scholarly journals Hierarchical 3-D Registration of Computed Tomography to Ultrasound Using Reinforcement Learning

10.29007/12lv ◽  
2020 ◽  
Author(s):  
Xuxin Zeng ◽  
Michael Vives ◽  
Ilker Hacihaliloglu
Keyword(s):  
Computed Tomography ◽  
Reinforcement Learning ◽  
Success Rate ◽  
Evaluation Studies ◽  
Computer Assisted ◽  
Registration Method ◽  
Q Learning ◽  
Target Registration Error ◽  
Ct Data ◽  
Computer Assisted Orthopedic Surgery

In ultrasound (US)-based computer-assisted orthopedic surgery (CAOS), accurate and robust intra-operative registration in real-time is vital in securing the reliable outcomes for surgical image guidance. For this purpose, we focus on developing a hierarchical registration method, using reinforcement learning (RL), for 3-D registration of pre-operative computed tomography (CT) data to intra-operative US. In the RL-based registration procedure, we proposed a supervised Q-learning framework for learning the sequence of motion action to achieve the optimal alignment. Within the approach, the agent was modeled using PointNet++ with the mis-aligned point set from US and CT as the input, and the next optimal action as the output. Evaluation studies achieved average target registration error (TRE) of 3.82 mm with success rate of 92.7% and an average time of 8.36 seconds. We achieve 57.1% improvement in success rate over state of the art.

Spatial Dysmorphology of the Foot in Apert Syndrome: Three-dimensional Computed Tomography

1995 ◽  
Vol 32 (3) ◽  
pp. 255-262 ◽  
Author(s):  
E. Dale Collins ◽  
Jeffrey L. Marsh ◽  
Michael W. Vannier ◽  
Louis A. Gilula
Keyword(s):  
Computed Tomography ◽  
High Resolution ◽  
Congenital Malformations ◽  
Three Dimensional ◽  
Apert Syndrome ◽  
Computer Assisted ◽  
Foot Surgery ◽  
Joint Morphology ◽  
Ct Data ◽  
Dimensional Surface

Computer assisted medical imaging was used to define the spatial dysmorphology of the foot in three patients with Apert syndrome and to correlate that dysmorphology with ambulation and footwear. Thin slice (2 mm), abutting, high resolution axial computed tomography (CT) foot scans were obtained. The CT data were post processed, using Analyze, to generate three-dimensional surface shaded and volumetric reformations. The reformatted images were evaluated by a bone and joint radiologist to identify abnormalities of bone shape, size, and orientation, of joint morphology, and of the foot as a whole. Five consistent findings were observed among the three pairs of feet: (1) anomalous great toes with phalangeal and metatarsal pathology; (2) simple syndactyly of toes 2-5; (3) fusions between metatarsals; (4) tarsal coalitions; and (5) limitations in commercial footwear. One patient underwent bilateral fifth metatarsal wedge osteotomies to facilitate the wearing of shoes. The dysmorphology of the Apert foot is a combination of congenital malformations and postnatal deformations, secondary to progressive synostosis. Prophylactic foot surgery may be indicated in Apert patients to facilitate shoe fitting.

scholarly journals Morphometric Evaluation of Detailed Asymmetry for the Proximal Humerus in Korean Population

Symmetry ◽  
2021 ◽  
Vol 13 (5) ◽  
pp. 862
Author(s):  
Eunah Hong ◽  
Dai-Soon Kwak ◽  
In-Beom Kim
Keyword(s):  
Proximal Humerus ◽  
Contralateral Side ◽  
Patient Specific ◽  
Computer Assisted ◽  
Patient Specific Instruments ◽  
Left And Right ◽  
Ct Data ◽  
Bilateral Differences ◽  
Computer Assisted Orthopedic Surgery ◽  
Morphometric Evaluation

