Igor Paredes

Surgical Phase Recognition (SPR) enables real-time, context-aware assistance during surgery, but its use remains limited by the cost and effort of dense video annotation. This study presents a Semi-Supervised Deep Learning framework for SPR in endoscopic pituitary surgery, aiming to reduce annotation requirements while maintaining performance. A Timestamp Supervision strategy is employed, where only one or two representative frames per phase are labeled. These labels are then propagated, creating pseudo-labels for unlabeled frames using an Uncertainty-Aware Temporal Diffusion (UATD) approach, based on confidence and temporal consistency. Multiple spatial and temporal architectures are evaluated on the PituPhase–SurgeryAI dataset, the largest publicly available collection of endoscopic pituitary surgeries to date, which includes an outside-the-body phase. Despite using less than 3% of the annotated data, the proposed method achieves an F1-score of 0.60 [0.55–0.65], demonstrating competitive performance against previous Supervised approaches in the same context. Removing the recurrent outside-the-body phase reduces misclassification and improves temporal consistency. These results demonstrate that uncertainty-guided Semi-Supervision is a scalable and clinically viable alternative to fully Supervised Learning for surgical workflow analysis.

Pituitary apoplexy is a clinical syndrome that can cause vision loss and oculomotor (OM) deficits. Neurosurgical decompression is still considered the treatment of choice for those with visual impairment. We aim to explore the influence of timing of neurosurgical decompression on visual outcome. Multicentre retrospective study of patients presenting with pituitary adenoma and pituitary apoplexy who were operated in three Spanish tertiary hospitals between 2008 and 2022. We identified 49 patients with pituitary apoplexy with visual symptoms and were operated. Among the visual symptoms, 35 patients (73%) had visual acuity (VA) impairment and 33 (67%) had an OM nerve paresis. Twenty-three of the 35 patients improved their VA, and 24 of the 33 improved their OM nerve paresis after surgery. The median time between apoplexy to surgery was 10 days (range 0-164). Regarding visual acuity, 50% (3 out of 6) of those operated within 3 days of apoplexy showed improvement, compared to 69% (20 out of 29) of those operated beyond (p = 0.37).Regarding OM paresis, 62.5% (5 out of 8) of those operated within 3 days of apoplexy improved, compared to 76% (19 out of 25) of those operated beyond (p = 0.56).Mean apoplexy to surgery time was 18 ± 19 days in those whose VA improved (vs. 32 ± 51; p = 0.46); and mean apoplexy to surgery time for those whose OM symptoms improved was 21+/- 35 days (vs. 24 +/- 37; p = 0.86). Our results suggest that timing of neurosurgical decompression does not affect the visual outcome of patients with pituitary apoplexy.

Background There is skepticism about the benefit of surgery in elderly patients affected by traumatic brain injury (TBI) due to the negative effect of age on the outcome and surgical complications. However, there are few studies that have investigated differences in patient’s outcome between surgically and conservatively managed patients after adjusting for the imbalance in preinjury characteristics and clinical and radiological features. The primary aim of this study was to evaluate the effect of early surgery on mortality and functional recovery in a cohort of older adults with acute traumatic intracranial lesions after adjustment by Propensity Score (PS) matching. Materials and methods We conducted a retrospective cohort study on older adult patients (≥ 65 years) admitted for TBI between 2013 and 2023 to a single level 1 trauma center. Patients were categorized based on whether they underwent early surgery (< 48 h after TBI) for a space-occupying lesion evacuation. PS model was constructed based on age, frailty, comorbidities (Charlson comorbity index and American Society of Anaesthesiologists score), anticoagulants, hypoxia, shock, pupillary abnormalities and GCS motor response upon admission, midline shift, basal cistern effacement, volume of subdural and intracerebral hematomas, and limitation of life-sustaining treatment decisions.The effect of early surgery on 30-day mortality and unfavorable functional outcomes (GOSE 1–3) at 6 and 12 months were investigated after matching by paired test. Results We identified and reviewed 301 patients who met all inclusion criteria and contained no exclusions. After matching, 62 patients (31 pairs of conservative and surgical patients) remained as the matched datasets. Our key finding was that older adult TBI patients who underwent early surgery had a statistically significant reduction in the risk of 30-day mortality (OR 0.313, 95% CI 0.114–0.853, p = 0.023) and unfaourable outcome at 12 months after TBI (OR 0.286, 95% CI 0.094–0.868, p = 0.027). Conclusions Early surgery was associated with decreased 30-day mortality and better functional outcome at 12 months after TBI in older adults with few comorbidities and good functionality when clinically affected by acute traumatic intracranial lesions with mass effect.

Background and Purpose: Syndrome of the trephined or sinking skin flap syndrome is an underdiagnosed condition of craniectomized patients that usually improves after cranioplasty. Among the pathophysiological theories proposed, the changes of cerebral blood perfusion (CBP) caused by cranial defects might have a role in the neurological deficiencies observed. We aim to assess the regional cortex changes in CBP after cranioplasty with Technetium 99m hexamethylpropylene‐amine oxime (99mTc‐HMPAO) SPECT‐CT.Methods: Twenty‐eight craniectomized patients subject to cranioplasty were studied with 99mTc‐HMPAO SPECT‐CT in three different times, before cranioplasty, a week, and 3 months after. The images were processed with quantification software comparing CBP of 24 cortical areas with a reference area, and with a database of controls. A mixed effects model and T‐Student were used.Results: CBP increased significantly in both hemispheres after cranioplasty, either using ratio (β = .019, p‐value = .030 first postsurgical SPECT‐CT and β = .021, p‐value = .015 in the second study, vs. presurgical) or Z‐score (β = .220, p‐value = .026 and β = .279, p‐value = .005, respectively). Nine areas of the damaged side had a significant lower CBP ratio and Z‐score than the undamaged. Posterior cingulate showed an increased CBP ratio (p‐value = .034) and Z‐score (p‐value = .028) in the first postsurgical SPECT‐CT. These posterior cingulate changes represent a 4.83% increase in ratio and 91.04% in Z‐Score (p‐value = .035 and .040, respectively).Conclusion: CBP changes significantly in specific cortical areas after cranioplasty. Posterior cingulate changes might explain some improvements in attention impairments. SPECT‐CT could be a useful tool to assess CBP changes in these patients and might be helpful in their clinical management.

