Introduction:

Most of the published papers on AI based diagnosis have focused on the algorithm's diagnostic performance in a ‘binary’ setting (i.e. disease vs no disease). However, no study evaluated the actual value for the clinicians of an AI based approach in diagnostic. Detection of Traumatic thoracolumbar (TL) fractures is challenging on planar radiographs, resulting in significant rates of missed diagnoses (30-60%), thus constituting a field in which a performance improvement is needed. Aim of this study is therefore to evaluate the value provided by AI generated saliency maps (SM), i.e. the maps that highlight the AI identified region of interests.

Introduction and Objective

Up to 30% of thoracolumbar (TL) fractures are missed in the emergency room. Failure to identify these fractures can result in neurological injuries up to 51% of the casesthis article aimed to clarify the incidence and risk factors of traumatic fractures in China. The China National Fracture Study (CNFS. Obtaining sagittal and anteroposterior radiographs of the TL spine are the first diagnostic step when suspecting a traumatic injury. In most cases, CT and/or MRI are needed to confirm the diagnosis. These are time and resource consuming. Thus, reliably detecting vertebral fractures in simple radiographic projections would have a significant impact. We aim to develop and validate a deep learning tool capable of detecting TL fractures on lateral radiographs of the spine. The clinical implementation of this tool is anticipated to reduce the rate of missed vertebral fractures in emergency rooms.

Infection remains among the first reasons for failure of joint prosthesis. Currently, the golden standard for treating prosthetic joint infections (PJIs) is two-stage revision. However, two-stage procedures have been reported to be associated with higher costs and possible higher morbidity and mortality, compared to one-stage. Furthermore, recent studies showed the ability of a fast-resorbable, antibacterial-loaded hydrogel coating to reduce surgical site infections after joint replacement, by preventing bacterial colonization of implants. Aim of this study was then to compare the infection recurrence rate after a one-stage, cemenless exchange, performed with an antibacterial coated implant versus a standardized two-stage revision procedure.

In this two-center prospective study, 22 patients, candidate to revision surgery for PJI, were enrolled to undergo a one-stage revision surgery with cementless implants, coated intra-operatively with a fast-resorbable, antibiotic-loaded hyaluronan and poly-D,L-lactide based hydrogel coating (“Defensive Antibacterial Coating”, DAC, Novagenit, Italy). DAC was reconstructed according to manufacturer indications and loaded with Vancomycin or Vancomycin + Meropenem, according to cultural examinations, and directly spread onto the implant before insertion. This prospective cohort was compared with a retrospective series of 22 consecutive patients, matched for age, sex, host type, site of surgery, that underwent a two stage procedure, using a preformed, antibiotic-loaded spacer (Tecres, Italy) and a cementless implant. The second surgery, for definitive implant placing, was performed only after CRP normalization and no clinical sign of infection. Clinical, laboratory and radiographic evaluation were performed at 3, 6 and 12 months, and every 6 months thereafter. Infection recurrence was defined by the presence of a sinus tract communicating with the joint, or at least two among the following criteria: clinical signs of infections; elevated CRP and ESR; elevated synovial fluid WBC count; elevated synovial fluid leukocyte esterase; a positive cultural examination from synovial fluid; radiographic signs of stem loosening.

The two groups did not differ significantly for age, sex, host type and site of surgery (18 knees and 4 hips, respectively). The DAC hydrogel was loaded intra-operatively, according to cultural examination, with vancomycin (14 patients) or vancomycin and meropenem (8 cases). At a mean follow-up of 20.2 ± 6.3 months, 2 patients (9.1%) in the DAC group showed an infection recurrence, compared to 3 patients (13.6%) in the two-stage group. No adverse events associated with the use of DAC or radiographic loosening of the stem were observed at the latest follow-up months.

This is the first report on one-stage cementless revision surgery for PJI, performed with a fast-resorbable antibacterial hydrogel coating. Our data, although in a limited series of patients and at a relatively short follow-up, show similar infection recurrence rate after one-stage exchange with cementless, coated implants, compared to two-stage revision. These findings warrant further studies in the possible applications of antibacterial coating technologies to treat implant-related infections.

Development of antibacterial surfaces or coatings to prevent bacterial adhesion and hence colonization of implants and biofilm formation is an attractive option, in order to reduce the tremendous impact of implant-related infections associated with modern surgery. To overcome the lack of in vivo and clinical models, able to evaluate the performance of anti-adhesive coatings, we designed an in vitro experimental setting that allows to quantitatively evaluate the ability of a coating to reduce bacterial adhesion on a given surface; this model may efficiently serve as a surrogate endpoint to validate anti-adhesive medical devices and compounds. Here we report the results the evaluation of the anti-adhesive properties of a patented, fast-resorbable hydrogel coating, (“Defensive Antibacterial Coating”, DAC).

Sterile sandblasted titanium discs of approximately 5cm² surface area were used as substrates for bacterial adhesion. The gel was prepared as follows: syringes prefilled with 300 mg of DAC powder (Novagenit Srl) were reconstituted with 5 ml of sterile water to obtain a hydrogel with a DAC concentration of 6%. Two experiments were conducted. In the first, 200 mg of hydrogel were homogenously spread on the surface of titanium disc, with the spreading device provided by the manufacturer. Both coated and uncoated substrates (controls) were overlaid with a standardized inoculum (10⁸ CFU/ml) of a wild methicillin-resistant Staphylococcus aureus strain, previously isolated from a peri-prosthetic joint infection, for 15, 30, 60 and 120 minutes. Afterwards, non-adherent bacteria were removed by rinsing with sterile saline. The remaining adhered cells were seeded on agar plates for CFU count. In the second experiment, the discs were first inoculated with bacterial cells followed by a treatment with the hydrogel and bacterial count as described above. Ten discs were used for each condition and each time interval (total 160 discs).

The adhesion density of S. aureus on titanium discs pre-treated with DAC was significantly lower than that observed on untreated controls at each time point. In particular, the average number of adherent bacteria at 15, 30, 60 and 120 minutes of incubation, was respectively reduced by 86.8%, 80.4%, 74.6% and 66.7%, compared to controls (p<0.001). DAC treatment of discs with previously adhered S. aureus reduced bacterial adhesion, at 15, 30, 60 and 120 minutes of incubation, by, respectively, 84.0% (p<0.05), 72.8%, 72.3% and 64.3% (p<0.001), compared to untreated controls.

Our results shows that DAC, “Defensive Antibacterial Coating”, has anti-adhesive properties that allow to reduce bacterial adhesion on a sanded titanium surface by more than 80%, even in the presence of remarkably high bacterial loads (10⁸ CFU/ml), of multi-resistant bacteria (MRSA) and even in the case of previous contamination. Providing anti-adhesive properties to a surface with a fast-resorbable coating may be a safe option to protect inorganic and organic surfaces and biomaterials. Those observation could be the pre-requisite for its in vivo application.