Computer-assisted orthopedic surgery and patient-specific instruments are widely used in orthopedic fields that utilize contralateral side bone data as a template to restore the affected side bone. The essential precondition for these techniques is that the left and right bone features are similar. Although proximal humerus fracture accounts for 4% to 8% of all fractures, the bilateral asymmetry of the proximal humerus is not fully understood. The aim of this study is to investigate anthropometric differences of the bilateral proximal humerus. One hundred one pairs of Korean humerus CT data from 51 females and 50 males were selected for this research. To investigate bilateral shape differences, we divided the proximal humerus into three regions and the proximal humerus further into five sections in each region. The distance from the centroid to the cortical outline at every 10 degrees was measured in each section. Differences were detected in all regions of the left and right proximal humerus; however, males had a larger number of significant differences than females. Large bilateral differences were measured in the greater tubercle. Nevertheless, using contralateral data as a template for repairing an affected proximal humerus might be possible.

scholarly journals Reinforcement Learning to Improve Image-Guidance of Ablation Therapy for Atrial Fibrillation

2021 ◽  
Vol 12 ◽  
Author(s):  
Laila Muizniece ◽  
Adrian Bertagnoli ◽  
Ahmed Qureshi ◽  
Aya Zeidan ◽  
Aditi Roy ◽  
...  
Keyword(s):  
Atrial Fibrillation ◽  
Reinforcement Learning ◽  
Success Rate ◽  
Recurrence Rate ◽  
Image Guidance ◽  
Treatment Success ◽  
Patient Specific ◽  
Functional Information ◽  
Q Learning ◽  
Af Recurrence

Atrial fibrillation (AF) is the most common cardiac arrhythmia and currently affects more than 650,000 people in the United Kingdom alone. Catheter ablation (CA) is the only AF treatment with a long-term curative effect as it involves destroying arrhythmogenic tissue in the atria. However, its success rate is suboptimal, approximately 50% after a 2-year follow-up, and this high AF recurrence rate warrants significant improvements. Image-guidance of CA procedures have shown clinical promise, enabling the identification of key patient anatomical and pathological (such as fibrosis) features of atrial tissue, which require ablation. However, the latter approach still suffers from a lack of functional information and the need to interpret structures in the images by a clinician. Deep learning plays an increasingly important role in biomedicine, facilitating efficient diagnosis and treatment of clinical problems. This study applies deep reinforcement learning in combination with patient imaging (to provide structural information of the atria) and image-based modelling (to provide functional information) to design patient-specific CA strategies to guide clinicians and improve treatment success rates. To achieve this, patient-specific 2D left atrial (LA) models were derived from late-gadolinium enhancement (LGE) MRI scans of AF patients and were used to simulate patient-specific AF scenarios. Then a reinforcement Q-learning algorithm was created, where an ablating agent moved around the 2D LA, applying CA lesions to terminate AF and learning through feedback imposed by a reward policy. The agent achieved 84% success rate in terminating AF during training and 72% success rate in testing. Finally, AF recurrence rate was measured by attempting to re-initiate AF in the 2D atrial models after CA with 11% recurrence showing a great improvement on the existing therapies. Thus, reinforcement Q-learning algorithms can predict successful CA strategies from patient MRI data and help to improve the patient-specific guidance of CA therapy.

Computer-Assisted Categorizing of Head Computed Tomography Reports for Clinical Decision Rule Research

2006 ◽  
Vol 48 (5) ◽  
pp. 551-557.e25
Author(s):  
Stephen P. Wall ◽  
Oliver Mayorga ◽  
Christine E. Banfield ◽  
Mark E. Wall ◽  
Ilan Aisic ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Decision Rule ◽  
Clinical Decision ◽  
Computer Assisted ◽  
Clinical Decision Rule ◽  
Head Computed Tomography

The diagnosis of early pneumoconiosis in dust-exposed workers: comparison of chest radiography and computed tomography

2021 ◽  
pp. 028418512110225
Author(s):  
Hideyuki Hayashi ◽  
Kazuto Ashizawa ◽  
Masashi Takahashi ◽  
Katsuya Kato ◽  
Hiroaki Arakawa ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Chest Radiography ◽  
Pulmonary Emphysema ◽  
Early Stage ◽  
Pathological Changes ◽  
Occupational History ◽  
Exposed Workers ◽  
History Of ◽  
Ct Data ◽  
Positive Results

