, Sebastian & Ziemer, Tim, “A Psychoacoustic Sound Design for Pulse Oximetry”, in: 25th International Conference on Auditory Display (ICAD2019), http://doi.org/10.21785/icad2019.024, Newcastle, Jun 2019.
Abstract
Hide Abstract, Full paper
Abstract
========
Oxygen saturation monitoring of neonates is a demanding task, as oxygen saturation (SpO2) has to be maintained in a particular range. However, auditory displays of conventional pulse oximeters are not suitable for informing a clinician about deviations from a target range. A psychoacoustic sonification for neonatal oxygen saturation monitoring is presented. It consists of a continuous Shepard tone at its core. In a laboratory study it was tested if participants (N = 6) could differentiate between seven ranges of oxygen saturation using the proposed sonification. On average participants could identify in 84% of all cases the correct SpO2 range. Moreover, detection rates differed significantly between the seven ranges and as a function of the magnitude of SpO2 change between two consecutive values. Possible explanations for these findings are discussed and implications for further improvements of the presented sonification are proposed.
|
■ T. Siriapisitha, W. Kusakunnirana, P. Haddawy, 3D segmentation of exterior wall surface of abdominal aortic aneurysm from CT images using variable neighborhood search, Computers in Biology and Medicine, 107, pp 73-85, April 2019.
Abstract
Hide Abstract, Full paper
Abstract
========
A 3D model of abdominal aortic aneurysm (AAA) can provide useful anatomical information for clinical management and simulation. Thin-slice contiguous computed tomographic (CT) angiography is the best source of medical images for construction of 3D models, which requires segmentation of AAA in the images. Existing methods for segmentation of AAA rely on either manual process or 2D segmentation in each 2D CT slide. However, a traditional manual segmentation is a time consuming process which is not practical for routine use. The construction of a 3D model from 2D segmentation of each CT slice is not a fully satisfactory solution due to rough contours that can occur because of lack of constraints among segmented slices, as well as missed segmentation slices. To overcome such challenges, this paper proposes the 3D segmentation of AAA using the concept of variable neighborhood search by iteratively alternating between two different segmentation techniques in the two different 3D search spaces of voxel intensity and voxel gradient. The segmentation output of each method is used as the initial contour to the other method in each iteration. By alternating between search spaces, the technique can escape local minima that naturally occur in each search space. Also, the 3D search spaces provide more constraints across CT slices, when compared with the 2D search spaces in individual CT slices. The proposed method is evaluated with 10 easy and 10 difficult cases of AAA. The results show that the proposed 3D segmentation technique achieves the outstanding segmentation accuracy with an average dice similarity value (DSC) of 91.88%, when compared to the other methods using the same dataset, which are the 2D proposed method, classical graph cut, distance regularized level set evolution, and registration based geometric active contour with the DSCs of 87.57 ± 4.52%, 72.47 ± 8.11%, 58.50 ± 8.86% and 76.21 ± 10.49%, respectively.
|
2018
|
■ W. Kusakunniran, T. Chaiviroonjaroen. Automatic Cattle Identification based on Multi-Channel LBP on Muzzle Images. Proc. of the 3rd Int'l Conference on Sustainable Information Engineering and Technology (SIET 2018), Malang, Indonesia, Nov 2018.
Abstract
Hide Abstract
Abstract
========
Every individual is unique. Biometrics authentication is mainly used to distinguish and identify each person. Parts of human body which are widely used for identification are fingerprints, faces and iris. In recent years, biometrics authentication begins to be used on animals, for the sake of population control, legal ownership and trade, and disease surveillance. This paper focuses on the automatic identification of cattles. Similar to human’s fingerprint in term of uniqueness, cattle’s muzzle is used in the identification process. In the conventional way, plastic ear tags are used to identify individual cattles. However, they can be worn down or lost easily. In addition, microchips are also used and implanted into cattles. This could injure them or cause some sickness. It is also expensive and requires human experts for the implant process. This paper introduces a novel solution using biometric images for the cattle identification. The proposed method extracts features from muzzle images using histogram of multi-channel Local Binary Pattern (LBP). This feature extraction is processed on sub-images to preserve the local spatial information of the muzzle patterns. Then, Support Vector Machine (SVM) is employed as the main classifier. The proposed method is evaluated using the published dataset containing 31different cattles. It achieves the perfect performance of 100% accuracy.
|
■ Vajsbaher T., Schultheis, H., & Francis, N.K (2018). Spatial Cognition in Minimally Invasive Surgery: A Systematic Review. BMC Surgery. 18(94). DOI: 10.1186/s12893-018-0416-1
Abstract
Hide Abstract, Full paper
Abstract
========
Background Spatial cognition is known to play an important role in minimally invasive surgery (MIS), as it was found to enable faster surgical skill acquisition, reduce surgical time and errors made and significantly improve surgical performance. No prior research attempted to summarize the available literature, to indicate the level of importance of the individual spatial abilities and how they impact surgical performance and skill acquisition in MIS.
