The proposed study is addressed to introduce new Image-Guided Surgery (IGS) tools to assist mini-invasive surgical procedures of anterior rectal resection (TME; TA-TME, TTSS) performed with laparoscopic or robotic procedure. In details, the idea is to provide augmented reality (AR) guidance during robotic-assisted and laparoscopic surgeries, by overlaying the preoperative 3D virtual anatomical models to intraoperative surgical images (3D AR guidance).

To optimize the intraoperative view during the 3D AR guidance, AI-based algorithms will be developed to allow the real time detection and segmentation of surgical instruments, needed to provide instrument de-occlusion during AR robotic surgery.

We use 3D modelling technology in surgical planning (7 case) and intra-operative navigation ( 2 cases). The pilot study focused to 3D virtual reconstruction of the pelvis, rectum and neurovascular structure to the test the feasibility of virtual reality to this type of anatomy. Implementation of reconstruction using 3D nerve sequence (3 Tesla MRI) was used for the last 3 cases.

After the creation of a complete virtual model of pelvis and its structures the last two models were applied in the operating-room during a laparoscopic rectal resection with the ausilium of AI. The test has showed good results: a good overlap of the 3D structures to the real organs of the pelvis. The focus of this research was on developing support tools aimed at enhancing surgical safety. AR can assist surgeons in identifying vascular and nerve structures that are not always clearly visible during minimally invasive procedures, compensating for the lack of tactile perception and thereby improving overall surgical safety.

The next step of the study is to evaluate its benefits and limitations in clinical practice to reduce the post-operative complications and oncological recurrence.

In Italy, colorectal cancers account for 13% of newly diagnosed cancers per year, making it the third most frequent neoplasm in men (after lung and prostate cancers) and the second most frequent in women (after breast cancer); the second most frequent in terms of mortality in both sexes. Of these, about 30% can be traced to a rectal location .

Although over the years there has been a progressive development in chemo-radiotherapy and the introduction of new procedures and increasingly conservative protocols, to date demolitive surgery represents the gold standard in the radical treatment of rectal neoplasms, despite being one of the most complex operations to be performed in high-volume centres.

Surgical outcomes have improved considerably over the last 20 years due to the introduction of total mesorectal excision (TME), which consists of the complete removal of the mesorectum (the lymphoadipose tissue surrounding the rectum) while preserving the pelvic nerve plexuses. Despite improvements in preoperative staging, knowledge of anatomy, and the introduction of neo-adjuvant therapy, the recurrence rate of rectal cancer and postoperative genitourinary dysfunction still remains high. In recent years, there is an increasing interest in image-guided surgery (IGS) to improve the effectiveness of surgery and postoperative outcomes.

Primary objective:

To test whether the use of virtual and augmented reality, based on virtual reality reconstructions, during minimally invasive surgery for rectal cancer can allow a reduction in circumferential margin oncology positivity rates and an increase in the completeness of total mesorectal excision, thus representing an overall improvement in technical-surgical efficacy compared to the approach without augmented reality.

Secondary objectives:

To test whether the application of virtual reality and augmented reality allows a reduction in complications related to rectal resection surgery and postoperative mortality.

ENDPOINT/S:

Primary endpoints

• Circumferential margin oncological positivity rate (CRM) (margin of healthy tissue between any neoplastic structure present in the mesorectum and the mesorectal resection margin itself, which is considered positive when less than 1 mm) measured on anatomopathological assessment of the tumour
• Completeness of total mesorectal excision (TME) understood as the finding on anatomopathological examination of completely intact mesorectal fascia.

Secondary endpoints

Complications related to rectal resection surgery that will be examined are:

• Rate of genitourinary dysfunction 1 year after surgery (assessed by administration of questionnaires: International index of Erectile Funcion for men; PISQ-12 for women.
• Intraoperative and postoperative complication rate at 30 days and their severity (measured according to the Clavien-Dindo classification (28).
• Dehiscence rate of the colo-rectal or colo-anal anastomosis at 1 year after surgery, understood as discontinuity of the suture rhyme evidenced by the presence of at least one of these elements alteration of the clinical picture characterised by abdominal pain; fever or appearance of enteric material from abdominal drains; alteration of laboratory tests; elevation of phlogosis indices; evidence on abdominal CT of perianastomotic collections or extraluminal presence of orally administered mdc.
• Mortality at 30 days after surgery The study is experimental with historical control. The design involves the comparison of perioperative, oncological and functional outcomes in two cohorts of patients undergoing surgical treatment of rectal cancer: a retrospective cohort of patients undergoing 'classic' minimally invasive surgery and a prospective cohort undergoing minimally invasive surgery with the application of 3D modelling, VR and AR.

STUDY POPULATION The cohort of patients undergoing rectal surgery with the aid of augmented reality will be prospectively enrolled consecutively indicatively from 01/01/2024 and in any case after the approval of this study by the CE AVEC and the issuance of the General Management clearance.

The comparison cohort will be constructed by identifying patients treated with minimally invasive rectal surgery between January 1, 2017 and December 31, 2023 , and matched 2:1 with the group of patients treated with AV and AR surgery for the following characteristics: gender, age group, oncological stage, possible use of neoadjuvant therapy.