Background Chest radiography (CR) is employed as the evaluation of pneumoconiosis; however, we sometimes encounter cases in which computed tomography (CT) is more effective in detecting subtle pathological changes or cases in which CR yields false-positive results. Purpose To compare CR to CT in the diagnosis of early-stage pneumoconiosis. Material and Methods CR and CT were performed for 132 workers with an occupational history of mining. We excluded 23 cases of arc-welder’s lung. Five readers who were experienced chest radiologists or pulmonologists independently graded the pulmonary small opacities on CR of the remaining 109 cases. We then excluded 37 cases in which the CT data were not sufficient for grading. CT images of the remaining 72 cases were graded by the five readers. We also assessed the degree of pulmonary emphysema in those cases. Results The grade of profusion on CR (CR score) of all five readers was identical in only 5 of 109 cases (4.6%). The CR score coincided with that on CT in 40 of 72 cases (56%). The CT score was higher than that on CR in 13 cases (18%). On the other hand, the CT score was lower than that on CR in 19 cases (26%). The incidence of pulmonary emphysema was significantly higher in patients whose CR score was higher than their CT score. Conclusion CT is more sensitive than CR in the evaluation of early-stage pneumoconiosis. In cases with emphysema, the CR score tends to be higher in comparison to that on CT.

scholarly journals Personalized project recommendations: using reinforcement learning

2019 ◽  
Vol 2019 (1) ◽  
Author(s):  
Faxin Qi ◽  
Xiangrong Tong ◽  
Lei Yu ◽  
Yingjie Wang
Keyword(s):  
Reinforcement Learning ◽  
User Behavior ◽  
Collaborative Work ◽  
Recursive Least Squares ◽  
The Internet ◽  
Dynamic Impact ◽  
Rls Algorithm ◽  
Trust Value ◽  
Q Learning ◽  
Actual Evaluation

AbstractWith the development of the Internet and the progress of human-centered computing (HCC), the mode of man-machine collaborative work has become more and more popular. Valuable information in the Internet, such as user behavior and social labels, is often provided by users. A recommendation based on trust is an important human-computer interaction recommendation application in a social network. However, previous studies generally assume that the trust value between users is static, unable to respond to the dynamic changes of user trust and preferences in a timely manner. In fact, after receiving the recommendation, there is a difference between actual evaluation and expected evaluation which is correlated with trust value. Based on the dynamics of trust and the changing process of trust between users, this paper proposes a trust boost method through reinforcement learning. Recursive least squares (RLS) algorithm is used to learn the dynamic impact of evaluation difference on user’s trust. In addition, a reinforcement learning method Deep Q-Learning (DQN) is studied to simulate the process of learning user’s preferences and boosting trust value. Experiments indicate that our method applied to recommendation systems could respond to the changes quickly on user’s preferences. Compared with other methods, our method has better accuracy on recommendation.

4D Computed Tomography for Dynamic Upper Airway Evaluation in Robin Sequence

2021 ◽  
pp. 019459982110021
Author(s):  
Austin S. Lam ◽  
Michael D. Bindschadler ◽  
Kelly N. Evans ◽  
Seth D. Friedman ◽  
Jeffrey P. Otjen ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Upper Airway ◽  
Quantitative Information ◽  
4D Ct ◽  
Robin Sequence ◽  
Analysis Tools ◽  
Airway Caliber ◽  
4D Computed Tomography ◽  
Ct Data ◽  
Sample Case