Method Psychological and medical databases were systematically searched to identify studies directly exploring spatial cognition in MIS learning and performance outcomes. Articles written in the English language articles, published between 2006 and 2016, investigating any and all aspect of spatial cognition in direct relation to influence over performance or learning of MIS, were deemed eligible.
Results A total of 26 studies satisfied this criterion and were included in the review. The studies were very heterogeneous and the vast majority of the participants were novice trainees but with variable degree of skills. There were no clinical studies as almost all studies were conducted on either box trainers or virtual reality simulators. Mental rotation ability was found to have a clear impact on operative performance and mental practice was identified as an effective tool to enhance performance, pre-operatively. Ergonomic set-up of the MIS equipment has a marked influence on MIS performance and learning outcomes.
Conclusions Spatial cognition was found to play an important role in MIS, with mental rotation showing a specific significance. Future research is required to further confirm and quantify these findings in the clinical settings.
|
■ Vajsbaher, T., Müller, M., Schultheis, H. & Francis, N.K. (2018). Surgeons Attitudes and Opinions Towards Training in Laparoscopic Surgery. Journal of Surgical Endoscopy, Special Issue: Proceedings of the 26th International Congress of the European Association for Endoscopic Surgery (EAES).
Abstract
Hide Abstract, Full paper
Abstract
========
Background: The past two-decade has seen dramatic and impressive technological and scientific advances in laparoscopic surgery, revolutionising the manner in which laparoscopic procedures are planned, performed and learned. The impact of such change is already evident, as today, laparoscopic surgery is considered the ’gold standard’ for many operative procedures. As the field of minimally invasive surgery (MIS) continues to evolve, and further technological innovations are implemented, the surgeons’ own attitudes and opinions towards the impact of such a paradigm shift on training remains under-investigated.
Aim: The aim of this on-going study is to explore the attitude and opinion of general and visceral surgeons towards teaching and learning of laparoscopic surgery.
Method: An online questionnaire, which is currently piloted in collaboration with the Professional Association of German Surgeons (BDC) and the German Society of Surgeons (DGCH), has been made available to all members of both societies. The survey targets laparoscopic surgeons from all levels (residents and consultants) and across all surgical fields in MIS. The questionnaire aims to capture the surgeons’ opinion regarding their own laparoscopic skill competency (e.g., How would you rate your own expertise in laparoscopic surgery?) and their opinions on which specific cognitive, technical or non-technical factors they find to be most challenging when learning laparoscopy. Surgeons are also asked to report on their attitudes towards teaching and learning of laparoscopy, which will be compared with the attitudes and opinions of learners.
Results: We aim to collect around 200–250 responses from both laparoscopic trainers (registrars and consultants) and learners (surgical residents), which will be made available to be presented at the conference in London. So far 20 responses were collected, 17 were from trainers and 3 from trainees. Results obtained so far show trainees to lack significant guidance and confidence in their ability to perform laparoscopy, with trainers reporting increased stress when operating with the resident by expressing frustration towards teaching.
Conclusion: We hope to report on the surgeons’ attitudes and opinions towards teaching and learning in laparoscopic surgery, in aim to offer a greater insight into the potential issues surrounding laparoscopic education, today.