Objectives The large number of infected patients requiring mechanical ventilation has led to the postponement of scheduled neurosurgical procedures during the first wave of the COVID-19 pandemic. The aims of this study were to investigate the factors that influence the decision to postpone scheduled neurosurgical procedures and to evaluate the effect of the restriction in scheduled surgery adopted to deal with the first outbreak of the COVID-19 pandemic in Spain on the outcome of patients awaiting surgery. Design This was an observational retrospective study. Settings A tertiary-level multicentre study of neurosurgery activity between 1 March and 30 June 2020. Participants A total of 680 patients awaiting any scheduled neurosurgical procedure were enrolled. 470 patients (69.1%) were awaiting surgery because of spine degenerative disease, 86 patients (12.6%) due to functional disorders, 58 patients (8.5%) due to brain or spine tumours, 25 patients (3.7%) due to cerebrospinal fluid (CSF) disorders and 17 patients (2.5%) due to cerebrovascular disease. Primary and secondary outcome measures The primary outcome was mortality due to any reason and any deterioration of the specific neurosurgical condition. Second, we analysed the rate of confirmed SARS-CoV-2 infection. Results More than one-quarter of patients experienced clinical or radiological deterioration. The rate of worsening was higher among patients with functional (39.5%) or CSF disorders (40%). Two patients died (0.4%) during the waiting period, both because of a concurrent disease. We performed a multivariate logistic regression analysis to determine independent covariates associated with maintaining the surgical indication. We found that community SARS-CoV-2 incidence (OR=1.011, pConclusions Patients awaiting neurosurgery experienced significant collateral damage even when they were considered for scheduled procedures.

Objective: To assess the effect of the first wave of the SARS-CoV-2 pandemic on the outcome of neurosurgical patients in Spain. Settings: The initial flood of COVID-19 patients overwhelmed an unprepared healthcare system. Different measures were taken to deal with this overburden. The effect of these measures on neurosurgical patients, as well as the effect of COVID-19 itself, has not been thoroughly studied. Participants: This was a multicentre, nationwide, observational retrospective study of patients who underwent any neurosurgical operation from March to July 2020. Interventions: An exploratory factorial analysis was performed to select the most relevant variables of the sample. Primary and secondary outcome measures: Univariate and multivariate analyses were performed to identify independent predictors of mortality and postoperative SARS-CoV-2 infection. Results: Sixteen hospitals registered 1677 operated patients. The overall mortality was 6.4%, and 2.9% (44 patients) suffered a perioperative SARS-CoV-2 infection. Of those infections, 24 were diagnosed postoperatively. Age (OR 1.05), perioperative SARS-CoV-2 infection (OR 4.7), community COVID-19 incidence (cases/10 ⁵ people/week) (OR 1.006), postoperative neurological worsening (OR 5.9), postoperative need for airway support (OR 5.38), ASA grade ≥3 (OR 2.5) and preoperative GCS 3–8 (OR 2.82) were independently associated with mortality. For SARS-CoV-2 postoperative infection, screening swab test 5 people/week) (OR 1.011), preoperative cognitive impairment (OR 2.784), postoperative sepsis (OR 3.807) and an absence of postoperative complications (OR 0.188) were independently associated. Conclusions: Perioperative SARS-CoV-2 infection in neurosurgical patients was associated with an increase in mortality by almost fivefold. Community COVID-19 incidence (cases/10 ⁵ people/week) was a statistically independent predictor of mortality. Trial registration number: CEIM 20/217.

Cranioplasty after decompressive craniectomy (DC) has been found to improve the neurological condition. The underlying mechanisms are still unknown. The aim of this study is to investigate the roles of the postural changes and atmospheric pressure (AP) in the brain hemodynamics and their relationship with clinical improvement. Seventy-eight patients were studied before and 72 h after cranioplasty with cervical and transcranial color Doppler ultrasound (TCCS) in the sitting and supine positions. Craniectomy size, shape, and force exerted by the AP (torque) were calculated. Neurological condition was assessed with the National Institutes of Health Stroke Scale (NIHSS) and the Barthel index. Twenty-eight patients improved after cranioplasty. Their time elapsed from the DC was shorter (214 vs 324 days), preoperative Barthel was worse (54 vs 77), internal carotid artery (ICA) mean velocity of the defect side was lower while sitting (14.4 vs 20.9 cm/s), and torque over the craniectomy was greater (2480.3 vs 1464.3 N*cm). Multivariate binary logistic regression showed the consistency of these changes. TCCS findings were no longer present postoperatively. Lower ICA (defect side) velocity in the sitting position correlates significantly with clinical improvement. Greater torque exerted by the AP might explain different susceptibilities to postural changes, corrected by cranioplasty.