«MEDUSA DELIVERABLE D122 Demonstration report & Clinical validation ...»
D122 Demonstration report & Clinical validation
Project number: ITEA 10004
Document version no.: V1.0
Edited by: WP1 partners
ITEA Roadmap domains:
Major: Content & Knowledge
ITEA Roadmap categories:
Minor: Network & computing MEDUSA ITEA 10004 WP1 Deliverable D121 Page 2 of 12 HISTORY Document Date Remarks Updated by version # 1 30-11-2015 Final, public Jordi Borst MEDUSA public 02/12/2015 MEDUSA ITEA 10004 WP1 Deliverable D121 Page 3 of 12
TABLE OF CONTENTS1 INTRODUCTION
1.1 Aim of activity
1.2 Overview of Medusa demonstrators
2 ACUTE SCENARIO: ISCHEMIC STROKE DEMONSTRATION
3 ACUTE SCENARIO: TRAUMA DEMONSTRATION
4 POST TRAUMATIC COMA DEMONSTRATION
5 CONTOURING FOR ONCOLOGY DEMONSTRATION
6 COMPUTER ASSISTED IN VITRO DIAGNOSTICS DEMONSTRATION........... 9
7 COLLABORATIVE DECISION SUPPORT FOR ONCOLOGYDEMONSTRATION
MEDUSA public 02/12/2015 MEDUSA ITEA 10004
1.1 Aim of activity This document provides the report of the user experiences of the demonstrators See D121 Demonstrator plan). The experience of the user is reported per demonstration.
The demonstrations and discussions with the clinical experts are recorded for most demonstrations. In this document we briefly report what the clinical expert liked from the Medusa system and what they recommend to implement. We also report what we have learned from the clinical experts.
1.2 Overview of Medusa demonstrators Acute scenario: Ischemic Stroke demonstration Acute scenario: Trauma demonstration Post traumatic Coma demonstration Contouring for oncology demonstration Computer assisted in vitro diagnostics demonstration Collaborative decision support for oncology demonstration
2 Acute scenario: Ischemic Stroke demonstration Demonstration date: 10-11-2015 Location: Academic Medical Center Amsterdam Clinical experts present: a Neuroradiologist and an Emergency radiologist The demonstration and discussion with the clinical experts is recorded (see Google drive, see reference section for the movies) Likes Fast time to treatment leads to better outcome for patients with a stroke. The Medusa system saves valuable time in the diagnosis of stroke since part of the investigation can now be performed in the ambulance.
Using the Medusa system a decision can be made to which hospital a patient should be transported. This saves a lot of time because the patient goes directly to the hospital that can provide the level of care this patents needs and does not have to be transferred to another hospital if the initial hospital cannot provide the needed level of care.
It is very valuable that the vital signs can be streamed from the ambulance to the hospital. This allows the doctors to inspect for example the blood pressure and give instructions to the paramedic if needed.
The interactive medical image viewer allows a direct second opinion or expert opinion in the acute situation.
Recommendations Add a map with the location of the ambulance and the estimated arrival time. The care team in the hospital can see exactly where the ambulance is and how much time they have to prepare before the patient arrives.
Add an easy form to score the stroke severity using The National Institutes of Health Stroke Scale (NIHSS) or the The Los Angeles Motor Scale (LAMS) scoring systems. Based on the score on these scales and the professional opinion of the medical doctor in the hospital patients can be triaged in the ambulance.
Patients with a severe stroke can be directly transported to a specialised intervention centre. This saves valuable time since these patient do not need to be transferred.
Add Medusa in the intervention room and operating room. This allows direct collaboration with remote colleagues during the medical procedure.
The video conferencing can also be used to record the conversation between the patient and the medical doctor. The patient can view the conversation later in case important information was forgotten. It is difficult for patients to remember information of such a conversation especially if they received some bad news.
It would be valuable if future Eletronic Patient records can be integrated into Medusa.
Learning points Post processing of the medical images is nice but is less important in acute situation. It is more than sufficient that medical doctors can view the medical images regardless of their location. It is important that there is no latency in the interaction with the image viewer. Fast and reliable internet connection is important Patient privacy is very important and the cloud should be secure.
3 Acute scenario: Trauma demonstration Demonstration date: 14-10-2015 and 20-11-2015 (and 26-10-2015) Location: Academic Medical Center Amsterdam (Hôpital Pitié Salpétrière in Paris) Clinical experts present: Trauma surgeons (Trauma radiologists) The demonstration and discussion with the clinical experts on 14-10-2015 has been recorded (see Google drive, see reference section for the movies) Likes The experts are enthousiastic about the continuous stream of vital sign data that is captured starting in the ambulance and shared in real-time with medical doctors in one or more hospitals. Currently vital sign information from the ambulance is piecemeal and does not allow to identify trends. As vital sign information is crucial to diagnosis, the experts expect that the immediate and continuous availability of vital sign data will help significantly reduce under triage and over triage. From all major trauma patients, 20% currently ends up in a local hospital whereas they should have been taken to a level 1 trauma centre. A higher accuracy in triage benefits the patient and the hospital (efficient use of resources).
The alerts generated by the decision support application will help the trauma team to keep track of changes in the patient’s condition, esp. to detect dangerous combinations of factors that require immediate action. It is also well-received that decision support rules can be defined by the medical doctors in the context of a medical protocol, and that the parameters and values used in the rules are configurable. By defining medical protocols for specific groups of patients, the decision support allows to standardise and finetune the trauma care provided.
Recommendations As the MEDUSA services require an internet connection and internet coverage may be low or absent in parts of the country, it should be possible to capture and display vital sign data and run decision support on a local network in the ambulance.