|
■ Vajsbaher, T. & Schultheis, H. (2018). A survey of surgeon’s perception and awareness of the role of spatial cognitive abilities in surgical learning. Springer Lecture Notes in Artificial Intelligence: Proceedings of the 11th International Conference on Spatial Cognition. https://doi.org/10.1007/978-3-319-96385-3
Abstract
Hide Abstract
Abstract
========
Laparoscopy is considered the gold-standard surgical procedure for diagnosis and treatment of most abdominal and pelvic-related conditions. However, although such minimally invasive technique offers countless advantages to the patient, they come at a cost: surgeons must learn to overcome a unique set of visuo-spatial cognitive and dexterity challenges, which are associated with the application of the technique. Although the existing literature provides valuable insights about the overall role of spatial cognition in laparoscopic learning, one fundamental question remains unanswered: How are these fundamental spatial cognitive challenges perceived and understood by surgeons themselves. Method. In aim to gather further insight regarding the influence of spatial cognition in surgical learning among medical professionals, an online questionnaire was circulated to members of the Professional Association of German Surgeons (BDC) and German Society of Surgeons (DGCH). The responders were asked to rank-order a list of technical, cognitive or psychological factors they found to be most challenging when learning to operate laparoscopically. Results. A clear generational difference in awareness of spatial cognition was observed; 1) Younger surgeons showed greater awareness of the taxing conditions related to spatial abilities in laparoscopy, compared to their senior colleagues, 2) lack of tactile feedback was identified as the most challenging factor for both younger and senior surgeons and finally, 3) novice surgeons identified the ability to visually process the 3D internal structures through a 2D monitor in direct relation to specific spatial louses to be most challenging. Discussion. Spatial cognition plays a notable role in influencing the learning of minimally invasive surgical skills. Nonetheless, its nature and challenges appear to be largely underestimated by the senior surgeons in charge of training, leading to the hypothesis that this is most likely one of the reasons for the observed learning curve, as no support and/or training for developing spatial skills is offered to residents with weaker spatial abilities. Conclusion. The new generation of surgeons appears to be more aware of the influential nature of spatial cognitive abilities, with lack of tactile feedback and impaired spatial orientation causing the most difficulties for novices learning minimally invasive surgery.
|
■ P. Haddawy, M. Su Yin, T. Wisanrakkit, R. Limsupavanich, P. Promrat, S. Lawpoolsri and P. Sa-angchai, Complexity-Based Spatial Hierarchical Clustering for Malaria Prediction, Journal of Healthcare Informatics Research, 2018, https://doi.org/10.1007/s41666-018-0031-z.
Abstract
Hide Abstract
, Full paper
Abstract
========
Targeted intervention and resource allocation are essential in effective control of infectious diseases, particularly those like malaria that tend to occur in remote areas. Disease prediction models can help support targeted intervention, particularly if they have fine spatial resolution. But, choosing an appropriate resolution is a difficult problem since choice of spatial scale can have a significant impact on accuracy of predictive models. In this paper, we introduce a new approach to spatial clustering for disease prediction we call complexity-based spatial hierarchical clustering. The technique seeks to find spatially compact clusters that have time series that can be well characterized by models of low complexity. We evaluate our approach with 2 years of malaria case data from Tak Province in northern Thailand. We show that the technique’s use of reduction in Akaike information criterion (AIC) and Bayesian information criterion (BIC) as clustering criteria leads to rapid improvement in predictability and significantly better predictability than clustering based only on minimizing spatial intra-cluster distance for the entire range of cluster sizes over a variety of predictive models and prediction horizons.
|
■ C. Sa-ngamuang, P. Haddawy, V. Luvira, W. Piyaphanee, S. Iamsirithaworn, S. Lawpoolsri, Accuracy of Dengue Clinical Diagnosis with and without NS1 Antigen Rapid Test: Comparison between Human and Bayesian Network Model Decision, PLOS Neglected Tropical Diseases, 12(6): e0006573, June 2018.
Abstract
Hide Abstract
, Full paper
Abstract
========
Differentiating dengue patients from other acute febrile illness patients is a great challenge among physicians. Several dengue diagnosis methods are recommended by WHO. The application of specific laboratory tests is still limited due to high cost, lack of equipment, and uncertain validity. Therefore, clinical diagnosis remains a common practice especially in resource limited settings. Bayesian networks have been shown to be a useful tool for diagnostic decision support. This study aimed to construct Bayesian network models using basic demographic, clinical, and laboratory profiles of acute febrile illness patients to diagnose dengue. Data of 397 acute undifferentiated febrile illness patients who visited the fever clinic of the Bangkok Hospital for Tropical Diseases, Thailand, were used for model construction and validation. The two best final models were selected: one with and one without NS1 rapid test result. The diagnostic accuracy of the models was compared with that of physicians on the same set of patients. The Bayesian network models provided good diagnostic accuracy of dengue infection, with ROC AUC of 0.80 and 0.75 for models with and without NS1 rapid test result, respectively. The models had approximately 80% specificity and 70% sensitivity, similar to the diagnostic accuracy of the hospital’s fellows in infectious disease. Including information on NS1 rapid test improved the specificity, but reduced the sensitivity, both in model and physician diagnoses. The Bayesian network model developed in this study could be useful to assist physicians in diagnosing dengue, particularly in regions where experienced physicians and laboratory confirmation tests are limited.
|
■ P. Haddawy, A.H.M. Imrul Hasan, R. Kasantikul, S. Lawpoolsri, P. Sa-angchai, J. Kaewkungwal, P. Singhasivanon, Spatiotemporal Bayesian Networks for Malaria Prediction, Artificial Intelligence in Medicine, 84, pp 127-138, 2018.
