Computer-aided detection of pulmonary embolism (CAD):

Slides:



Advertisements
Similar presentations
3D Imaging Research Massachusetts General Hospital Harvard Medical School 3D Computer-Aided Diagnosis (CAD) in Radiology Hiro Yoshida, PhD.
Advertisements

CARDIOVASCULAR IMAGING WITH COMPUTED TOMOGRAPHY(CT)
Cardiac applications of 16 slice MDCT :Initial experience
CAD Performance Analysis for Pulmonary Nodule Detection: Comparison of Thick- and Thin-Slice Multi- detector CT Scans Randy D Ernst 1, Russell C Hardie.
Novel single-source high-pitch protocol for CT angiography of the aorta: comparison to high-pitch dual-source protocol in the context of TAVI planning.
ACR and SBI Statement Margarita Zuley, MD Associate Professor, Radiology Medical Director, Breast Imaging University of Pittsburgh.
Computer Aided Diagnosis: CAD overview
Liver Segmentation Using Active Learning Ankur Bakshi Allison Petrosino Advisor: Dr. Jacob Furst August 21, 2008.
Special Imaging Techniques Chapter 6 Bushong. Dynamic Computed Tomography (DCT) Dynamic scanning implies 15 or more scans in rapid sequence within one.
What Are They? Computed Tomography Angiography (CTA)
Radiologic investigation of Chest and CVS diseases
Radiology Research at BUMC Jorge Soto, MD Vice Chair Research, Radiology Associate Professor, Radiology 12/19/07.
Multislice CT Coronary Angiography
Saira Ahmad UOG. CAT Scans CAT Scans ( Computerized axial tomography) Topic:
Number of Entry Tears Is Associated With Aortic Growth in Type B Dissections Ann Thorac Surg March 28, 2013 Thoracic Aortic Research Center, University.
EVALUATION OF ACUTE PULMONARY EMBOLISM USING SPIRAL CT SCAN 1 Sudan University of Science and Technology, College of Medical Radiologic Science, P.O.Box.
Revision Dr Mohamed El Safwany, MD.. Liver CT Blood circulation in the liver comprises two major components: the hepatic artery and the portal vein. After.
Radiographic Evaluation of a Pulmonary Embolism Dr Mohamed El Safwany, MD.
THE DICOM 2013 INTERNATIONAL CONFERENCE & SEMINAR March 14-16Bangalore, India On handling low-dose CT images in the absence of reliable DICOM header information.
Intracranial AVMs: Comparison of volumes generated from orthogonal measurements and integrated 3D analysis Faiz I Syed MD MS 1, Lubomir Hadjiiski PhD 1,
IN COMPUTED TOMOGRAPHY
Computed Tomography Angiography (CTA). What is CT Angiography? An examination that uses x-rays to visualize blood flow in arterial and venous vessels.
CAD Performance Analysis for Pulmonary Nodule Detection: Comparison of Thick- and Thin-Slice Helical CT Scans Randy D Ernst 1, Russell C Hardie 2, Metin.
Intended learning outcome The student should learn at the end of this lecture procedures of CT pulmonary angiography.
MEASUREMENTS OF RADIATION DOSES IN MULTISLICES COMPUTED TOMOGRAPHY EXAMINATIONS A. ELMAHDI*, A. SULIEMAN *Presenting author 1 Sudan Atomic Energy Commission,
Michael Ficorelli.  To describe clinical indications for C.T.A. examinations in the circulatory system. To understand and recognize anatomy and landmarks.
Part No...., Module No....Lesson No
Issues in diagnosis of VTE in Pregnancy Ng Heng Joo Department of Haematology Singapore General Hospital.
CT ANGIOGRAPHY Dr Mohamed El Safwany, MD. Intended learning outcome The student should learn at the end of this lecture CT IMAGE OF THE BLOOD VESSEL OPACIFIED.
Diagnostic Imaging on Intracranial Atherosclerotic Stenosis Eduardo Freire Mello Department of Interventional Neuroradiology Hospital Espanhol, Salvador.
Kamran M 1, Deuerling-Zheng 2, Mueller-Allissat B 2, Grunwald IQ 1, Byrne JV 1 1. Oxford Neurovascular and Neuroradiology Research Unit, University of.