INTERVENTION/TREATMENT UNDER STUDY The preliminary step for the creation of a three-dimensional model of an anatomical district is the segmentation of a diagnostic image of the patient (e.g. CT or MRI), acquired as per standard clinical practice.

Commercial software (D2P from 3D Systems; Mimics from Materialise), which is CE (and FDA) certified as DM, class IIa (II), is used to perform the segmentation of CT/RM images. 3D modelling or '3D modelling' is the process of defining a three-dimensional shape in a virtual space, generated on a computer. The 3D models, obtained as the final step of segmentation, can be processed (e.g. execution of cuts, unions/volume intersections, realisation of cavities, shells, etc.) using CE/FDA-certified 3D modelling software for medical use (3-Matic, Materialise), in order to prepare them for surgical simulation. The validation of the 3D model will be performed by the radiologist and the clinician by comparing the reconstructed structures with those identified by CT and MRI sequences.

For the study in question, the 3D anatomical models reconstructed from the patient's imaging can be explored by the surgeon in the preoperative phase through standard 2D monitors and immersive VR vision systems (e.g. software package 'D2P Advanced VR (Volume and Mesh views)' by 3D Systems/Oqton and relative Vive viewer with control handpieces), also available within the eDIMES Lab and under proprietary licence For the intra-operative visualisation of the virtual 3D anatomical model, superimposed in AR on the laparoscopic intra-operative image, a medical PC (Werth PRO 600, Noris Medical devices networking) will be used, equipped with the relative medical keyboard and mouse (medical silicone keyboard mod. in/sk307, optical mouse IN/SM502 in sanitised silicone) and a medical monitor (MD-2402 clinical review monitor, Neovo). See details in the relevant Technical Annexes made available.

The medical PC will be connected to the laparoscopy column in the operating room for the sole purpose of taking the video signal from the endoscope of the laparoscopy column. The intraoperative video image of the endoscope will be used within the medical PC to obtain the AR visualisation, i.e. to superimpose the 3D virtual model reconstructed in the preoperative phase on this intraoperative image. It is specified that the medical PC used in the study does not require a connection to the hospital network or interfacing with the SIO (hospital information system).

By means of a commercial software for live video streaming (vMIX, StudioCoast Pty Ltd, Robina, Queensland, Australia), the 3D virtual anatomical model will be superimposed in AR on the surgical video stream of the endoscope and aligned to the corresponding real structure by bioengineering technical staff present in the operating room (previously authorised). For navigation and manipulation of the virtual 3D model, a 6-degree-of-freedom mouse (SpaceMouse, 3D Connexion, Munich, Germany) may be used in addition to the medical optical mouse associated with the PC, which will be appropriately bagged with a sterile surgical instrument bag.

The result of the superimposition will be displayed on the medical monitor associated with the PC, or it can be transferred to one of the monitors already present in the room, in parallel with the standard intra-operative laparoscopic image display only. Therefore, the AR display will in no way alter the normal intra-operative laparoscopic display, which will still be available to the surgeon on a room monitor, as per normal care pathway.

Prior to use in the operating room, the connection between the medical PC introduced to take the laparoscopic image and the laparoscopic column in the operating room will be tested by the Clinical Engineering Service of the Experiment Centre, in order to carry out mainly electrical safety checks.

The above-mentioned technologies for 3D modelling, VR/AR vision and the medical PC used for intraoperative AR visualisation will be made available by the Study and Research Centre (CSR) eDIMES Lab of the Department of Medical and Surgical Sciences (DIMEC), University of Bologna.

Study duration The study includes a prospective recruitment phase lasting 3 years followed by 1 year of follow-up for the evaluation of outcomes. At the end of the prospective recruitment, during the last year of follow-up, the match will be performed with patients from the retrospective cohort. This will be followed by a 3-month period for data analysis.

The end date is scheduled for December 2027

VISITS AND EVALUATIONS The first follow-up will be performed at 30 days after surgery to assess short-term complications (anastomosis dehiscence, reintervention, complications according to Clavien Dindo's classification). Thereafter, follow-up will be completed at one year after surgery at three-month intervals according to normal oncological follow-up protocols for patients undergoing rectal surgery for primary cancer, in order to assess functional and oncological outcomes.

Inclusion Criteria:

• Indication for radical transabdominal surgery for primary rectal cancer
• Resection and anastomosis or abdominal-perineal amputation surgery
• Signature of informed consent
• Patient's age ≥18 years.
• CT images acquired in the arterial venous and urographic phase with section thickness:1.25/2.5 mm, level range:0.8/0.2 mm.
• Pelvic MRI images acquired with section thickness of 1.5 mm or images acquired with 3 tesla MRI.

Exclusion Criteria:

• Patients who have already undergone previous rectal surgery
• Relapse of previous rectal neoplasm
• Neoplasm located in other pelvic organs, infiltrating the rectum
• Indication for endoscopic or transanal treatment
• Disease with peritoneal localisation (carcinosis)
• Surgery for palliative purposes
• Imaging performed elsewhere