Beneficiaries
       SURG (Lead)                                
       AUH (co-Lead)
       AU
       SWU
       RTU
       IAAD
       CORR                                                                                  
                                                                           WP7   2
                             Objectives
                 To annotate a large dataset of laparoscopic videos.
                 To create an algorithm that enables automatic detection of the 
                 division plane around the endometriotic lesion.
                 To integrate the algorithm into SurgAR software for real time 
                 augmented reality guidance of the surgeon in finding the division 
                 planes around the lesion.
                                                                                                       
                                                                                     WP7     3
                                    Tasks
                1 . Collecting videos with metadata [M3-M9]
                2.  Semantic annotation of laparoscopic videos [M9-M25]
                3. Deep learning algorithm dedicated to automatic detection of 
                division plane [M25-M33]
                4. Augmented reality to guide surgeons real time on the laparoscopic 
                screen [M33-M40]
                5. Integration and clinical validation of the algorithm in SurgAR 
                software [M40-M45]
                                                                                                          
                                                                                        WP7     4
                            Deliverables
        D.7.1 – Dataset of 1,000 laparoscopic dicomised videos [M9]
        Lead: AUH (P2) 
        Type: Videos
        D.7.2 – Dataset of 100,000-200,000 laparoscopic semantic annotated images [M25]
        Lead: SUW (P5) 
        Type: Images
        D.7.3 – Deep learning algorithm to automatically determinate the division plane [M33]
        Lead: RTU (P8)
        Type: Algorithm
        D.7.4 – Computer vision-based software for real time guided surgery of endometriosis [M45]
        Lead: SURG (P7)
        Type: Software                                                                                     
                                                                                         WP7     5
                                                       Data transfer procedure
                                                                                              Hospital
                                                                  Operating room
                                                                    laparoscopes                                  Virtual machine
                                                                    Video recording                                                                              SurgAR 
                                                                                                                                                                 storage
                                                             Hospital                                        Stored 
                                                             computer                                                           Anonymisation 
                                                                                                           dicomized            + export 
                                                                                                             videos
                                                                Identifying data removal
                                                                dicomization
                                                               Tools used by the                                       Automated process
                                                                     hospital
           •     Pipeline for transfer of anonymized video between a public healthcare structure and SurgAR
            
           •     Medexprim is an expert in secure medical data transfer
           •     Used in a three year collaboration with Clermont-Ferrand University Hospital