Impact of Contrast Media Concentration and kVp settings on Image Quality in CT Angiography of the Intracranial Vessels Birgitta Ramgren MD Roger Siemund.
MAGNETIC RESONANCE IMAGING AND MAGNETIC RESONANCE ANGIOGRAPHY IN DIAGNOSIS OF INTRACRANIAL ANEURYSMS Merhemic Z¹, Gavrankapetanovic F¹, Nikšić M¹, Avdagic.
Certainty of Stroke Diagnosis: Incremental Benefit with CT Perfusion over NC-CT & CTA Richard I. Aviv, Julia Hopyan, Anthony Ciarallo, et al (including.
Duel Acquisition Neck CTA/ CT for Pre-TLM H&N Ca Patient Evaluation Steven M. Weindling, M.D. Mayo Clinic Florida XIX Symposium.
M5L: servizio di analisi on demand di CT polmonari.
Pulmonary Embolism Pulmonary Embolism Ma hong Depart. of Medical Imaging, Xuzhou Medical College.
Date of download: 5/30/2016 Copyright © The American College of Cardiology. All rights reserved. From: Randomized Comparison of 64-Slice Single- and Dual-Source.
Wednesday Case of the Day History: CT pulmonary angiography (image shown) was performed on a 24-year-old female patient with pleuritic chest pain and increased.
J.Calatayud Moscoso del Prado, D. Castellón Plaza, C. Trinidad López, R. Prada González. POVISA, Vigo (SPAIN)
CT Image Quality for CT Pulmonary Embolism (CT PE) and Chest CT Aorta Acquired with Power Injection via an Arm Port Ryan Verity 1, David Leswick.
Pulmonary Embolism in Patients with Unexplained Exacerbation of COPD: Prevalence and Risk Factors Isabelle Tillie-Leblond, MD, PhD; Charles-Hugo Marquette,
Background Computed tomography coronary angiography is an accurate method for the non-invasive diagnosis of coronary artery disease Eur Heart J 2005;26:1482–1487.
Diagnosis Recitation. The Dilemma At the conclusion of my “diagnosis” presentation during the recent IAPA meeting, a gentleman from the audience asked.
Automated airway evaluation system for multi-slice computed tomography using airway lumen diameter, airway wall thickness and broncho-arterial ratio B.
Choosing Wisely : Radiology Perspective
Radiology 2005;236: Intern 陳君豪 2005/08/17
45-year-old woman with BMI of 23.1 and chest circumference of 94.0 cm. Axial CT images obtained at 120 kV and 200 mAs show ascending aorta with image noise.
Pulmonary Embolism: CT-ANGIOGRAPHY
The Evaluation of Suspected Pulmonary Embolism
Aortic Dissection.
Journal Reading: CT Angiography of Pulminary Embolism: Diagnostic Criteria and Causes of Misdiagnosis Radiographics 2004; 24(5):
Robust Lung Nodule Classification using 2
Lung Ventilation-Perfusion Scan (V/Q Scan) 2015/2016
Gateways to the heart – Incidental CT findings of anomalous systemic venous connections and the clinical challenges they present Hanzhou Li, Christopher.
Henderson JM, Moslim MA, Knox MK, Cowan NC
Coronary computed tomography angiography (CTA) obtained on a 35-year-old female presenting to an outpatient clinic with a history of unexplained syncope.
Applications of Diagnostic Imaging Computed Tomography
Diagnostic Imaging Required Reading: Pg
Introduction Results Material & Methods Conclusions
Strategies for Reducing Radiation Exposure From Multidetector Computed Tomography in the Acute Care Setting  Aaron Sodickson, MD, PhD  Canadian Association.
Outline Introduction Background Our Approach Experimental Results
Pulmonary Embolism /Pulmonary hypertension
CT Angiography of the Aorta: Contrast Timing by Using a Fixed versus a Patient-specific Trigger Delay In 108 participants undergoing CT angiography of.
Iodine Maps from Subtraction CT or Dual-Energy CT to Detect Pulmonary Emboli with CT Angiography: A Multiple-Observer Study Subtraction CT showed higher.
Limitations of two-dimensional (2D) measurements.
Margot Sommer et al. JIMG 2015;8:
Hisa Shimojima et al. JACEP 2018;4:
Evidence Based Diagnosis
Presentation transcript:

Computer-aided detection of pulmonary embolism (CAD): can junior radiologist benefit from consensus with CAD in detecting of segmental and subsegmental pulmonary embolism? Mr. chairman, ladies and gentlemen. I would like to introduce our work about new software tool for automated detection of pulmonary embolism. J. Baxa, J. Ferda, A. Bednarova, R. Vondrakova, H. Mirka, Radiology clinic, University Hospital Plzen, Czech Republic

introduction and goals PE-CAD: new automated software for detection of PE Syngo Circulation® - Siemens Medical Solutions  evaluate performance of PE-CAD  possible benefit for junior radiologist  compare results with other studies The pulmonary embolism CAD software is a new part of originally cardiac application distributed on workstations by Siemens Medical. We decided to find out performance of this prototype software, possible benefit for young radiologist and finally compare our results with other available studies. Zhou C et al. Preliminary investigation of computer.aided detection of pulmonary embolism in three-dimensional computed tomography pulmonary angiography images, Acad Radiol 2005

CT pulmonary angiography 64-slice CT (Somatom Sensation 64, Siemens) standard protocol: 64x0.6 mm; rotation 0.33 s; pitch 0.95; 120 kV, 140 eff. mAs contrast: 80 ml i.v. (Iomeron 400, Bracco ALTANA Pharma) + 60 ml flush of saline flow rate: 5 ml/sec + 3 ml/s bolus tracking: 80 HU at ascending aorta data set: slice 0.6 mm, increment 0.4 mm, soft tissue kernel (Siemens B30f – medium smooth) We reviewed our set of CT pulmonary angiographies performed on 64-slice CT with this standard protocol. We administrated 80 ml of contrast agent with concentration 400 mg of iodine per ml and flush of 60 ml of solution saline. Bolus tracking was placed in ascending aorta with threshold of 80 HU. Standard data sets used for evaluation of pulmonary vessels were reconstructed with slice 0.6 mm, increment 0.4 mm and soft tissue kernel.

patient selection segmental and subsegmental embolism 18 positive (9a 9,  58.1) 18 negative (13a 5,  60.7) 78 emboli (41 seg., 37 subseg.);  4.33; <1;9> board of 2 certified radiologist with specialisation in thoracic imaging CAD ………………………… Syngo Circulation® (Siemens) JR …………………………… junior radiologist (3 years praxis) JR+CAD …………………… consensus of JR with CAD SR …………………………… senior radiologist (15 years praxis) In our study, we concentrated on segmental and subsegmental embolism. So we picked up 18 positive and 18 negative examinations. Total count of emboli by this patients was 78, 41 segmental and 37 subsegmental, it means more than 4 emboli per patient. Minimal count by one patient was 1 embolus and maximum 9 emboli. At first, the data sets of our selection were applied to CAD software. Then same selection had been analyzed by junior radiologist with time limit of 10 minutes and after completing and storing was results compared with results of CAD. For control, we also applied same selection to one of senior radiologist at our clinic.

25 y.o. female false positive Just one short example of young female 25 years old, using hormonal contraception with short episode of dyspnoea and swelling of right leg. On the first image you can see basic background of PE-CAD application with list of 3 possible emboli offered by software. Here is first one in artery of third segment, a second one in bifurcation of lower lobe artery and third one in segmental branch for tenth segment. This was ideal example, but often CAD offers false negatives and more often false positives. Most frequent false positives are: pulmonary vein, azygos vein, hilar or intrapulmonary lymph nodes.

results – „per embolus“ total segmental subsegmental CAD Sensitivity 56% 73% 38% PPV 42% JR Sensitivity 83% 95% 70% PPV 93% 100% 85% JR+CAD Sensitivity 87% 97% 76% PPV 96% 100% 85% SR Sensitivity 92% 100% 89% PPV 96% 100% 94% And now results in per embolus analysis. Overall „per embolus“ sensitivity of CAD reached 56%, especially at subsegmental level was sensitivity of CAD very low. Sensitivity of junior radiologist was more than 80% and especially in detection of subsegmental embolisation is evident benefit in sensitivity after confrontation with results of CAD. Engelke C et al. Computer-assisted detection of pulmonary embolism: performance evaluation in consensus with experienced and inexperienced chest radiologists. Eur Radiol 2007

results – „per patient“ CAD - sensitivity 83%, specificity 61% - PPV 68%, NPV 79% JR + CAD - sensitivity 95%, specificity 89%, - PPV 90%, NPV 94% SR - sensitivity 100%, specificity 94% - PPV 95%, NPV 100% In per patient analysis, true positive subject is settled, if at least one embolus is truely found. Young radiologist in cooperation with CAD reached very high sensitivity, 95%.

comparison results of CAD in recent studies are different study no. emboli SE_CAD total segmental subsegmental False positive Plzen, CZE (Siemens) 78 57% 73% 38%  3.1 Würzburg, GER 820 34% 41%  28%   4.1 Munich, GER 147 82%   3.9 Aachen, GER (Siemens – new version) 215 83% 87% 77%  4.0 results of CAD in recent studies are different all studies proved contribution for young radiologist (4-10%) new version of PE-CAD is really better in comparison high average of false positives Comparison of CAD sensitivity with other studies evaluating same CAD software is shown in this table. As you can see results are a little bit different, but all studies proved significant contribution of CAD to sensitivity of young radiologist. Colleagues from Aachen had possibility to evaluate a new version of CAD from Siemens and their results are very promising. Persisting problem is a high rate of false positive findings. Das M et al. Computer-aided detection of pulmonary embolism: influence on radiologists´ detection performance, Eur Radiol 2008

conclusions PIOPED II: pulmonary CTA (pCTA) - first line modality in detection of pulmonary embolism (PE) PE-CAD: innovation in CAD systems easy and fast usage x lower sensitivity great potential as a „second reader“ new development is necessary In conclusion, according to study PIOPED II, pulmonary CTA is gold standard in detection of pulmonary embolism, so number of pulmonary CTA´s is growing. PE-CAD is innovation in CAD system, which is very simple and fast to use, but has lower sensitivity. According to our and other results has CAD software from Siemens great potential as a „second reader“ for junior radiologist, but new development is neccesary. Chan HP et al. Computer-aided diagnosis of lung cancer and pulmonary embolism in computed tomography-a review. Acad Radiol 2008 May;15(5):535-55.