Thorough assessment of dynamic upper airway obstruction (UAO) in Robin sequence (RS) is critical, but traditional evaluation modalities have significant limitations. Four-dimensional computed tomography (4D-CT) is promising in that it enables objective and quantitative evaluation throughout all phases of respiration. However, there exist few protocols or analysis tools to assist in obtaining and interpreting the vast amounts of obtained data. A protocol and set of data analysis tools were developed to enable quantification and visualization of dynamic 4D-CT data. This methodology was applied to a sample case at 2 time points. In the patient with RS, overall increases in normalized airway caliber were observed from 5 weeks to 1 year. There was, however, continued dynamic obstruction at all airway levels, though objective measures of UAO did improve at the nasopharynx and oropharynx. Use of 4D-CT and novel analyses provide additional quantitative information to evaluate UAO in patients with RS.

scholarly journals Size selection of the Ambu AuraOnce laryngeal mask in Chinese men weighing >70 kg: a pilot study

2021 ◽  
Vol 49 (5) ◽  
pp. 030006052110166
Author(s):  
Jiahui Chen ◽  
Chunhuan Chen ◽  
Wei Xu ◽  
Xiaoguang Zhang
Keyword(s):  
Computed Tomography ◽  
Success Rate ◽  
Fiberoptic Bronchoscopy ◽  
Laryngeal Mask ◽  
Insertion Time ◽  
Air Leakage ◽  
Chinese Men ◽  
Laryngeal Mask Insertion ◽  
Laryngeal Anatomy ◽  
Rate Requirement

Objective To collect computed tomography data of the laryngeal anatomy of Chinese men and to determine the feasibility of using the size 4 Ambu AuraOnce laryngeal mask (Ambu A/S, Copenhagen, Denmark) in Chinese men weighing >70 kg. Methods This prospective study involved men who underwent surgery from May 2018 to January 2019 at Jinshan Hospital. Pharyngeal and laryngeal parameters were measured by computed tomography. The laryngeal mask insertion success rate, requirement for tracheal tube insertion, laryngeal mask insertion time, fiberoptic bronchoscopy grading, air leakage pressure, and pharyngeal complications were analyzed. Results In a comparison of the size 4 and 5 Ambu AuraOnce devices, the first insertion success rate was 100% and 87% and the three-times insertion success rate was 100% and 93%, respectively, with no significant differences. However, the insertion time was significantly different at 19.6 ± 5.9 versus 31.1 ± 11.2 s, respectively, and the proportions of fiberoptic grading levels were also significantly different. There were no significant differences in the air leakage pressure or pharyngeal complications. Conclusion The size 4 Ambu AuraOnce is more adequate than the size 5 for Chinese men weighing >70 kg, with a shorter insertion time and higher fiberoptic bronchoscopic grading.

scholarly journals Integrating Production Planning with Truck-Dispatching Decisions through Reinforcement Learning While Managing Uncertainty

Minerals ◽  
2021 ◽  
Vol 11 (6) ◽  
pp. 587
Author(s):  
Joao Pedro de Carvalho ◽  
Roussos Dimitrakopoulos
Keyword(s):  
Reinforcement Learning ◽  
Discrete Event ◽  
Mining Operations ◽  
Fixed Sequence ◽  
Q Learning ◽  
Reward Function ◽  
Copper Gold ◽  
Mining Complex ◽  
Learning Reinforcement ◽  
Operational Plan

This paper presents a new truck dispatching policy approach that is adaptive given different mining complex configurations in order to deliver supply material extracted by the shovels to the processors. The method aims to improve adherence to the operational plan and fleet utilization in a mining complex context. Several sources of operational uncertainty arising from the loading, hauling and dumping activities can influence the dispatching strategy. Given a fixed sequence of extraction of the mining blocks provided by the short-term plan, a discrete event simulator model emulates the interaction arising from these mining operations. The continuous repetition of this simulator and a reward function, associating a score value to each dispatching decision, generate sample experiences to train a deep Q-learning reinforcement learning model. The model learns from past dispatching experience, such that when a new task is required, a well-informed decision can be quickly taken. The approach is tested at a copper–gold mining complex, characterized by uncertainties in equipment performance and geological attributes, and the results show improvements in terms of production targets, metal production, and fleet management.

Sign in / Sign up

Export Citation Format

Share Document