Abstract
Hide Abstract
, Full paper
Abstract
========
Targeted intervention and resource allocation are essential for effective malaria control, particularly in remote areas, with predictive models providing important information for decision making. While a diversity of modeling technique have been used to create predictive models of malaria, no work has made use of Bayesian networks. Bayes nets are attractive due to their ability to represent uncertainty, model time lagged and nonlinear relations, and provide explanations. This paper explores the use of Bayesian networks to model malaria, demonstrating the approach by creating village level models with weekly temporal resolution for Tha Song Yang district in northern Thailand. The networks are learned using data on cases and environmental covariates. Three types of networks are explored: networks for numeric prediction, networks for outbreak prediction, and networks that incorporate spatial autocorrelation. Evaluation of the numeric prediction network shows that the Bayes net has prediction accuracy in terms of mean absolute error of about 1.4 cases for 1 week prediction and 1.7 cases for 6 week prediction. The network for outbreak prediction has an ROC AUC above 0.9 for all prediction horizons. Comparison of prediction accuracy of both Bayes nets against several traditional modeling approaches shows the Bayes nets to outperform the other models for longer time horizon prediction of high incidence transmission. To model spread of malaria over space, we elaborate the models with links between the village networks. This results in some very large models which would be far too laborious to build by hand. So we represent the models as collections of probability logic rules and automatically generate the networks. Evaluation of the models shows that the autocorrelation links significantly improve prediction accuracy for some villages in regions of high incidence. We conclude that spatiotemporal Bayesian networks are a highly promising modeling alternative for prediction of malaria and other vector-borne diseases.
|
■ M. Su Yin, P. Haddawy, S. Suebnukarn, P. Rhienmora, Automated Outcome Scoring in a Virtual Reality Simulator for Endodontic Surgery, Computer Methods and Programs in Biomedicine, 153, pp 53-59, 2018.
Abstract
Hide Abstract
, Full paper
Abstract
========
BACKGROUND AND OBJECTIVE:
We address the problem of automated outcome assessment in a virtual reality (VR) simulator for endodontic surgery. Outcome assessment is an essential component of any system that provides formative feedback, which requires assessing the outcome, relating it to the procedure, and communicating in a language natural to dental students. This study takes a first step toward automated generation of such comprehensive feedback.
METHODS:
Virtual reference templates are computed based on tooth anatomy and the outcome is assessed with a 3D score cube volume which consists of voxel-level non-linear weighted scores based on the templates. The detailed scores are transformed into standard scoring language used by dental schools. The system was evaluated on fifteen outcome samples that contained optimal results and those with errors including perforation of the walls, floor, and both, as well as various combinations of major and minor over and under drilling errors. Five endodontists who had professional training and varying levels of experiences in root canal treatment participated as raters in the experiment.
RESULTS:
Results from evaluation of our system with expert endodontists show a high degree of agreement with expert scores (information based measure of disagreement 0.04-0.21). At the same time they show some disagreement among human expert scores, reflecting the subjective nature of human outcome scoring. The discriminatory power of the AOS scores analyzed with three grade tiers (A, B, C) using the area under the receiver operating characteristic curve (AUC). The AUC values are generally highest for the {AB: C} cutoff which is cutoff at the boundary between clinically acceptable (B) and clinically unacceptable (C) grades.
CONCLUSIONS:
The objective consistency of computed scores and high degree of agreement with experts make the proposed system a promising addition to existing VR simulators. The translation of detailed level scores into terminology commonly used in dental surgery supports natural communication with students and instructors. With the reference virtual templates created automatically, the approach is robust and is applicable in scoring the outcome of any dental surgery procedure involving the act of drilling.
|
■ Dwisaptarini A, Suebnukarn S, Rhienmora P, Koontongkaew S, Haddawy P. Effectiveness of the multilayered caries model and visuo-tactile virtual reality simulator for minimally invasive caries removal: A randomized controlled trial. Operative Dentistry, 43(3), pp E110 – E118, May/June 2018.