The vital signs application and decision support application will also be very useful for monitoring patients with a chronic disease, a development that hospitals are preparing for.
Both hospitals think that ambulance services will be very interested in the applications of this demonstrator and strongly recommend to involve ambulance services in followup discussions on further development and implementation.
The applications in the Trauma demonstrator are also relevant for other emergency cases: cardiac and stroke.
Learning points In France the ambulance staff may include a medical doctor (not just paramedics), but there also it is considered useful to be able to collaborate in real-time with colleagues elsewhere. The vital signs display showing trends and the alerts generated by decision support continue to be important.
Continuously capture and display a vital sign like Respiratory Rate is an innovation in itself as this information is lacking nowadays.
The demonstrator tends to slow down or cannot be fully used when accessed on a tablet or other mobile device.
4 Post traumatic Coma demonstration Demonstration date: October 26, 2015 Location: Paris, La Pitié Salpêtrière Hospital, Radiology Service Clinical experts present: Prof. Ph. Grenier, Prof. D. Dormont, Prof. O. Lucidarme Technical staff: Rahul Gaurav, R&D engineer The demonstration and discussion with the clinical experts could not be recorded;
however, a functional demo was filmed during the Best MEDUSA workshop and available under the conditions specified by the MEDUSA consortium. (see Google drive, see reference section for the movies Likes
- the usage of the virtualized application does not require from the clinical expert any technical insight related to the cloud
- the information required in medical decision making is not downgraded during the cloud portability process
- all the functionalities available in the local set-up are preserved in cloud deployment
- the possibility of using the application from a mobile terminal
- the collaboration functionality which was added by MEDUSA
- the collaboration between two mobile terminals was fluid
- the possibility of using unitary identification
- the possibility of displaying personalized content Recommendations
- prepare an awareness discourse for popularizing the application virtualization inside the medical community Learning point
- the fact that the cloud distribution implicitly preserves and reinforce the intellectual property & rights related to the application is very valuable for medical applications; this should be clearly explained to the medical community More details are available in D132 – Validation of MEDUSA and final impact assessment on healthcare.
5 Contouring for oncology demonstration Demonstration date: 12/11/2015 Location: Gustave Roussy Hospital Clinical experts present: Radiotherapist, Nuclear Medicine Physician Likes The concept to access to the application and to patient's data from anywhere • and not only from the interpretation room;
The concept to access to the MEDUSA framework using only a web browser;
• The capability for many users to interact in real time using the contouring tools.
• Recommendations In order to push MEDUSA framework to the routine, an effort is expected to improve the fluidity of image display. Especially, the MIP 3D view is not really usable within the MEDUSA framework when it is needful in nuclear medicine for tumor localization.
Learning point Contouring task for oncology is the most tedious task in order to prepare a treatment plan for radiotherapy. Many contouring applications have been developed for that purpose. The manufacturers always tried to facilitate this task by providing useful methods and ergonomic tools dedicated to fast and effective contouring.
Medusa will take over standard standalone applications if its workflow does not complicate or slow the way users work.
6 Computer assisted in vitro diagnostics demonstration Demonstration date: 25/11/15 Location: Imstar, Paris Clinical experts present: Michel Soussaline, MD Likes Expert appreciated the digital image quality, considering it the key argument for the replacement of the microscope use by the computer-based imaging. The interface of the virtual slide module is considered easy, intuitive and sufficiently responsive for everyday diagnostic work. Facility of the access to the slide archive in the digital form is a very positive point.
The quality of the automated processing is good; the cell segmentation and classification procedures miss virtually none of the abnormal cells. However, certain amount of the false positives (not abnormal cells pre-classified as possibly abnormal cells) is present after the automated processing.
The procedure of review and validation of the pre-selected cells, presenting them one by one and requiring the expert to either validate or reject them, is clear, easy to use and sufficiently responsive. Possibility to perform the review and validation process by multiple users at different locations at the same time is very much appreciated, as it simplifies the knowledge sharing and transfer.
Additional plus is the traceability of the review, validation and the diagnostic process;
as the system actually records the user choice concerning every cell, it is easy to return to the diagnostic data later if needed.
Recommendations The login procedure includes some repetitive steps (configuring the application list), it would be nice to store the configuration for every user.
Image data transfer from the microscope workstation to the storage used by the platform could be simplified.
Virtualisation of the microscope slides can enhance the international scientific/educational collaboration.
A deeper integration between the microscope workstation and the Medusa storage would be the next important thing to do
7 Collaborative decision support for oncology demonstration Demonstration date: 27/10/2015 Location: Pitié Salpétrière Hospital, Paris Clinical experts present: Prof. Philippe Grenier, Prof. Olivier Lucidarme The demonstration made to the clinical experts is recorded (see Google drive, see reference section for the movies) Likes The concept of lossy-to-lossless progressive quality access of the remote • DICOM images which allows clinical diagnosis to start before the end of data exchange.
There is total transparency for the clinical staff of the on-going data transfer.
• The clinical staff can concentrate on his diagnosis task.
Knowledge of the image quality level: any image investigated has the quality • level displayed as a color code. Even if visually the quality seems unaltered, the clinician can see if he is not looking at lossless quality data and can immediately request that level. This guarantees that wrong medical decisions cannot be due to the image quality.
In order to exploit this data exchange concept more largely and expand it to interconnecting small/private clinical settings, a mechanism requiring the remote DICOM data compression to JPEG2000 at the distant setting, prior to download it as progressive quality at the local setting, should be implemented.