Abstract
Hide Abstract
, Full paper
Abstract
========
This work presents the multilayered caries model with a visuo-tactile virtual reality simulator and a randomized controlled trial protocol to determine the effectiveness of the simulator in training for minimally invasive caries removal. A three-dimensional, multilayered caries model was reconstructed from 10 micro-computed tomography (CT) images of deeply carious extracted human teeth before and after caries removal. The full grey scale 0-255 yielded a median grey scale value of 0-9, 10-18, 19-25, 26-52, and 53-80 regarding dental pulp, infected carious dentin, affected carious dentin, normal dentin, and normal enamel, respectively. The simulator was connected to two haptic devices for a handpiece and mouth mirror. The visuo-tactile feedback during the operation varied depending on the grey scale. Sixth-year dental students underwent a pretraining assessment of caries removal on extracted teeth. The students were then randomly assigned to train on either the simulator (n=16) or conventional extracted teeth (n=16) for 3 days, after which the assessment was repeated. The posttraining performance of caries removal improved compared with pretraining in both groups (Wilcoxon, p<0.05). The equivalence test for proportional differences (two 1-sided t-tests) with a 0.2 margin confirmed that the participants in both groups had identical posttraining performance scores (95% CI=0.92, 1; p=0.00). In conclusion, training on the micro-CT multilayered caries model with the visuo-tactile virtual reality simulator and conventional extracted tooth had equivalent effects on improving performance of minimally invasive caries removal.
|
■ T. Siriapisith, W. Kusakunniran, P. Haddawy, Outer wall segmentation of abdominal aortic aneurysm by variable neighborhood search through intensity and gradient spaces, Journal of Digital Imaging, (in press) 2018.
Abstract
Hide Abstract
, Full paper
Abstract
========
Aortic aneurysm segmentation remains a challenge. Manual segmentation is a time-consuming process which is not practical for routine use. To address this limitation, several automated segmentation techniques for aortic aneurysm have been developed, such as edge detection-based methods, partial differential equation methods, and graph partitioning methods. However, automatic segmentation of aortic aneurysm is difficult due to high pixel similarity to adjacent tissue and a lack of color information in the medical image, preventing previous work from being applicable to difficult cases. This paper uses uses a variable neighborhood search that alternates between intensity-based and gradient-based segmentation techniques. By alternating between intensity and gradient spaces, the search can escape from local optima of each space. The experimental results demonstrate that the proposed method outperforms the other existing segmentation methods in the literature, based on measurements of dice similarity coefficient and jaccard similarity coefficient at the pixel level. In addition, it is shown to perform well for cases that are difficult to segment.
|
■ Vannaprathip, N., Haddawy, P., Schultheis, H., Suebnukarn, S., Limsuvan, P., Intaraudom, A., Aiemlaor, N. and Teemuenvai, C., 2018, June. A Planning-Based Approach to Generating Tutorial Dialog for Teaching Surgical Decision Making. In International Conference on Intelligent Tutoring Systems (pp. 386-391). Springer, Cham.
Abstract
Hide Abstract
, Full paper
Abstract
========
Teaching surgical decision making aims at enabling students to choose the most appropriate action relative to the patient’s situation and surgical objectives. This requires a deep understanding of causes and effects related to the surgical domain as well as being aware of key properties of the current situation. To develop an intelligent tutoring system (ITS) for teaching situated decision making in the domain of dental surgery, in this paper, we present a planning-based representation framework. This framework is capable of representing surgical procedural knowledge with respect to situation awareness and algorithms that utilize the representation to generate rich tutorial dialog. The design of the tutorial dialogs is based on an observational study of surgeons teaching in the operating room. An initial evaluation shows that generated interventions are as good as and sometimes better than those of experienced human instructors.
|
■ Jasper van de Ven, Ahmed Loai Ali, Thomas Barkowsky, Christian
Freksa, Michael Epprecht, Thatheva Saphangthong and Peter Haddawy, Mobile Decision Support for Yellow-Spined Bamboo Locust Plague Intervention in Lao PDR, Proc. 14th Conf. on Location Based Services, Zurich, Jan 2018.
Abstract
Hide Abstract
, Full paper
Abstract
========
Location-based services and crowdsourced applications provide support for governments and other groups for plague and small disaster intervention. In this work in progress paper we report on an extension of the Mobile4D application to aid the government of Lao People’s Democratic Republic (Lao PDR) in dealing with the current yellow-spined bamboo locust plague. That is, we introduce the general project and approach, the Mobile4D application and specifically its locust module, report on intermediate results, and illustrate next steps to extend the support capabilities to other problems, e.g., vector-borne diseases.
|